U.S. Army Medical Department, Office of Medical History
Skip Navigation, go to content

HISTORY OF THE OFFICE OF MEDICAL HISTORY

AMEDD BIOGRAPHIES

AMEDD CORPS HISTORY

BOOKS AND DOCUMENTS

HISTORICAL ART WORK & IMAGES

MEDICAL MEMOIRS

AMEDD MEDAL OF HONOR RECIPIENTS External Link, Opens in New Window

ORGANIZATIONAL HISTORIES

THE SURGEONS GENERAL

ANNUAL REPORTS OF THE SURGEON GENERAL

AMEDD UNIT PATCHES AND LINEAGE

THE AMEDD HISTORIAN NEWSLETTER

Chapter XI

Contents

949

SECTION III

NEUROSURGERY

CHAPTER XI

TECHNIQUE OF NERVE SURGERY

INTRODUCTION 

In presenting a system of technique on the operative treatment of peripheral nerve lesions, it is assumed that the reader is familiar with the histopathology and pathologic physiology of peripheral nerve tissue when visited with disease or subjected to injury: it is only on the basis of a thorough understanding of the histologic and physiologic principles involved that a rational operative treatment may be founded. Fortunately, these principles, placed pon a sound footing more than two decades ago, have been amply verified by subsequent experimentation and consistently upheld by the innumerable clinical observations afforded by the vast amount of material supplied by the casualties of the World War. The conclusions arrived at by Howell and Huber in 1891, in their many experimental studies of peripheral nerve physiology, degeneration, and regeneration may still serve as a general text for a rational system of procedure in the operative treatment of peripheral nerve lesions.

Probably in no other type of surgery will the end results so emphatically demonstrate the necessity of subserving physiologic principles. It has been generally known that the fibers of one nerve trunk, when anastomosed to the distal segment of another nerve, will regenerate satisfactorily through the adopted nerve: this fact was utilized in hypoglosso-facial anastomosis for the correction of facial paralysis, though the principle governing this procedure was not based upon a complete physiologic appreciation of all the factors involved. This operation demonstrates the principle of anatomic surgery, in which fibers are supplied to the denervated muscles of the tongue by the anastomosing of one nerve into another. The physiologic principle ignored in this type of anastomosis is manifest in the end results, namely, the paralyzed facial muscles, though having regained some motor power, still remain immobile to emotional facial expression, contracting synchronously with lingual movements during mastication or deglutition. While the paralyzed facial muscles were re-innervated, the innervation from a physiologic standpoint was doomed to failure, because the new fibers innervating the facial musculature are hypoglossal fibers, capable of subserving hypoglossal function only-motor function to the tongue--which is in no way correlated to the emotional reactions expressed through the facial musculature. This represents but one type of physiologic transgression in the surgery of nerves.

 Most peripheral nerves are both motor and sensory in functions; the motor portion of the nerve trunk, derived through a neuraxon outgrowth from the motor cells in the anterior horns of the spinal cord, carry efferent motor impulses: the sensory portion of the nerve trunk is composed of fibers which carry


950

sensory or afferent impulses through the posterior horns of the cord. These motor and sensory fibers, though anatomically incorporated within a single nerve trunk, are from a physiologic standpoint totally unrelated; this well-known physiologic fact, when given due consideration, immediately elevates the operation of nerve suture from the domain of anatomic to that of physiologic surgery. The anatomic principle of simply obtaining a satisfactory end-to-end approximation of a divided peripheral nerve, adequate to permit the passage of regenerating fibers, does not suffice in nerve surgery, for it may, from a physiologic standpoint, be very imperfect. in that the motor fibers in regenerating may enter sensory channels leading to sensory terminations, and these motor fibers, not being able to subserve sensory function, would be physiologically lost.

A study of the end results following nerve suture has amply demonstrated that perfect anatomic sutures are frequently attended by only a partial restoration of function, and occasionally by its total absence. At best, the unqualified anatomic principle of nerve suture, without regard to the physiologic differentiation of fibers, can be nothing more than a hit-or-miss method of procedure. If in the approximation of the ends of the divided nerve the surgeon is fortunate enough to approximate motor fibers to the motor channels in the distal segment and sensory fibers to their respective channels, the ultimate results-if regeneration is unimpeded in both nerve and muscle-should be a complete restoration of function. The possibility of a physiologic approximation, if left wholly to chance as in an ungoverned anatomic suture, can be seldom expected, though in the majority of instances some motor fibers will probably (by chance) reach motor channels and regenerate to a motor termination; no doubt many will be lost, and our experience seems to indicate that the regenerating motor and sensory fibers are incapable of selecting their respective physiologic channels. The principle of physiologic approximation resolves itself almost entirely into the prevention of torsion of the nerve trunk during suture; the most important factor involved in the technique of nerve surgery depends upon our ability to prevent or correct this unfortunate incident. While end-to-end approximation is absolutely essential to neuraxon regeneration, physiologic approximation is indispensable to the restoration of function.

An adequate conception of physiologic principles will immediately enable the surgeon to disregard many of the older operative procedures of nerve anastomosis, more or less brilliant from a purely anatomic standpoint. hut absolutely futile when considered in the light of our present knowledge. An appreciation of the fact that each fiber within a nerve trunk has to a greater or lesser degree some functional individuality should lead the surgeon to a realization that, though it is within his power to change the anatomic course of these fibers, their physiologic attributes remain unchanged; at best, his most painstaking and careful operative manipulation, however ingenious from an anatomic standpoint, loses in functional value in direct proportion to the perversion of the original physiologic pattern. Satisfactory end results in nerve surgery depend primarily upon a full appreciation of physiologic function and our ability to conserve the anatomic characteristics compatible with the normal expression of that function.


951

THE GENERAL TECHNIQUE

ANATOMIC REQUIREMENTS

Primarily it is essential that the operator have that practical working knowledge of peripheral nerve anatomy which is acquired and maintained only by frequent recourse to the dissecting room. It is particularly important that he be familiar not only with the general course and relationship of nerve trunks, but also the origin and distribution of the various branches its they arise; it is not only essential that these branches be preserved in dissection, but they must also be exposed and accurately identified in order to determine, by their intraneural position, the intraneural topography of the nerve trunk. While no dogmatic rule may be laid down regarding the point of separation-superficial origin-of branches from the parent trunk, their intraneural course, as a rule, is sufficiently constant to permit a fair degree of accuracy in topographical identification. It is usually necessary that the particular branch under consideration be liberated for a short distance up the nerve trunk, in order that its exact intraneural position be accurately determined. Occasionally a motor branch may have its superficial origin from the sensory side of the nerve trunk, but its deep intraneural position will usually be found to be in that sector of the nerve trunk which contains the motor bundles. The exposure and identification of motor and sensory branches, arising from a peripheral nerve, is of exceedingly great value in topographical localization, as their intraneural localization will nearly always indicate that portion of the nerve trunk which subserves their respective function, either motor or sensory.

PREPARATION OF AN EXTREMITY FOR OPERATION

It is important in nerve operations that the skin of the entire extremity he sterilized from a point some distance above the proposed incision. The necessity of sterilizing the peripheral portion of the extremity is due to the fact that manipulation of the entire limb is frequently necessary during an operation; the observation of the entire limb during electrical stimulation of the nerve is essential to determine the action of individual muscles. This is particularly important in the hand. Occasionally a sterile rubber glove may be used to cover the hand of the patient, when adequate sterilization is doubtful. The glove usually will permit observation of muscle action. The draping of an extremity should always he arranged in such a manner as will facilitate observation and manipulation without contamination; the surgeon should personally direct the placing of protective sheets and towels with this object in view.

POSITION

The position of an extremity, for operation upon one of its principal nerves, should be planned with as much forethought as might be given to any other step of the operation; the surgeon should endeavor, by the effective placing of sand bags, to maintain the extremity in a position conducive to the comfort of the patient as well as to his own convenience. A third assistant, or a sterile nurse with specific instructions, should be retained to maintain the extremity in the desired position of flexion, when flexion-relaxation is utilized in the correction of continuity defects.


952

ANESTHESIA

Local anesthesia, in the writer's experience. has supplanted all other types of anesthesia for peripheral nerve work. Its proper use is attended with complete analgesia, with the retention of motor function. It permits the use of the electro-anatomic method of funicular identification; it greatly diminishes the oozing encountered in scar tissue dissections: and avoids the disagreeable complications attending inhalation anesthesia.

The following technique is used in produtinlg local anesthesia: At 1 percent solution of novocaine or procaine is used, to each ounce of which 15 minims of a fresh 1:1,000 adrenalin chloride solution has been added. The line of proposed skin incision is carefully infiltrated, producing a continuous line of intradermnal wheals. The needle is now passed into the subcutaneous tissue for a depth of 112 cm., where 1 c.c. of the solution is injected; the needle is then directed deeper and a second injection is made. Without withdrawing the needle, which should be sufficiently long to reach the nerve, these injections are made at an increasing depth until the tissues surrounding the nerve have been reached. These subcutaneous injections are made about 2 or 2 ½ cm. apart, along the entire line of the proposed incision. At each injection, in deep infiltration, particularly in the region of a large vessel, the piston of the syringe should be slightly withdrawn before any of the solution is injected, to determine the possibility of an intravascular penetration; if blood is aspirated the needle should be redirected. It is usually neither necessary nor desirable to inject the nerve until it is ready for section. unless pain is experienced. After the field has been thoroughly infiltrated, pressure is used for a few minutes to prolmote the diffusion of the solution througth the surrounding tissues. If it be necessary to carrv a dissection to the surface of a bone, its periosteum should be infiltrated. When the incision and surgical manipulations are confined to the infiltrated areas, they are absolutely painless and frequently bloodless; and, as most peripheral nerve operations are attended with much scar tissue dissection, this bloodless field is highly desirable.

EXPOSURE OF THE NERVE

The importance of long incisions and adequate exposures can not be too greatly emphasized in the surgery of peripheral nerves. It has usually been our experience that at seemingly adequate incision will require lengthening before the operative procedure is completed. This possibility should always be borne in mind during the preparation of the skin anil draping of the extremity for by so doing the surgeon will avoid the risk which attends invasion of parts not adequately prepared.

One of the greatest mistakes a surgeon can make is to attempt the exposure of a nerve in a scar-invaded region; for safety's sake alone, without a consideration of other advantages, the surgeon should make it his invariable rule to expose the nerve above and below the lesion, when the location of the lesion will permit. The nerve, having been exposed in a region where normal anatomic relationships prevail, may then be followed through the scar-invaded area where these anatomic guides have been distorted or lost.


953

The prevention of scar tissue following nerve operations is of paramount importance; and while it cannot be wholly obviated, there is no doubt but that it may be greatly mitigated by a careful operative technique. Scar tissue following surgical intervention is usually the result of tissue trauma, inadequate hemostasis, and infection. Tissue trauma may be reduced to a minimum by careful, clean-cut dissections, confined when possible to normal lines of cleavage. A common source of tissue trauma lies in the use of artery forceps and the ligation of vessels. When it is necessary to clamp a vessel, no more tissue than is absolutely necessary should he crushed to secure the vessel. The habit of crushing, and ligating the entire area around a vessel is greatly to be deplored, as it is accompanied by considerable tissue strangulation and subsequent necrosis, inviting infection and scar formation. Blunt and finger dissection, which is so frequently associated with tissue tearing and extensive trauma has no place in the surgery of peripheral nerves. The use of antiseptics, such as iodine, within the wound is to be condemned. It is desirable to cover the skin edges with clamped towels, as complete skin sterilization is often questionable. Parts of the incision which are not receiving immediate attention many be prevented from drying and protected from contamination by cotton pads saturated with warm saline. Scar tissue dissections, which under general anesthesia would be associated with constant oozing, are often rendered blood- less by infiltration anesthesia, and by the time the constricting action of the adrenalin has disappeared clotting has occurred in those oozing points which so often defy ligation. A time-saving procedure in the dissection of neurovascular bundles is preliminary control of the circulation by isolation of the veins below the lesion and the artery above, which structures may be temporarily constricted by tape to control hemorrhage following vascular accidents. Nothing is more deplorable than to see a surgeon blindly attempting to grasp an unseen bleeder in a pool of blood: in nerve surgery such a procedure is fraught. with so great a degree of danger to nerve trunks and branches that it is inexcusable, in that it may usually be obviated by using sufficient precaution for vascular control.

DEFECTS IN NERVE CONTINUITY

When at nerve is divided its ends tend to retract in the surrounding loose connective tissue, where they become fixed with the process of healing and the formation of scar tissue, making a defect in nerve continuity. In certain instances a portion of the nerve trunk may be completely destroyed or traumatized to such an extent as to obliterate its recognizable anatomic continuity. More frequently, however, the nerve so suffers from the devastating effect of infection and subsequent scar-tissue proliferation that large sections must be sacrificed before normal-appearing nerve bundles are found in its ends.

Various procedures have been recommended for the correction or filling of nerve defects, many of which have little, if any, physiologic grounds for support. Technically, some of the procedures recommended for the filling of defects, such as the interposition of tubes and various other foreign materials, may appear to be very ingeniously conceived, but from a physiologic standpoint they were often so defective that the probability of recovery was actually


954

minimized by the procedure. To permit nerve regeneration, direct anatomic approximation of the nerve ends must he attained. In certain instances when this is found to be impossible, the use of the autogenous cable graft is justified for filling defects, but at best it seems to promise but a very moderate degree of functional restoration. The interposition of heterogenous nerve grafts has been so universally unsuccessful in clinical experience that their use is to be discouraged. Various procedures, such as nerve implantation into neighboring nerves and the neurotization of paralyzed muscles, have so little to recommend them from a practical standpoint at the present time that they may be dismissed from the category of effective clinical procedures.

The following methods of correcting continuity defects in peripheral nerves are the only ones whose functional end results, in our experience, justify usage:(1) Primary nerve stretching; (2) flexion-relaxation; (3) nerve transposition;(4) stretching with secondary suture (two-stage operation); (5) autogenous cable grafts; (6) viable neuroplastic transplants; (7) bone shortening.

PRIMARY STRETCHING

Primary stretching will overcome many of the lesser defects. It consists of freeing the nerve trunk beyond the area of scar tissue fixation and drawing its ends together by gentle traction. Extensive mobilization of a nerve trunk permits a proportionately greater stretching. It is particularly necessary to freely mobilize the nerve beyond the region of extraneural adhesions, but in doing this, great care should be exercised in avoiding injury to branches.

FLEXION-RELAXATION

The majority of nerve defects may be corrected by taking advantage of the relaxing effect upon a nerve trunk, by flexion of governing joints. The approximation of the nerve ends is made possible, after the nerve is prepared for suture, by flexing the governing joint, in which position it is maintained, after the nerve is approximated, by suitable splints until the wound is healed; after which the mobilizing splints are occasionally changed to permit a gradual extension of the flexed extremity. In our study of end results of nerves under tension or those stretched by the foregoing method, we have not found that this stretching seriously interferes with regeneration; in fact, some of our best end results have been observed in cases where approximation was attainable only under considerable tension. In our early experience, when flexion-relaxation was used in overcoming large defects, we maintained the limb in the position of primary flexion for a period of six weeks; later, however, this period of fixed flexion was reduced to two weeks, after which the gradual extension of the extremity was permitted at the rate of approximately ten degrees every second day.

NERVE TRANSPOSITION

The course of certain nerves in the arm may be made more direct and thereby shortened by transferring them from a dorsal to a more ventral plane. This procedure is commonly used to great advantage in overcoming the more extensive defects.


955

The ulnar nerve in its normal position behind the internal condyle is not relaxed by flexion of the elbow until it has been transposed to a position anterior to the internal humeral condyle. This procedure not only somewhat shortens its course but also renders it susceptible to flexion-relaxation. The musculospiral nerve is occasionally transposed, being directed from its posterior humeral course to one anterior to the humerus. Large defects in the median often require for their correction a transposition of the nerve from its deep position below the superficial head of the pronator teres to one ventral to this muscle.a

In the transposition of nerves it is essential that the operator be familiar with the various branches given off from that portion of the nerve trunk requiring mobilization; these branches must be freed some distance up the nerve trunk by an intraneural dissection to permit mobilization without their avulsion. It is also essential to place topographical markings on both proximal and distal segments of the nerve by means of well placed identification sutures, before the position of these segments has been disturbed and anatomic relationships lost, in mobilization.

STRETCHING, WITH SECONDARY SUTURE (TWO-STAGE OPERATION)

After a divided nerve has been exposed and found to present a defect, the surgeon should determine means of correcting this defect before the neuroma anti scar tissue have been excised from the ends of the nerve. If it be found that the defect will not lend itself to correction by one of the foregoing methods, the operator must then resort to the two-stage operation. This consists inextensive mobilization of the nerve trunk after it has been marked with topographical identification sutures, and all possible relaxation has been attained by transposition and flexion-relaxation. The unsectioned nerve ends are drawn together by one or two strong chromic catgut traction sutures using sufficient tension to permit, if possible, an overlapping of the nerve ends to a point approximating that which will correct the defect, after the sear tissue has been resected; or at least, to obtain as much overapproximation as possible. The wound is now closed, and after one week the extension of the extremity begun, the object being a gradual stretching of the nerve trunk. When this has been accomplished by complete extension of the extremity and the wound is entirely healed so that it is safe for a secondary aseptic invasion, the nerve is again exposed; the neuroma and scar tissue are resected and the freshened ends approximated by again utilizing the assistance of flexion-relaxation. The extremity is splinted in this position for a period of two weeks, after which extension is gradually encouraged. Many of the defects encountered in extensive lesions of nerves have been corrected by a two-stage operation, and only when the nerve defect is of such magnitude as not to be corrected by the above method are we justified in resorting to grafting.

NERVE GRAFTS

From an experimental standpoint, nerve grafting has attracted considerable interest and much has been expected from this procedure; from a clinical standpoint

a The technique of transposing the above nerves is considered in detail under the surgery of special nerves.


956

however, the end results have been very disappointing. A review of end results in so far as functional restoration is concerned, in 15 cases in which the operations were done by different operators, shows that functional restoration was nil, with the exception of one instance in which there was some return of sensation in a low median graft. (This restoration consisted, after four years, in simply the restoration of a nondiscriminative sensation in the index and middle fingers. All sensation, including touch, pin prick, heat, and cold, were interpreted as tingling-differentiation not being possible.) In the writer's experience, a number of homogenous refrigerator grafts were tried out; in each instance the condition of the wound was favorable and remained so throughout the healing process, though in no case was there any clinical evidence of regeneration. These grafts when subsequently removed showed a total absence, histologically, of nerve fiber regeneration, and the proximal end of the nerve trunk exhibited a secondary neuroma. In most of the above instances it was later found possible to obtain satisfactory end-to-end approximation by the above-described two-stage operation. With increasing experience in dealing with large defects, we found that in most instances it was possible to repair the defect by a two-stage operation. thereby almost entirely eliminating the necessity for grafts. Occasionally, however, defects are found of such magnitude that the only possibility of nerve repair lies in grafting, but as a rule such cases, by virtue of their severity, the extent of tissue destruction and scar tissue formation, promise but little in the way of satisfactory functional restoration. It must be remembered that the experimental work of nerve grafting which has been attended with such satisfactory results has been done in relatively small defects under extremely favorable conditions from a surgical standpoint. Such satisfactory conditions are rarely encountered clinically because of the severity of the original lesion which makes nerve grafting necessary.

AUTOGENOUS CABLE GRAFTS

Autogenous cable grafting consists in exposing the end is of both proximal and distal segments of the divided nerve and preparing them for suture. The length of the defect between the nerve ends is measured and the grafts to be inserted are selected from convenient sensory nerves. In the arm the following nerves may be utilized: Radial (sensory portion of the musculospiral), sensory portion of the musculocutaneous, internal cutaneous. In the leg: Sensory portion of the musculocutaneous, tibial, and peroneal recurrent, the sural branches of the peroneal nerve. One or several segments from one or more of the sensory nerves, of sufficient length to fill the defect, are removed and transplanted to fill the gap between the ends of the damaged nerve. These sensory branches are usually so small that several are required in making a cable which will approximate in size the nerve trunk. After a cable graft of sufficient size a1l length is obtained, it is carefully anastomosed to the ends of the nerve trunk and in this way made to bridge the defect. It is essential to obtain accurate end-to-end approximation between both ends of the nerve trunk and graft. If a defect in a small nerve is being bridged, it may be possible to obtain sensory graft of approximate size, thereby eliminating the necessity of using


957

several grafts as it, cable. It is particularly important that a favorable bed be found to protect the grafted nerve. The writer, in the removal of a number of unsuccessful grafts, has found them converted into strands of fibrous tissue; in at few instances some regenerating neuraxons Were found to have penetrated the proximal end of the implant, but in none had they succeeded In completely traversing, the graft.

VIABLE NEUROPLASTIC TRANSPLANTS

In irreparable defects of combined nerve lesions, such as the median and ulnar or the tibial and peroneal portions of the sciatic trunk, the writer, by sacrificing the nerve of lesser importance, devised a method of filling the defect in the more important nerve with at viable transplant. In this procedure for example for, in a combined median and ulnar irreparable defect, the median considered to be the nerve of greater importance, the end of its proximal segment is anastomosed by end-to-end suture to the proximal end of the ulnar nerve--the nerve of lesser importance--permitting the median fibers

FIG. 150.– Bundle or “cable” graft, using an autosensory nerve for repair of the defect. (Ney, Annals of Surgery, 1921)

to regenerate around the loop) so formed by the anastomosis and up the ulnar trunk. The ulnar nerve is divided the required distance (length of median defect) above its anastomosis to permit the degeneration of its fibers in that portion of the trunk which subsequently is to be used as a transplant, thereby preparing it to receive the regenerating fibers of the mendian nerve. The trunk of the sacrificed ulnar nerve remains undisturbed in its original position, to conserve its nutrition during the migration of median fibers through its entire length. Before the transplant is utilized, it is allowed one month for each inch of its length plus one month's grace, to assure regeneration through its entirety. The upper end of the adopted trunk may be placed subcutaneously if desired, where it ran he readily palpated and percussed to determine the growth of nerve fibers to its extremity--the presence of regenerated fibers may be readily demonstrated by Tinel's sign, thus assuring the surgeon of the viability of the transplant before it is turned down at the second stage of the operation.

The advantage of the viable neuroplastic transplant is that its nutrition is maintained during neuraxon regeneration, and at the time of use it has, to


958

all intents and purposes, become a viable part of the nerve trunk, containing intact and viable neuraxons. It then may be united to the distal segment of the nerve with all the advantages of a direct approximation. The writer has restricted its use to the repair of inoperable combined lesions in which he felt justified in sacrificing a permanently disabled neighboring nerve. The possibilities of the viable neuroplastic transplant will probably be extended with experience as it seems at the present time to be our only practical method of filling nerve defects. The results following the use of the viable transplant are comparable to those following direct approximation.

BONE SHORTENING
          
Bone shortening is occasionally justified for the correction of large nerve defects, though the arm should not be shortened until full advantage is taken of other means to correct the defect. Shortening of the upper extremity does not, of itself, produce serious disability, but obviously such a procedure would not be practical in the lower extremity. Before recourse is taken to bone shortening the two-stage operation of nerve stretching should be tried; if this fails by a few centimeters to correct the defect, humeral shortening to this extent may be practiced to assist the other procedure. When a section is removed from the humerus for the purpose of shortening the extremity to permit coaptation of divided nerves, the resection is usually done in its lower third, care being taken to prevent injury to the musculospiral nerve if the exposure is carried into the middle third. The bone may be divided with a Gigli saw, after which it is dislocated through the incision and a section sufficiently long to correct the defect removed. The ends of the bone are approximated before suturing the nerve, by bone plates or a medullary graft, or by their combination. It is particularly important in approximating the nerve and closing the incisions to prevent unnecessary movement which might disturb the position of the bone or throw undue strain upon the bone plate, or graft. The arm should be carefully immobilized with plaster to a body cast which has been applied before the operation, and so maintained for a period of six weeks or until firm bony union has taken place. A window should he made in the cast which will permit inspection of the wound and change of dressings when necessary. The skin is sutured with catgut to prevent the disturbance of suture removal.
 
ANKYLOSED JOINTS WITH NERVE DEFECTS

Occasionally an ankylosis of the elbow or knee joint will prevent the utilization of flexion-relaxation in overcoming continuity defects in nerve. The advisability of resorting to arthroplasty for the correction of ankylosis in these joints will depend greatly upon the disability resulting from the combined nerve and joint lesion. Much has been said for and against arthroplasty as a practical procedure; many surgeons of experience are in favor of allowing these joints to remain ankylosed, providing their position is useful. The writer is inclined to believe that in many instances very satisfactory results will follow a well-planned and carefully executed arthroplasty; this is particularly true of the elbow, when the ligaments have not been extensively destroyed.


959

A flail-joint, however, is more disabling than one ankylosed in the position of maximum usefulness. If arthroplasty has made it possible to repair the nerve and restore function in the paralyzed muscles, even though it result in a flail-joint, it has served at useful purpose, for the joint may again be ankylosed in a convenient position.

IRREPARABLE NERVE DEFECTS

The neurologic surgeon, in about 5 percent of cases, will meet continuity defects in injured nerves of a magnitude so great as to preclude the hope of successful nerve repair. In such instances he must resort to various supplementary procedures which will promise some degree of ultimate functional usefulness in the injured extremity. From a motor standpoint, tendon transplantation frequently offers a fairly satisfactory solution of the problem. In other instances, the immobilization of joints must be resorted to; this procedure is of value in deltoid paralysis, in upper plexus lesions with loss of elbow flexion, and in stabilizing the foot in irreparable lesions of the sciatic nerve. These various supplementary surgical procedures, though ordinarily falling within the realm of orthopedic surgery, should be done, or at least directed, by the neurologic surgeon, who is familiar with the neurologic aspects of the case. This is particularly important in tendon transplantations, which require a thorough knowledge of the existing motor status of each individual muscle utilized in the transplantation, if one is to obtain the most satisfactory functional end results. b

TORSION OF THE NERVE TRUNK DURING SUTURE

A most serious eventuality in the repair of nerves is torsion of the nerve trunk during suture. Many surgeons have been content to obtain a satisfactory end-to-end approximation of the divided nerve, but a physiologic appreciation of the functional individuality of the various bundles within the nerve trunk should tend to discourage such a procedure as being manifestly inadequate, for the final results are greatly inferior to those of a technique where due regard is paid to physiologic approximation. There are very few instances in which it is not possible to obtain at least some idea of the topographical anatomy of the nerve trunk before suture, and the surgeon who has not the ability, time, or patience to use the physiologic technique in all its phases is not justified in attempting the repair of peripheral nerves.

There is no evidence at our command, and much to the contrary, to verify the ability of motor and sensory fibers to find their respective physiologic channels during regeneration through the distal segment of a divided nerve trunk. If motor fibers are placed in contact with the sensory channels of the lower segment, these regenerating motor fibers will follow the sensory channels, into which they have been directed, to a sensory termination, and be physiologically lost; it is not possible for motor fibers to subserve the function of sensation, nor are sensory fibers capable of carrying efferent motor impulses

b Supplementary procedures applicable to individual nerves are considered in detail under the surgery of special nerves.


960

to muscles, if, during suture, they should be directed down motor channels. This physiologic misdirection of nerve fibers is probably responsible for much of the defective end results attending the surgery of peripheral nerves; and while our present technique is not sufficiently refined to differentiate accurately the position within the nerve trunk of these respective motor and sensory bundles, it is possible for the surgeon, with a fair degree of accuracy, to maintain the physiologic alignment of the nerve during suture by the use of the following methods: (1) Identification sutures; (2) forceps identification; (3)anatomic or branch identification; (4) electro-anatomic method.

IDENTIFICATION SUTURES

Identification sutures are placed at a corresponding point in both proximal and distal segments of the nerve sheath before the normal position of the nerve has been disturbed by dissection; the purpose of these sutures is to identify a definite circumferential point upon the nerve trunk which will permit, during approximation, a restoration of original anatomic alignment without torsion. It is well that these sutures be placed some distance above and below the lesion where normal anatomic landmarks prevail and where the nerve has not been subjected to torsion through the contraction of scar tissue. This procedure is undoubtedly the simplest and perhaps the most efficacious method of preventing torsion. Two sutures may be placed, one above the other, to assist in more accurate alignment. When the scar tissue is resected from the nerve ends preparatory to suture, these identification markings should be lined upmost carefully and the approximation sutures placed in a manner which will maintain this alignment.

FORCEPS IDENTIFICATION

From the standpoint of technique, forceps identification offers certain advantages and may be used in conjunction with identification sutures, or independently. As the nerve trunk is exposed and before it has been completely freed from its bed, the nerve sheath is grasped on each side with very fine mosquito hemostats. These forceps are so placed on both proximal and distal segments that they locate exactly the center of the medial and lateral aspect of the nerve, for the purpose of permitting alignment; they also serve efficiently to stabilize the nerve while it is being sectioned. Unless care is used, the operator may rotate the nerve trunk and produce torsion by alignment of lateral forceps with medially placed forceps, if the forceps used on each side of the nerve trunk are of the same pattern. (To prevent this, the writer identifies both laterally placed forceps by tying a suture to the thumb piece, which will serve for lateral identification in the alignment.)

ANATOMIC OR BRANCH IDENTIFICATION

This may be utilized in many locations to effect satisfactory alignment: it is particularly valuable in secondary sutures, when the surgeon has reason to believe that due regard had not been paid to physiologic alignment during the primary suture; it is also of value in checking up the accuracy of identification sutures.


961

A careful study of a series of dissections has convinced the writer that both motor and sensory bundles occupy within the nerve trunk a very definite and relatively constant position. While no dogmatic rule may be given as to the position of exit of a nerve branch from the parent trunk, it may be generally stated that if the nerve sheath be opened for a short distance a given branch will invariably be found to originate from a definite position within the nerve trunk. If a motor branch be followed a short distance up the nerve trunk, it will be found to occupy a definite location which indicates the position of the motor bundles in the nerve trunk. The identification of motor branches springing, from both segments of the divided nerve, when these branches have received proper intraneural localization, will assist materially in accurately approximating motor segments during suture.

To reiterate, if a motor branch to a given muscle is identified and followed a short distance up the nerve trunk until it has assumed a fixed intraneural position, this position may be definitely regarded as the location of the motor portion of the nerve. It may be termed the motor sector; conversely, the other side of the nerve would be the sensory sector. The utilization of this information serves topographical localization and physiologic identification

FIG. 151.- Diagram showing the necessity of determining the intraneural location of a given branch. A motor branch may leave the sensory side of a nerve trunk, and intraneural dissection is required to establish establish its deep origin which represents the motor sector of the trunk; the superficial origin of nerve branches is often misleading the study of the topographical and varies considerably

when other methods, such as identification when other methods, such as identification sutures, are not to be relied upon. With the study of the topographical anatomy of individual nerves, this method of branch identification becomes of increasing value. Obviously, the surgeon who expects to use this method must familiarize himself with the intraneural anatomy of the nerve upon which he contemplates surgical invasions.c

THE ELECTRO-ANATOMIC METHOD

This method presents certain refinements in topographical localization, unattainable by purely anatomic methods. In the lower segment, the motorportion of the nerve trunk is localized by the anatomic or branch identification method as described above; in the upper segment, the sensory side of the nerve trunk is localized by weak faradic stimulation of its various bundles, after the nerve sheath has been opened 2 cm. above the nerve end. The technique of

c Under the surgery of special nerves such detailed information as is now available regarding the method of anatomic identification is considered.


962

this procedure consists in exposing the nerve by infiltration anesthesia, without injection of the nerve trunk proper. (See technique of local anesthesia, p. 952).Two centimeters above the scar invaded end, the nerve sheath is opened longitudinally and its pearly white bundles exposed; two or three of these bundles are isolated from intraneural connective tissue by gently inserting and spreading the blades of fine-pointed iris scissors parallel with the course of the bundles. A single bundle is then lifted upon a fine glass hook and subjected to very gentle faradic stimulation with a fine needle-pointed bipolar electrode. If a sensory bundle be stimulated, the patient will complain of a tingling sensation in the

FIG. 152.- Electro-anatomic method of topographical identification. A, Motor branches in both proximal and distal segments serve to identify the motor sector of the nerve. B, Motor sector of distal segment is identified by the position of motor branches. The proximal segment, having no branches to permit anatomic identification of intraneural topography, the nerve sheath is opened and the nerve bundles Identified by faradic irritation, which is most effective in determining sensory bundles; occasionally the stimulation of bundles at the end of a resected nerve may give some assistance in topographical identification, though it is of less value and the stimulation is difficult to control

normal cutaneous area supplied by that bundle. If the current be too strong. the patient will complain of pain. If a motor bundle be stimulated, the patient will experience a sensation of muscular effort due to stimulation of myosensory fibers. For example, in the musculospiral nerve, if a motor bundle be subjected to mild faradic irritation in the arm, and the patient experiences a sensation of " pulling up " of the wrist or straightening of the fingers, the surgeon will know that he is stimulating motor fibers to the carpal extensors or to the extensor communis digitorum, and he has thus localized the motor side of the nerve trunk above the lesion. If, however, the patient experiences a


963

tingling sensation which is localized on the dorsum of the thumb or hand, the surgeon is then stimulating the sensory area of the nerve trunk.

The electro-anatomic method of funicular identification consists, therefore, in locating the motor portion of the nerve trunk in the distal segment, by identifying the intraneural position of motor branches, while in the upper segment of the nerve trunk the sensory portion is localized by weak faradic stimulation; in approximation, the surgeon is thereby able to effect a certain degree of physiologic alignment. This method is of value principally in areas where the upper segment of the divided nerve is not giving off motor branches, such as the median nerve in the upper arm. After the identification of the physiologic constituents of the proximal segment, the nerve trunk may be anesthetized with novocaine, and the neuroma and scar tissue resected, preparatory for approximation. In the electrical identification of bundles, it is essential to begin with the weakest possible current, the needle points of the electrode being separated not more than 2 mm.

Electric funicular identification is an extremely delicate procedure, and it requires the greatest care and gentleness, particularly to insure a current sufficiently weak to avoid pain on stimulating sensory bundles; a current which can be felt on the tip of the tongue is often too strong for sensory bundle stimulation.

BUNDLE IDENTIFICATION

After resection of the scar tissue from the ends of the nerve, the individual nerve bundles stand out prominently, and some surgeons have attempted to effect funicular identification by bundle matching. In Ney's experience this has not proved feasible, because the bundles of the lower segment, which are but the empty channels from which the neuraxons have disappeared by degeneration, in no way correspond either in number or in size with the intact bundles of the proximal segment.

APPROXIMATION TECHNIQUE

After both segments of the divided nerve have been exposed and precautions taken to prevent torsion during approximation, and the surgeon has assured himself of the possibility of correcting the defect, the nerve ends maybe prepared for suture. The resection of nerve ends, however, must not be done if there is any possibility of having to resort to the two-stage operation to correct a defect.

In preparing a nerve for suture, the neuroma of the proximal end is resected in serial sections about 1 mm. thick until normal-appearing nerve bundles present themselves throughout; this resection should be continued until the nerve bundles are free from scar tissue. The resection of the nerve end should be made with a very sharp instrument, and for this purpose a new safety razor blade held in strong forceps serves admirably. If it is desired to preserve the sectioned ends for further study, the order of their arrangement may be maintained by stringing them upon a piece of suture material, after which they may be placed in alcohol or formalin. The end of the distal segment does not present a typical neuroma as does the proximal end and is often


964

FIG. 153.- Application of forceps to an immobilized nerve during section. (Ney. Annals of Surgery, 1921)

FIG. 154.- Sectioning of nerve ends for removal of neuroma and scar tissue. Method of preserving sections in order of removal. (Ney. Annals of Surgery, 1921)


965

found to taper off into the surrounding scar tissue, in which it is lost. It should be sectioned carefully beyond the point of scar invasion to a point where the bundles stand out freely, using the same technique as that used in sectioning the proximal end.

After the nerve ends have been prepared, the first fine silk suture is placed, passing just through the nerve sheath, not deep enough to encroach upon or to constrict the bundles. The first suture is of great importance, in that it

FIG. 155.- A, Technique of end-to-end suture, showing the placing of identification sutures before a nerve is removed from sear tissue. B, Exposed nerve before resection of neuroma and scar from its ends. C, Nerve resected, identification forceps applied, and three quadrant sutures placed. (Ney. Annals of Surgery, 1921)

should remain the guide to alignment in the placing of subsequent sutures, and in the placing of this suture the alignment of identification sutures should be assured. The second and third sutures are now placed in the center of the medial or lateral quadrant of the nerve trunk. The ends of these sutures are rasped by forceps and not tied until all quadrant and intermediate sutures are accurately placed. In placing the fourth or posterior quadrant suture, the nerve trunk is rotated by passing the forceps holding the lateral suture


966

under the nerve trunk, rotating it to the medial side, while the medial qualdrant suture is pulled over to the lateral side. This rotates the nerve trunk so that its posterior quadrant is exposed, when its suture may be accurately placed. Between each quadrant suture one or more intermediate sutures maybe required, depending upon the size of the nerve trunk. Adjacent quadrant sutures are held taut, during the placing of each intermediate suture, to assist inaccurate spacing. A sufficient number of sutures should always be used to assure perfect approximation of the nerve sheath, thereby preventing the neuraxons from wandering outside the nerve trunk at the suture line, where they would be

FIG. 156.- A, Rotation of the nerve for the purpose of placing the posterior quadrant suture. B, Intermediate sutures placed and all sheath sutures held in position to prevent rotation in placing a tension suture. (Ney. Annals of Surgery, 1921)


physiologically lost and tend to form a sensitive neuroma. After all suture shave been placed the nerve should be relaxed by flexing the governing joint and the nerve ends gradually approximated by taking in the slack of the sutures. If considerable tension upon the sutures is necessary to produce approximation, a tension suture of plain catgut may be passed completely through the nerve trunk 1 cm. beyond its ends to diminish the tension on sheath sutures. It is essential in placing the tension suture that the forceps holding the sheath sutures be equally arranged on each side, corresponding to the sheath sutures of that side, and these sutures held taut to prevent distortion of alignment during its insertion. Care must be exercised to prevent tearing of the sheath sutures


967

FIG. 157.- A, Approximation by tension suture. B, Order in which sheath sutures are tied after nerve is approximated by the tension suture. C, End-to-end suture completed. (Ney. Annals of Surgery, 1921)

FIG. 158.-The V section of a small distal segment used for the same purpose as the diagonal section in Figure 159. (Ney. Annals of Surgery, 1921)


968

when force is required to draw the tension suture through the nerve trunk. With all sheath sutures held taut in their consecutive arrangement, the tension suture is tied as the nerve ends are approximated. The sheath sutures are individually picked up and tied, beginning with those placed proximal to the tension suture; as the sheath sutures are being tied, every second suture is left uncut to facilitate rotation of the nerve trunk for subsequent inspection of the suture line. If sheath approximation is found to be imperfect at any point additional sutures may be placed for its correction. Frequently, after sectioning the nerve ends, there is some bleeding, which may usually be controlled by

FIG. 159.- Diagonal section of distal segment where it is smaller than proximal segment, for the purpose of securing accurate sheath approximation. (Ney. Annals of Surgery, 1921)

wet cotton compresses. As the nerve is being approximated, care should be used to wash away any blood clot which might have formed over the end of the sectioned nerve. This clot should be washed away, not wiped, as the traumatism of wiping will frequently dislodge the clot, occluding the end of the vessel, and the bleeding will be renewed.

THE PREPARATION OF THE NERVE BED

When the original nerve bed contains an excess of sear tissue, a satisfactory new bed may be provided by placing a few deep sutures through the adjacent muscles, which may be drawn together under the nerve, thereby burying the old bed and providing one more favorable to the needs of regeneration. It must be remembered in the preparation of a nerve bed, however, that scar tissue may be limited only by the avoidance of tissue trauma and by


969

adequate hemostasis; therefore, in interposing adjacent muscular tissue for the formation of the new bed, it is desirable that untraumatized muscle be used when available. When scar tissue is excessive and normal muscular tissue is not available for the preparation of the nerve bed, the nerve may be brought to a superficial position. Occasionally in extensively scarred areas, neither muscle nor superficial fat will be found available for the protection of the nerve: it then becomes necessary to transpose the nerve by directing its course through adjacent intermuscular planes. When, however, the length of the nerve does not permit transposition, the surgeon must resort to plastic procedures in which normal skin and superficial fascia from adjacent regions are shifted to cover the scarred area. When adjacent fat is not available for plastic procedures of this type, skin and fat from the abdominal wall may be transplanted by the pedunculated flap method.

After a satisfactory bed has been provided, all blood clots should be gently washed away by a running stream of warm saline solution. The common habit of wiping incisions with gauze, instead of sponging, is to be discouraged as it causes unnecessary trauma to the wound and tends to increase oozing. It is usually advisable to drain all incisions, when there has been extensive dissection, by one or two well-placed pieces of rubber dam, which should be adequate to care for any subsequent oozing. This drainage should be removed at the end of 48 hours.

PARTIAL NERVE LESIONS

Occasionally a missile will sever only a part of a nerve trunk, causing a partial lesion; or a portion of the nerve may suffer accidental or intentional surgical injury during debridement, or in the removal of a nerve tumor. In several instances, the writer has found fragments of bone driven into the nerve trunk: in one case the musculospiral nerve had been penetrated by a hypodermic needle during typhoid inoculation, resulting in a permanent injury to the fibers supplying the extensor indicis. The majority of gunshot wounds involving the sciatic trunk are partial lesions, usually involving its external or peroneal portion. The frequency with which partial lesions are met adds greatly to their importance and demands special consideration, for in the surgical treatment of such lesions it is essential that the operator. in his efforts at repair, does not inflict additional surgical trauma to such intact nerve fibers as might have escaped the original injury. A careful preoperative determination of individual muscle function is extremely important, inasmuch as it should indicate the exact extent of nerve injury and incidentally determine the functional importance of intact nerve fibers.

TECHNIQUE OF REPAIR IN PARTIAL LESIONS

If the nerve trunk, on exposure, exhibits a central or lateral neuroma, partial excision and approximation are required. If the preoperative examination indicates a high degree of functional integrity, the surgeon may rest assured that the “neuroma" is composed mostly of scar tissue, and its removal should


970

FIG. 160.- A, Partial lesion of a nerve trunk. B, Isolation of the interrupted portion from the physiologic normal portion. Quadrant sutures placed for approximation. C, Approximation in partial suture, such as a partial division of the sciatic nerve, showing relaxed undivided portion of the nerve. (Ney. Annals of Surgery, 1921)

FIG. 161.- A, Partial lesion of a nerve trunk, where gross anatomic isolation of functionally intact portion of a nerve can not be made as in Figure 160. B, Opening of the nerve sheath, showing in- volvement of bundle. C, V-shaped incision of sheath by which approximation of the bundle is made possible. (Ney. Annals of Surgery, 1921)


971

be attempted only after the nerve trunk has been longitudinally opened for some distance, and the normal bundles carefully separated from those incorporated in scar. Such bundles as appear to pass directly into the scar tissue should be carefully isolated above and below the lesion and subjected to weak faradic stimulation with a needle-point bipolar electrode. When faradic stimulation of a nerve bundle, distal to the lesion, is accompanied by muscular contraction (not due to diffusion of current), the surgeon may feel assured that the bundle is anatomically intact and should be dissected from the scar tissue with great care, avoiding traction, as nerve bundles are very easily torn. Bundles which give no response during electrical stimulation may not be anatomically divided, but only subjected to a degree of compression sufficient to destroy their conductivity; they should be carefully followed through the scar tissue, if possible, to determine their anatomic continuity. If a bundle shows scar tissue infiltration, the infiltrated area should be resected and the bundle ends approximated by a single suture of arterial silk passed directly through the bundle. When bundles are not readily approximated because of a defect, a V-shaped portion of the nerve sheath may be resected, corresponding in width to the length of the defect; linen sutures may be used to approximate this portion of the sheath, and with its approximation the ends of the individual bundles will be brought together, overcoming their defect; bundle sutures may then be tied, care

FIG. 162.- A, Approximation of the divided bundle accomplished by relieving tension in the approximation of the resected sheath, being used to obtain accurate B, The thickened sheath is not entirely closed for fear of strangulation; the defect is covered by a fat transplant. (Ney. Annal of Surgery, 1921)

being used to obtain accurate end-to-end approximation without strangulation. After approximation of individual bundles has been satisfactorily effected, the nerve sheath should be only partially closed; the unclosed portion may be covered with some surrounding untraumatized tissue such as intact muscle sheath, or if this is not available, by a pedunculated or free fat transplant, extreme care being given to hemostasis. In partial lesion of the sciatic trunk in which preoperative examination reveals a total loss of peroneal or tibial function, the nerve should be exposed sufficiently low to obtain the natural line of cleavage between its component parts, which upon their separation may be sutured in their entirety, as individual nerves. When a nerve trunk is found to present evidence of almost complete anatomic division, retaining only a small degree of functional integrity, one is justified in resecting the entire trunk and effecting a complete suture, but all partial lesions retaining important functional integrity should be subjected to individual bundle examination and repair.

COMPRESSION AND STRANGULATION LESIONS

Nerve trunks, while retaining their anatomic continuity, may be compressed to such a degree as to interrupt their physiologic continuity. This compression,


972

if excessive, may result in complete strangulation with neuraxon degeneration. Strangulative lesions with degeneration (do not present the typical neuroma, as is common to anatomic divisions, and the nerve can usually be traced through scar tissue as a more or less intact anatomic structure. Nerves may be compressed by extraneural scar tissue, bone callous, or even external pressure, such as ill-fitting splints, crutches, or the edge of operating tables. Surgical intervention is not indicated in the milder degrees of compression until sufficient time has elapsed for evidence of spontaneous regeneration. When, however, after three months there exists no sign of regeneration and the exact location of the lesion is known, the surgeon is justified in exploration. While extraneural lesions are probably the most common cause of nerve compression, two other types often exist which far more frequently demand surgical intervention: Compression or strangulation from nerve sheath involvement, and intraneural fibrosis.

NEUROLYSIS

Many surgeons have been content with an external neurolysis when exposure revealed the nerve trunk to be anatomically intact; such a procedure has frequently given very disappointing end results, obviously, because of inadequate bundle decompression. When a

FIG. 163.- A, Physiologic interruption of a nerve; nerves with this appearance are occasionally considered as having an "internal neuroma." B, Showing the enlargement to be due to a greatly posed and surrounding scar thickened nerve sheath, producing compression or strangulation. C, Perifunicular adhesions following the prolonged use of a tourniquet. The method of separation of adhesions, as in A and B. The sheath is left open to prevent future strangulation, and the exposed funiculi are protected by adjacent fatty tissues when present, or fat transplantation. (Ney. Annals of Surgery, 1921)

compression lesion has been exposed and surrounding scar tissue removed (which constitutes an external neurolysis), the nerve trunk should be carefully palpated; if it be found soft, without evidence of a thickened sheath or intraneural fibrosis, the external neurolysis should be deemed sufficient.

When the nerve has an indurated appearance and presents undue hardness on palpation, decompression, or internal neurolysis is indicated. The nerve sheath is longitudinally incised, exposing the indurated portion, and gently held open with fine traction sutures or mosquito forceps. This intraneural exposure will usually reveal the nature of the compression, whether it be due to a thickened indurated nerve sheath or to an intraneural fibrosis. When the compression is confined to a thickened and indurated nerve sheath, the exposed bundles, though showing signs of compression, will be found intact, readily


973

isolated, and free from scar tissue. Such a condition is usually relieved by the simple decompressive procedure of opening the nerve sheath. When the sheath exhibits marked thickening the thickened portion may be resected, no attempt being made at closure. Occasionally the intraneural exposure reveals" matting" of the bundles by adhesions; intraneural neurolysis is then indicated. The bundles are gently dissected free from the scar tissue by inserting the blades of fine-pointed iris scissors which are spread, gently separating the bundles from the perifunicular adhesions. If, however, the bundles are found imbedded in a dense mass of scar in which they lose their anatomic identity, resection of this area is indicated, and the nerve united according to the usual technique for end-to-end suture.

THE ELECTRICAL EXAMINATION OF NERVES AT OPERATION

Important information may be obtained by the routine use of faradic stimulation of nerves when exposed at operation. The battery required is the ordinary type of single-cell faradic medical battery, in which the strength of current may be regulated by moving the core. A fine-pointed bipolar electrode is connected to the battery with plain copper or silver wires, which may be sterilized by boiling. Small glass beads may be threaded on each wire to prevent short circuiting. Ordinarily the bare ends of the wire may be used as a bipolar electrode, but for the more delicate procedure of funicular identification needle points should be used, these points having a separation at their end of approximately 2 mm. In beginning the test the weakest possible current should be used, the strength of which may be increased as desired, by the attendant manipulating the coil. The operator should assure himself that the connections are intact and the battery is working properly, by stimulating neighboring exposed muscular tissue whose nerve supply is normal.

When a compressed nerve is exposed at operation. it should be stimulated both above and below the lesion. If weak faradic stimulation above the lesion gives a response in the paralyzed muscles, it is doubtful whether the lesion exposed is responsible for any serious compression. Where such a condition is found to exist before the nerve has been decompressed, the diagnosis of an organic lesion is questionable, providing all the muscles respond to stimulation. It is possible, however, that the surgeon is dealing with a partial lesion and is stimulating intact fibers. A careful preoperative examination, however, should have demonstrated the probability of such a condition. Occasionally, an anatomically intact nerve may be exposed in a compression lesion, where stimulation above the lesion fails to elicit a response, while stimulation below the lesion gives definite muscular contractions; such findings are found only in the mildest grades of nerve compression, and a careful neurolysis will usually result in the passage of electrically excited impulses beyond the lesion, providing the nerve has not been traumatized nor stretched during the neurolysis. Trauma during neurolysis will frequently cause the disappearance of the electrical conductivity previously found.


974

ELECTRICAL IDENTIFICATION OF PHYSIOLOGIC COMPONENTS OF A DIVIDED NERVE

When a nerve is exposed under local infiltration anesthesia and before the nerve trunk has been anesthetized, it is possible by opening the nerve sheath of the upper segment to effect physiologic identification of the exposed bundles by weak faradic stimulation. The stimulation of a sensory bundle elicits a tingling sensation localized in the normal cutaneous sensory area of that nerve; if the stimulation is too intense, considerable pain may be experienced. Stimulation of a motor bundle is not associated with pain unless the current is diffused to sensory bundles. The stimulation of myosensory fibers, usually found in a motor bundle below a nerve plexus, gives a sensation of muscular action. (See technique of electro-anatomic funicular identification p. 963.) In partial lesions of nerves the physiologic intact portion of the nerve trunk may be determined by faradic stimulation.

The success of electrical identification of nerve bundles depends primarily upon an extremely careful and gentle technique. If the current used is a bit too strong, the patient is caused pain and the tests are difficult to control. The current also has a tendency to diffuse throughout the nerve trunk, stimulating all fibers and rendering the test valueless. When carefully and accurately applied it is an extremely valuable adjunct in identifying the physiologic components of a nerve trunk. In long-existing paralysis the stimulation of motor bundles will occasionally give no myosensory reaction; the same istrue in the stimulation of motor bundles above a nerve plexus. The most reliable results obtained are from stimulating a completely formed branch, before it has left the parent trunk.

SECONDARY OPERATIONS FOR DEFECTIVE REGENERATION

Inasmuch as defective regeneration may take place in both nerve and muscle, it is essential in the absence of regeneration to decide definitely the nature of the regenerative defect.

MUSCLE REGENERATION

Primarily, degeneration in paralyzed muscles following peripheral nerve injuries is due to denervation; the longer a muscle remains denervated the more intense will the degenerative changes become. In muscles of small size these degenerative changes seem to become extreme, and after a period of 12 or 18 months the degeneration has progressed to a point where recovery is extremely doubtful. Small muscles, like the intrinsic muscles of the hand and foot, being innervated at a great distance, seldom recover after nerve injury if deprived of their nerve supply longer than one year. Therefore, secondary operations are, as a rule, not indicated for regenerative failure in muscles of small size with distal innervation. In muscles of larger bulk, whose nerve supply is usually derived at a higher level, the chances for regeneration are far more favorable; in regeneration of the injured nerve they receive nerve fibers much earlier than the smaller, more distally located muscles, and


975

their larger bulk insures their receiving a greater number of regenerated motor fibers. In the upper extremity regeneration occurs in about the same degree in muscles of comparative size, whether they be innervated by the musculospiral, median, or ulnar nerves, providing the lesion of these respective nerves occurs at about the same level and the period of their denervation is the same. In brachial plexus lesions, which are high lesions, the more distal muscles in the forearm, as a rule, show more extreme degeneration and less tendency to recover than those muscles deriving a higher innervation. In high sciatic trunk lesions recovery of the gastrocnemius is much more likely, because of its large bulk and comparatively high innervation, than is to be expected in the flexors or extensors of the toes. We may conclude, therefore, from a standpoint of muscle regeneration, that the longer a muscle remains paralyzed the greater will be its degeneration and the less likely will be motor recovery.

Reoperation on nerve trunks is indicated solely because of absent or defective neuraxon regeneration, and not because of defective muscle regeneration. While re-innervation is absolutely essential to muscle regeneration, it must be remembered that muscular tissue having undergone extreme degrees of atrophy will not regenerate to a degree of functional usefulness when reinnervated. It must also be remembered that defective regeneration frequently occurs in muscles which have undergone prolonged overstretching; also that muscles showing signs of regenerative changes may be late in responding to voluntary motor impulses, due to concomitant functional conditions and lack of reeducation. A muscle showing signs of increasing electrical and mechanical irritability is undoubtedly regenerating; a muscle completely devoid of electrical and mechanical irritability either has undergone complete degeneration, resulting in a fibrous transformation, or else it remains denervated.

NEURAXON REGENERATION

Neuraxon regeneration in its early stages may be determined only by Tinel's sign of formication. Regeneration, following suture of a peripheral nerve, occurs after the first month at the rate of about 3 cm. per month, providing regeneration is unimpeded. After the third month, formication should be elicited in the lower segment of the nerve; at the end of six months, it should he elicited 5 or 6 inches below the suture line with about the same intensity as the reaction obtained at the suture line. While this reaction gives evidence only of sensory fiber regeneration, it is probable that motor fibers contained in the same sheath have regenerated to the same degree. Tinel's sign is the only means we have at our disposal at the present time for determining neuraxon regeneration, until the regenerating motor and sensory fibers have reached their respective terminations and manifested their presence by a return of sensation and muscle regeneration. The absence of Tinel's sign in the distal segment of a sutured mixed peripheral nerve, after four to six months, definitely indicates an absence of neuraxon regeneration, providing this sign may be elicited with intensity at the suture line. When Tinel's reaction in the lower segment of the nerve is of considerably less intensity than at the suture line, defective regeneration may be definitely concluded. The state of a negative,


976

or seriously defective neuraxon regeneration calls definitely for surgical intervention.

In certain instances. neuraxon regeneration may progress favorably down the distal segment of the repaired nerve without restoration of motor or sensory function, due to torsion of the nerve trunk during suture, with misdirection of fibers, resulting in their complete physiologic loss. This accident may occur to any degree, and doubtless many defective regenerative results are due primarily to this cause. We may, in general, classify regenerative failure into three groups: (1) Absent or defective neuraxon regeneration; (2) defective regeneration, due to disturbance of nerve pattern during primary suture;(3) persistence of muscle degeneration after reinnervation.

1. The factors responsible for absent or defective neuraxon regeneration will usually be found at the suture line. They are:
a. Defective approximation, in which the proximal segment of the nerve presents a secondary neuroma which calls for resection and accurate approximation.
b. Construction in the region of the primary suture by a scar infiltrated nerve sheath not sufficiently resected at the primary operation, which calls for decompression. The surgeon should assure himself that the exposed intraneural contents are not subject to extensive scar invasion; if such be the case, complete resection and a new approximation will be indicated. If the intraneural contents are found to be soft to palpation without evidence of scar infiltration, the defect should be considered as having been corrected by the decompression.
c. Construction by perineural scar tissue is usually the result of excessive tissue trauma, defective hemostasis, or infection following suture. Occasionally the failure to provide a suitable bed for the repaired nerve may be responsible; this obstruction to regeneration sometimes follows the use of foreign tissue, which at one time was used extensively to protect the suture line. Undoubtedly, experience has shown that the best protection a nerve may have is its own sheath; that an excess of scar tissue subsequent to operation can usually be avoided by paying due respect to the above enumerated causes. After the nerve trunk has been freed from constrictive extraneural scar tissue and found to be free from intraneural scar invasion, it should be transposed to a new and more favorable bed.

2. Defective regeneration, due to torsion of the nerve trunk during the original suture, calls for resection of the original suture line and reapproximation; in this, the surgeon must resort to topographical identification, which is considered in detail under the surgery of individual nerves.

3. Failure of muscle regeneration does not call for surgical intervention, in so far as the nerve is concerned. When the degenerated condition of the muscles has persisted for two or three years following a nerve suture with successful neuraxon regeneration, they should be considered as irretrievable lost and recourse should be taken to such supplementary procedures as will assist in correcting the defect. (See irreparable lesions of special nerves, p. 959.)


977

TENDON TRANSPLANTATION

Tendon transplantation, in peripheral nerve lesions, is indicated for the correction of defective motor regeneration, whatever may be its cause. It should not be resorted to, particularly when it entails a sacrifice of muscles and tendons, as most transplants do, until the persistence of the motor defect is definitely established or the extent of the nerve injury precludes the possibility of repair. Two exceptions to this rule at the present time are pronator teres transplantation for extension of the wrist, and the author's transplant for restoring opponens function to the thumb; both of these procedures may he executed without jeopardizing subsequent muscle regeneration or interfering with muscle action after regeneration. In fact, both procedures, particularly the latter, tend to relax the paralyzed muscles, making conditions more favorable for their regeneration. In irretrievable lesions of a peripheral nature, tendon transplantation offers a very valuable means of restoring, to some extent at least, movements lost by the original paralysis.

The general principles governing tendon transplantation are primarily the same in both upper and lower extremities, though the chief desideratum in the upper extremity is mobility, while in the lower extremity stability for weight bearing is more essential. To hope for successful functional end results following tendon transplantation, the surgeon must adhere strictly to certain general principles which experience has proved to be indispensable; the neglect of these primary requisites is but to invite failure, following which the after results mav be more unsatisfactory than the former.

An important essential to the success of tendon transplantation lies in the mobility of those joints upon which the transplanted tendons act. A joint whose function is fixed in deformity will never respond to the action of transplanted tendons. Before tendon transplantation of any type is attempted, the surgeon should assure himself that the range of motion in these joints which are to be influenced by the transplantation must be fully adequate to functional needs and without restriction. An operation for tendon transplantation, therefore, must be definitely postponed until restricted joints are fully mobilized and tendon and fascial contractures completely corrected.

For restoration of function through tendon transplantation, only muscles with sufficient power and contractile vigor should be used. It is therefore essential that the surgeon be familiar with the ways and means of determining individual muscle function, and in the selection of a muscle for transplantation, its ability to do the proposed work should be definitely determined. A normally weak muscle should never be expected to assume a function requiring greater power than it ordinarily possesses. Though a muscle may hypertrophy, to a certain extent, to assume an additional burden, its power to do so is markedly limited and should not be too greatly relied upon.

While various rules have been promulgated from time to time, on the grounds of synergetic action, for the selection of muscles to be used in tendon transplantation, practical experience has demonstrated that certain deviations from these rules have been attended with considerable success. It is important, however, in selecting a muscle for the performance of a specific function that


978

it be adapted physiologically to that function. Although it is generally held that an antagonist should never be selected for tendon transplantation, it should be remembered that anatomic antagonists may not, in a strict sense, be physiologic antagonists. In the synergetic action of muscles, for the accomplishment of smoothly adjusted movements, a contraction in a certain group of muscles will call into play a relaxation of a physiologic antagonistic group. In the execution of certain other movements, however, a muscle anatomically antagonistic may contract simultaneously with its opponent, to fix a joint in a position which will supplement the action of the other muscles.

The determination of synergetic action in various movements is extremely complicated and can be comprehended only after a careful study of individual and combined muscle action. The surgeon is materially assisted, however, in the determination of those muscles of separate innervation, which contract simultaneously for the execution of a given movement, by the study of "trick or substitution" movements commonly seen in peripheral nerve injuries.

The maintenance of direct muscle pull from origin to insertion should be assured in tendon transplantation by long incisions and adequate tendon and muscle mobilization. A muscle with an obliquity of pull is much less effective than one pulling in a direct line. In the transplantation of tendons, therefore, free tendon mobilization and the avoidance of angulation are primary requisites.

The tunnel best adapted to transplanting is one directed beneath the fat of the superficial fascia, where adhesions are less apt to be dense. It is advisable, in planting the original incision, to avoid if possible the superimposing of skin and tendon sutures. The amount of tension exerted upon a transposed tendon is frequently responsible for success or failure. A muscle is an elastic structure and its inherent elasticity and tone will cause shortening when its tendon is divided. In making an anastomosis, it should be held sufficiently taut to restore its normal tension, but must in no wise undergo a degree of stretching sufficient to produce a paresis. The joint to be acted upon must be flexed or extended to the degree to which the transplanted tendon is expected to act. If it is desired to procure complete extension of hand and fingers, they must be so maintained, in extension, during approximation of the tendon, and the anastomos is effected under sufficient tension, that when the anastomosis is completed and the muscle completely relaxed, it still is under a degree of tension comparable to that existing in the muscle previous to its transplantation.

After suture, the member must be maintained continuously in the position under which it was placed during transplatation, lest the sutures in the anastomosed tendon be avulsed or broken. This requires considerable care and attention on the part of the assistant during closure of the integument and the application of dressings. It is usually advisable to prepare a suitably moulded splint in advance, in order that it may be slipped into position without delay and without jeopardizing the anastomosis. In the preoperative preparation of the splint, it should be moulded to maintain exactly that position which the tendon transplant is expected to effect during its maxi-mum action; however, the joints involved should not be sufficiently extended


979

or flexed to produce discomfort. The splint should be so arranged that removal of dressings is possible without disturbing the position of the extremity.

After the second week, passive movements may be instituted to prevent joint restriction, and to stretch the adhesions about the anastomosis before fixation occurs. These active and passive movements must be executed with extreme care, and controlled by one who is familiar with the nature of the operation and experienced in the treatment of these conditions. If healing of the skin is complete, gentle massage of the extremity may now be practiced.

After a period of one month the splint may be entirely removed during treatment, after which it is replaced. Two weeks later, the splint may be removed for a short time each day, which period may be lengthened as the transplanted muscles develop in strength. If after discarding the splint the muscles seem to become weaker, it should again be applied and worn continuously for a few weeks. A night splint is important and should be worn for a period of at least two months following the discarding of the day splint.

Reeducation consists in explaining to the patient the nature of the operation, demonstrating the original action of each muscle, and then the acquired action through tendon transplantation. It has been the writer's practice to have the patient instructed before the operation in the individual use of those muscles which will be used in transplantation. Frequently the use of electricity will demonstrate to a patient more clearly than any other means the action of his transplanted muscle. The progress in reeducation depends less upon the individual than upon the nature of his instruction. It is expedient that the surgeon keep these patients under his personal supervision until such a time as the ability is acquired to use his transplanted tendon properly.

Three months is usually required for the scar tissue, uniting the tendons, to become sufficiently hard to resist the strain of ordinary functional demand, and until this period has elapsed care should be used in preventing any undue strain falling upon the transplanted tendons.

INFECTED WOUNDS AND NERVE SURGERY

It is folly to attempt nerve repair in the presence of an infection or an open wound, and if during the course of a dissection pus is uncovered in bone or soft tissue, the attention should be directed to the removal of the infected area, and the nerve repair postponed until the infection is completely controlled and the wound entirely healed.

RECRUDESCENCE OF INFECTION IN CICATRIZED WOUNDS

Frequently our attention has been attracted to the recrudescence of an infection in wounds which have for many months been completely cicatrized. Wounds which have undergone a long period of suppuration may harbor latent infection for months or even years following their cicatrization, and during this period any form of tissue trauma, surgical or otherwise, may serve to initiate a recrudescence of the latent infection, which may be attended with high temperature, great pain, and all the symptoms of a local or even a general sepsis. It is essential, therefore, that a sufficient period of time be allowed to elapse following the healing of an infected wound to insure to some degree the


980

safety of a surgical invasion. When suppuration has attended the healing of a wound involving only soft parts, an aseptic surgical operation in the vicinity of this wound should not be attempted until cicatrization has been complete for a period of at least three months. When the suppurative process has involved osseous structures, the period of waiting should be prolonged to at least six months, and in many instances the operation will not be safe for a period of one year, particularly if there is a history of repeated breaking down of the wound. Inasmuch as this attitude of conservatism or watchful waiting may jeopardize the ultimate outcome of nerve repair, because of delayed intervention, a more active policy may be pursued for the clearing up of scars suspected of harboring infection.

We have satisfactorily used the following method of testing for latent infection: For two days previous to the test the patient is confined to bed and his temperature recorded every four hours, a leucocytic count having been made on two alternate days. The scar is then subjected to vigorous massage for a period of five minutes. Following this, for one week the four-hour temperature is recorded and daily leucocytic counts continued. Careful daily inspection of the wound is made for local signs of inflammation, such as swelling, redness, or increased local temperature. A wound harboring infection will usually reveal its presence by both a local and a general reaction. An elevation of temperature, an increased leucocytosis, or signs of a local inflammatory nature will usually be observed if there be a dormant infection. If no reaction be elicited, it may be considered safe to attempt a clean operation; if, however, a reaction does occur following the massive massage, excision of the scar tissue followed by wound sterilization and closure should be attempted, and the nerve operation delayed until this wound has been completely healed for a period of three months.

PRIMARY OPERATIONS

There can be no doubt regarding the advantage of immediate nerve suture following wounds in which, during examination, the nerve is found to be completely divided. Inasmuch as most lacerated wounds are infected, the question naturally resolves itself into a discussion of the advantages and disadvantages of suturing a nerve in the presence of infection, or where infection seems inevitable. Before the days of wound excision and successful sterilization, one could not suture a nerve and hope for satisfactory regeneration to any degree.

If wound excision and sterilization seem probable, the surgeon is justified in approximating the ends of the nerve with fine sutures, or even by a single tension suture, with the primary object of preventing retraction and fixation of the nerve ends, thereby diminishing the extent of the continuity defect during the process of healing and perhaps making less difficult a secondary operation. In any wound potentially infected, the excision of nerve ends and the extensive exposure necessary to satisfactory approximation should be vigorously discouraged, as this procedure invariably opens up new channels for infection, is rarely successful, and greatly complicates the almost inevitable secondary suture. It is therefore advisable in all open wounds, complicated by nerve injury, to first control infections, delaying the operation for nerve repair until such a time as it may be aseptically accomplished.


981

NERVE INJURY DURING DÉBRIDEMENT

The efficacy of wound débridement has become established. In the hands of the skilled and experienced surgeon, d bridement with primary suture has in many instances resulted in primary healing, but in the hands of a novice it is a procedure nothing less than hazardous. There can be little doubt that many nerves receive their first injury from the scalpel of the surgeon during débridement. The writer, on one occasion, witnessed the excision of 2 inches of the musculospiral nerve, during a spectacular débridement of a gunshot wound.

Every injured extremity, having undergone débridement or any other operative procedure subsequently revealing a nerve injury, casts serious reflection on the operating surgeon, if the preoperative record fails to record the presence of that nerve lesion. The ease and rapidity with which an injured extremity may be examined for the presence of a complicating nerve injury leaves no excuse for its oversight in the preoperative records; a later discovery certainly throws the burden of responsibility upon the operator.

AMPUTATION NEUROMAS

The development of a neuroma at the end of a nerve, particularly when the neuroma is incorporated within the scar of an amputation stump, may cause extreme discomfort or pain, which is intensified by the pressure of amputation prosthetic appliances. To prevent subsequent annoyance of a painful amputation neuroma, the nerve, at the time of amputation, should be withdrawn and resected at as high a level as possible and injected with 80 per cent or absolute alcohol. If a nerve is not injected with alcohol, a neuroma will invariably form, but if the neuroma is located some distance above the cicatricial tissue of the stump, it is seldom subjected to irritation.

TREATMENT

A neuroma incorporated in the scar tissue of a stump usually calls for removal, with high section of the nerve. Frequently it is not considered advisable nor desirable to increase the scar of the stump by an additional incision; in such instances the nerve involved may be exposed some distance above the neuroma, injected with alcohol and sectioned. This procedure is quite as satisfactory as the removal of the neuroma, as it permits degeneration of its contained nerve fibers. Occasionally, the surgeon may resort to a plastic procedure on the larger nerves to prevent the formation of a subsequent neuroma.(See fig. 164.)

It is essential in each instance to determine the nerve responsible for the irritation. Pressure or percussion in the painful region will usually elicit tingling, which is localized in the cutaneous area normally supplied by that nerve. For example, if after an amputation of the arm a painful neuroma develops in the stump, and percussion of that neuroma elicits tingling localized in the little finger, the surgeon will know that the neuroma is of the ulnar nerve; if it is located in the second or third fingers, the median nerve will require resection. It is important to accurately determine the nerve responsible for


982

the pain. The writer has on several occasions seen a persistence of pain after neuroma removal, obviously because there were several neuromas in the stumpand the neuroma responsible for the painful syndrome was unfortunately missed, a proper localization evidently not having been made.

FIG. 164.- Plastic procedures and alcoholic injection to prevent the formation of amputation neuromas

THE BRACHIAL PLEXUS

GENERAL ANATOMY

The brachial plexus is formed by the anterior primary divisions of the fifth, sixth, seventh, and eighth cervical and first thoracic nerves, located in the posterior triangle of the neck and in the axilla. While the brachial plexus varies often in its formation, in that the fourth cervical nerve and the first thoracic may give a greater or lesser number of fibers to its formation, its general anatomy is fairly constant. Minor variations in the origin of branches will occasionally be found, but these anomalies are not sufficiently great to modify the surgical principles involved in attempting physiologic reconstruction of the plexus.

The plexus first consists of three primary trunks, lying upon the scalenus medius muscle; the upper trunk is formed by the fifth and sixth cervical nerves, the middle trunk by the seventh cervical, and the lower trunk by a


983

union of the eighth cervical and first thoracic nerves. Each of these primary trunks subsequently divide into an anterior and posterior division, which ultimately form the three plexus cords. The anterior divisions of the upper and middle trunk form the outer cord, which terminates by dividing into two branches, the musculocutaneous and the outer head of the median. The anterior division of the lower trunk forms the inner cord, which ultimately breaks up to form the inner head of the median, the ulnar, and the greater and lesser internal cutaneous nerves. The posterior divisions of the three primary trunks unite to form the posterior cord of the plexus, which terminates by dividing into the circumflex and musculospiral nerves. In the region of the clavicle, the three primary trunks divide into their anterior and posterior divisions, which are often comparatively short and lie lateral to the subclavian artery in close relation to the apex of the lung. Just below the clavicle, the anterior and posterior divisions of the primary trunks arrange themselves into the three plexus cords, which surround the axillary artery in its second portion, the outer cord lying upon the outer side of that vessel; the inner cord on its inner side; and the posterior cord behind. In the region of the coracoid process of the scapula, these cords divide into their terminal branches to supply the musculature of the arm and provide sensation to its integument.

A number of branches are given off from the brachial plexus to muscles of the shoulder girdle; the course of most of these is relatively short and they do not often lend themselves readily to surgical repair. Attention, however, must be directed to some of the larger and more important branches, which must be recognized and preserved in the surgical reconstruction of the plexus.

The long or posterior thoracic nerve usually arises by three roots from the fifth, sixth, and seventh cervical nerves, close to the intervertebral foramina. This nerve is very long and passes down behind the brachial plexus and axillary vessels along the outer surface of the serratus magnus muscle, which it supplies.

The suprascapular nerve arises from the lateral border of the upper trunk and passing obliquely outward beneath the trapezius and omohyoid, supplies the supraspinatus and infraspinatus muscles. In the dissection of the upper trunk, the suprascapular nerve should be identified and carefully preserved, or individually sutured if necessary, to preserve or restore function in the spinatus muscles; this is particularly important in deltoid paralysis.

The subscapular nerves, usually three in number, arise from the posterior cord below the clavicle and supply the subscapularis, teres major, and latissimus dorsi muscles. The importance of preserving these branches is obvious in the presence of a deltoid paralysis.

The anterior thoracic nerves are two in number and supply the pectoral muscles. The external and larger of the two arises from the outer cord of the plexus and passing inward crosses the axillary artery and vein, where it is distributed to the under surface of the pectoralis major. The inner anterior thoracic nerve arises from the inner cord and passing behind the axillary artery, it curves forward between the axillary artery and vein to the undersurface of the pectoralis minor. Anterior to the axillary artery, the external and internal thoracic nerves are commonly united by a filament from the external nerve.


984

It should be remembered, therefore, in the dissection of the brachial plexus that in exposure of the upper trunk, the suprascapular nerve must be identified and preserved. In dissection of the posterior cord, the subscapular nerves require the same consideration, and in the dissection of the outer and inner cords, the external and internal anterior thoracics should be identified and preserved to retain pectoral function. The thoracic nerves are particularly liable to injury during dissection and isolation of the axillary vessels; their early identification during this procedure is important.

SURGERY

Traumatic lesions of the brachial plexus are fairly common in both civil and military life. The nature of these injuries may be direct, in which portions of the plexus are involved by penetrating wounds; or indirect, which are due more frequently to stretching of the plexus, the mechanism of which consists in forcibly depressing the shoulder while the neck and head are forced to the opposite side, greatly increasing the shoulder and neck angle. Indirect injury may result in lesions of varying degree, and in severe cases roots may be completely avulsed from the spinal cord. The avulsion more commonly takes place just external to the intervertebral foramina and often is confined to the upper roots of the plexus. Not infrequently, such injuries are limited to an overstretching of the upper trunk, which may or may not cause rupture of its intraneural constituents, the entire nerve sheath usually remaining intact. In such lesions spontaneous regeneration will probably occur under appropriate splinting and relaxation treatment. There are, however, certain anatomic factors in this region which tend to inhibit spontaneous regeneration. The brachial plexus is closely covered by the prevertebral layer of the deep cervical fascia, and indirect lesions of the plexus, with or without rupture, are often attended with hemorrhage under the fascia layer, which may be responsible for primary compression and ultimate more or less dense scar formation. These adhesions between the plexus and the prevertebral fascia are the most conspicuous features of the lesion, which upon exposure presents itself as a mass of scar tissue. It is usually only after careful sharp dissection of this fascia from the plexus that it is possible to proceed with identification of the individual anatomic plexus constituents, and not infrequently a thorough neurolysis of the plexus will be all that is required. While lesions of the upper trunk and outer cord are more common following indirect injury from stretching, lesions of the lower portion of the plexus more often result from direct trauma. The proximity of the lower plexus to the apex of the lung and important vascular channels makes penetrating wounds in this region particularly serious, frequently leading to an early fatal termination from hemorrhage. The relation of the trunk of the brachial plexus to the axillary artery is responsible for many combined lesions in which hematoma and subsequent scar tissue simultaneously compress the nerve trunks and vascular structures, combining ischemia with compression paralysis and often resulting in painful paralytic syndromes. Aneurysm is not an uncommon complication in lower plexus penetrating wounds. The relationship of the lower trunk of the plexus


985

to the first rib and the pleura are likewise anatomic conditions peculiar to the plexus, often greatly complicating the nerve lesion.

Exact anatomic knowledge and adequate exposure are essential in nerve surgery, regardless of the location; in the surgery of the brachial plexus these primary requisites are indispensable. At best, the end results of severe brachial plexus lesions are disappointing. Manifestly, many of these severe lesions are practically hopeless from a surgical standpoint, due to the fact that the lesion lies within or closely approximating the intervertebral foramina. It is well that the surgeon appreciate the technical difficulties which at present limit anatomic reconstruction in lesions involving the roots within or proximal to the intervertebral foramina. Plexus injuries associated with avulsion of the roots within the spinal canal are usually associated with some cord involvement, which will be evident on a systematic neurologic examination. Lesions within the intervertebral foramina will usually involve both anterior and posterior divisions of the cervical nerves, which will be evidenced by the addition of a dorsal anesthesia in the cervical region corresponding to the dorsal rami of the cervical nerves. The motor branches contained in the dorsal rami are distributed in a manner which does not readily lend itself to this differentiation; however, the sensory impairment referred to should place the involvement within the intervertebral foramina. The presence of a paralysis of the serratus magnus innervated through the long thoracic nerve whose branches are given off from the anterior divisions of the fifth, sixth, and seventh cervical nerves close to the intervertebral foramina will indicate a lesion in this location, particularly if the rhomboids are found to be involved. Therefore (1) brachial plexus lesions associated with spinal cord involvement usually preclude the possibility of repair; (2) lesions associated with a dorsal area of anesthesia corresponding to the dorsal cervical divisions likewise preclude the possibility of repair, due to involvement within the intervertebral foramina; (3) lesions associated with paralysis of the serratus magnus and rhomboids make extremely doubtful the possibility of repair, due to the proximity of the injury to the intervertebral foramina; (4) plexus lesions associated with sym- pathetic involvement (Horner's syndrome-myosis, enophthalmus, and narrowed palpebral fissure) indicate a lesion involving the eighth cervical nerve within or immediately adjacent to the intervertebral foramen, which greatly diminishes the possibility of repair in this region.

While the presence of the above types of brachial plexus lesions usually precludes the possibility of repair, the surgeon may feel the desirability of giving the patient the benefit of any doubt; the justification of assuming the risk of along, tedious exposure is found in the fact that in the hands of an experienced neurologic surgeon there is little chance of increasing the disability and there may be a possibility of reconstructing some portion of the plexus. However, upon exposure, if the surgeon is content with simply approximating nerves, without regard to original physiologic pattern, even though the operation be followed by a regeneration of nerve fibers, the physiologic outcome will be such a distortion of motor and sensory function, as to exclude any possibility of usefulness in the extremity. The writer has seen several instances of brachial plexus lesions, following suture, in which neuraxon regeneration was active,


986

but with distortion of nerve pattern, resulting in most perplexing combinations of motor response and sensory localization. At least in the repair of the brachial plexus, the surgeon should by all means endeavor to approximate corresponding roots, trunks, and terminal branches, though even when this is accomplished and regeneration of nerve fibers is successful, it will be found that if there has been a torsion of the nerve trunk during suture, the ultimate physiologic outcome in that particular trunk will be more or less defective.

These facts will serve to partially explain the many instances in which sutures of the brachial plexus have been attended with such poor functional results. In order to avoid these unsatisfactory terminal results, the surgeon should fortify himself with such degree of anatomic and physiologic familiarity with the brachial plexus as will permit him, during repair, to conserve, as much as possible, its original pattern. If the surgeon feels that he is not sufficiently familiar with the intrinsic plexus anatomy to practice physiologic reconstruction, he had better confine his efforts strictly to a careful and painstaking neurolysis--which conservative attitude, in all probability, in the great majority of cases, will be attended with a more satisfactory functional end result. When, however, there has been an actual division of a portion of the plexus, its repair should be conducted with all precautions toward the prevention of distortion of the nerve pattern, by eliminating torsion during suture.

Most direct lesions involving the brachial plexus above the clavicle result in injury to the primary trunks, or their anterior or posterior divisions, while lesions below the clavicle usually involve the cords or their terminal branches. Penetrating wounds seldom involve the entire plexus, and it is usually possible, by careful preoperative examination of individual muscle function, to determine the exact location of the lesion. The surgeon so armed with a precise knowledge of plexus anatomy and clinical localization, will be able, after identification of the various intact branches by faradic stimulation, to definitely locate the lesion. This having been accomplished, should complete division be found, he will be able to effect reconstructive suture by uniting respective trunks without torsion with the aid of identification sutures and branch identification. For example, if the lesion compromises the terminal portion of the inner cord, involving the inner head of the median, ulnar, and internal cutaneous, the relative position in reconstructive suture would be: The inner cord of the median to the upper or lateral part of the anterior division of the lower trunk; the internal cutaneous would occupy the lower and medial side of the trunk; the ulnar, the center of the trunk between the inner head of the median and the internal cutaneous. To prevent torsion of the proximal segment, the position of the eighth cervical nerve would indicate the upper or lateral portion of the root; the position of the first thoracic, the lower portion of the root; and the posterior division, its posterior surface. This type of anatomic reconstruction during suture, while by no means perfect, in that shunting of fibers is still possible in the terminal segments, will promise the highest degree of efficiency compatible with our present knowledge of ways and means to preserve intraneural topography. While this is purely an anatomic method, a careful study of the nerve trunk, its formation, the


987

position of terminal branches, etc., will usually serve to prevent serious torsion. The writer finds that the use of an enlarged anatomic chart in a convenient position in the operating room is of great service in assisting the visualization of reconstructive requirements in the surgery of brachial plexus lesions. The chart should indicate the respective muscles innervated by each primary, secondary, and terminal division of the plexus, so that in electrical stimulation, precision in identification may be facilitated. It is needless to state at this point that in preparing the field for operation the entire arm should be included, in order that it may be properly exposed and manipulated for inspection of muscle action during electrical stimulation of nerve trunks and branches.

TECHNIQUE OF EXPOSURE

The patient is placed upon his back with head abducted and rotated to opposite side, and shoulder depressed; a sand bag placed between the shoulders is often of value in maintaining this position. A "step-shaped" incision has been found the most desirable for complete exposure of the plexus. The upper longitudinal limb of the incision parallels the posterior border of the sternomastoid muscle to a point just below the clavicle; turning outward, the incision transverses the thorax along the lower border of the clavicle to the junction of its outer and middle third; from this point it passes down over the anterior aspect of the shoulder to the medial side of the coracoid process, and extends down the arm following the medial border of the biceps for a short distance below the insertion of the pectoralis major. The triangular flap of skin and platysma between the upper longitudinal and infraclavicular incision is reflected outward, at which time the external jugular vein is met and divided between forceps. To gain a better exposure of the lower trunk of the plexus, the later altendon of insertion of the sternomastoid muscle may be divided close to its attachment to the clavicle. The deep fascia now being exposed is divided along the course of the skin incision and retracted laterally with its underlying cushion of fat. During this procedure the posterior belly of the omohyoid muscle is exposed and may be divided to obtain a more satisfactory exposure of the plexus. The prevertebral layer of the deep fascia is now exposed and the several transverse cervical vessels lying in front of it are ligated and divided. The prevertebral fascia is divided longitudinally, and the lateral border of the scalenus anticus muscle identified and exposed as it lies under the sternomastoid. The upper trunk of the plexus, formed by the anterior divisions of the fifth and sixth cervical nerves, will be found emerging from the lateral borderof the scalenus anticus. These trunks lie upon the scalenus medius muscle, the fifth being found to emerge from behind the lateral border of the scalenusanticus, opposite the transverse process of the fifth cervical vertebra, and if followed downward, it will be found to unite with the sixth, forming the upper trunk of the plexus, which in turn divides into three branches, a smaller and lateral branch, the suprascapular nerve; an anterior branch (anterior division)uniting with the anterior division of the middle trunk to form the outer cord of the plexus; and a posterior branch (posterior division) uniting with the poste-rior divisions of the middle and lower trunks to form the posterior cord. If the


988

lesion extends below the upper trunk, involving its anterior or posterior division, it will be necessary to divide the clavicle in the exposure. This division of the clavicle should be done subperiosteally, in its middle third, with a Gigli saw. It is advisable to drill holes through the clavicle before its division, to facilitate its later approximation with wire.

When lesions involve the plexus in its lower or subclavicular region, the lower vertical limb of the "step-shaped" incision is carried down the arm as far as exposure seems necessary, usually some distance below the insertion of the pectoralis major muscle. The line of cleavage between the pectoralis major and deltoid is found and these muscles are separated and retracted; while

FIG. 165.- Infraclavicular exposure of brachial plexus. Line of cleavage deepened between pectoralis major and deltoid; forceps passed under pectoral tendon, preparatory to division. A, Deltoid; B, cephalic vein C, biceps; D, pectoralis major

the adjacent fibers of these two muscles run parallel, the cephalic vein indicates the line of cleavage. In making this separation, the vein should be allowed to remain on the deltoid side and be carefully preserved, the line of cleavage being followed to the clavicle. To obtain exposure of the subclavicular portion of the plexus, the insertion of the pectoralis major into the humerus and of the pectoralis minor into the coracoid process may be divided, and these muscles retracted medially; after this retraction, the subpectoral fascia is divided, exposing below the neurovascular bundle. The lateral or medial end of the divided clavicle may be retracted by strips of heavy sterile bandage. The cords of the brachial plexus may now be identified by following upward their


989

terminal branches. The identification of the median nerve, which is the most superficial, is facilitated when followed upward, by its division into its respective outer and inner heads, in the fork of which will be found the axillary artery. If the outer head of the median is followed upward, it is found to join with the musculocutaneous nerve, forming the outer cord of the plexus. Following the inner head upward, it will be found to unite with the ulnar nerve and the internal cutanei, forming the inner cord of the plexus; in this manner both outer and inner cords are readily identified. In the axilla, the axillary artery is found lying immediately posterior to the median nerve. The loose connective tissue

FIG. 166.- Infraclavicular exposure of brachial plexus; pectoralis major divided and reflected; neurovascular bundle exposed and sheath held open by forceps; biceps retracted laterally to facilitate exposure. A, Deltoid; B, pectoralis major tendon, divided; C, coracobrachialis; D, biceps; E, median, outer head; F. median, inner head; F, median nerve; U, ulnar nerve; I, internal cutaneous nerve; J, pectoralis major, reflected; K, neurovascular bundle; L, axillary artery

uniting these structures may be carefully separated, and posterior to the artery the musculospiral nerve is identified, which when followed upward is found to unite with the circumflex, and by this union to form the posterior cord. Probably the greatest difficulty in isolating the components of the neurovascular bundle below the clavicle is found in the bleeding from the division of numerous venous radicals. It is wise, therefore, early in dissection, to carefully isolate the axillary vein, beginning this dissection in the lower part of the incision. The presence of scar tissue greatly complicates the dissection of that portion of the plexus which lies in direct contact with the axillary artery and vein; if there is


990

evidence of an aneurysm, it is well to carefully expose the artery proximal to the aneurysm before its dissection is attempted, so that control of hemorrhage maybe assured in case of accident. In isolating the axillary vessels in this region, caution should be used toward identifying and preserving the anterior thoracic nerves, the external of which in arising from the outer cord of the plexus passes over the artery, while the internal, arising from the inner cord passes under the artery; both branches usually pass over the axillary vein. If the preoperative report shows the pectoral muscles to be intact, the identification of the anterior thoracic nerves may be facilitated by electrical stimulation. These nerves

FIG. 167.- Infraclavicular exposure of brachial plexus. Median nerve retracted medially, exposing origin of outer head of the median and musculocutaneous nerve, behind which is seen the musculospiral nerve passing over the latissimus dorsi tendon, and the circumflex nerve accompanied by the circumflex artery, passing through the quadrilateral space in the posterior wall of the axilla. A, Deltoid; B, coracobrachialis; C, musculospiral nerve; D, latissimus dorsi tendon; E, coracoid process; F, pectoralis minor, tendon divided; G, median, outer head; R, musculocutaneous nerve; I, median nerve; J, circumflex nerve; K, circumflex artery; L, pectoralis major, tendon divided

enter the muscles on their under surface. In exposing the lower trunk of the plexus, both artery and vein must be well isolated and retracted downward. A mastery of the situation in difficult brachial plexus dissections can be assured only after thorough isolation of the adjacent vascular channels, and the preliminary attention given to this aspect of the procedure, though often time-consuming, is in the end a time-saving procedure. Occasionally the artery, will be found so constricted by scar tissue as to render it a nonpulsating cord, which may closely resemble a nerve in appearance. In such cases the preliminary identification and exposure of the artery will avoid much trouble. It has occasionally been possible to restore pulsation in a compressed brachial


991

artery by the removal of constricting scar tissue. If this procedure fails to restore circulation, the vessel is probably obstructed by a thrombus. If collateral circulation is defective, arteriotomy may be attempted and the thrombus removed with the aid of a dull ring curette. After a free flow of blood has been obtained from both directions, the vessel may be closed with ordinary vascular sutures. When brachial plexus lesions are associated with irritative syndromes, the operator should never neglect the removal of scar tissue from the artery, and occasionally a decortication is indicated, after the method of Leriche. The importance of vascular lesions in this region, in the production of both painful and ischemic syndromes, can not be overestimated, and any effort devoted to the reconstruction or liberation of vascular channels is well worth while, even though collateral circulation seems to be adequate.

At this point it might be well to reiterate the necessity of physiologic preservation of the various plexus components and to emphasize the importance of conservatism in resection. Sutures of the plexus proper, because of distortion of normal nerve pattern, have given very unsatisfactory results; and unless the surgeon feels confident in his ability to preserve, at least to some degree, the original intraneural topography, far better results will probably attend a carefully executed external and internal neurolysis. The internal neurolysis, however, should be confined almost entirely to the resection of scar-invaded portions of the nerve sheath. Suture of the terminal divisions of the plexus is a far more satisfactory procedure in that it is possible to preserve in a large measure the original nerve pattern by the prevention of torsion during suture.d

THE CIRCUMFLEX NERVE

GENERAL ANATOMY

The circumflex nerve receives its fibers from the ventral division of the fifth and sixth cervical nerves through the posterior cord of the brachial plexus, in which it occupies a position lateral to the fibers forming the musculospiral, though within the same sheath. Near the outer border of the subscapularis the circumflex leaves the musculospiral to pass with the posterior circumflex vessels through a quadrilateral space in the posterior wall of the axilla, which is formed by the teres minor and subscapularis above, the teres major below, the long head of the triceps medially, and the humerus laterally. The nerve then winds around the surgical neck of the humerus posteriorly, and lying upon the under surface of the deltoid sends fibers to this muscle. In the region of the quadrilateral space the circumflex nerve divides in to an upper and a lower branch. The upper branch winds around the surgical neck of the humerus and supplies the bulk of the deltoid muscle; it also distributes cutaneous branches which pierce the muscle and supply the integument over the deltoid region. The lower branch gives off filaments to the teres minor and to the posterior fibers of the deltoid, after which it pierces the integument, which it supplies with sensation over the lower posterior deltoid region.

d The technique of physiologic approximation in suture of the terminal branches of the brachial plexus is considered in detail under the surgery of these respective individual nerves.


992

SURGERY

AXILLARY EXPOSURE

Exposure of the circumflex nerve is obtained, as in low lesions of the brachial plexus, with the arm in outward rotation and marked abduction. The incision is made over the course of the neurovascular bundle from the middle of the clavicle to a few centimeters below the tendon of the pectoralis major. The deltoid is separated from the pectoralis major and the cleft between these muscles deepened to expose the deep pectoral fascia. (See figs. 165,166, 167.) The pectoralis major is divided near its insertion into the humerus and reflected. The deep pectoral fascia is now divided to expose the neurovascular bundle. The median nerve and the axillary artery are identified and retracted medially, exposing the musculospiral nerve passing over the latissimus dorsi tendon. The musculospiral nerve is now followed upward toa, point near the insertion of the pectoralis minor, where the circumflex will be found joining its lateral side to form the posterior cord of the brachial plexus. When thus identified the circumflex nerve may be followed around the medial side of the neck of the humerus to where it leaves the axilla by passing through the quadrilateral space to the dorsum of the arm.

DORSAL OR LATERAL EXPOSURE

Occasionally the circumflex nerve is injured after it has left the axilla in its passage around the surgical neck of the humerus. To approach lesions in this region a dorsal or lateral exposure is needed. The patient lies on his side with the arm acutely flexed across the chest, giving access to the posterior deltoid region. A longitudinal incision is now made approximating closely the dorsal border of the deltoid. This muscle is elevated and retracted forward. Retraction may be facilitated by elevation of the arm, relaxing the deltoid. Occasionally it will be found advisable to divide some of the posterior deltoid fibers. The circumflex nerve will be found hugging the neck of the humerus, as it emerges from the quadrilateral space, accompanied by the posterior circumflex artery and its venae comites.

Isolated lesions of the circumflex nerve are comparatively rare, though deltoid paralysis is frequently associated with brachial plexus lesions, particularly those involving the upper trunk. Fortunately, the muscles of the shoulder girdle are supplied by branches given off from the brachial plexus at various points, so that a total paralysis of the shoulder is seldom encountered except in complete brachial plexus lesions. Individual nerves supplying the muscles of the shoulder and shoulder girdle run a relatively short course before breaking up into terminal branches to supply their respective muscles, so that their repair is as a rule exceedingly difficult; in fact their short course usually eliminates the possibility of effecting end-to-end approximation where a defect is present. Practically the only recourse the surgeon has when confronted with a serious defect in a nerve of short course is to resort to nerve grafting or viable transplants. This is particularly true in regard to isolated lesions of the circumflex nerve.


993
 
In irreparable lesions of the circumflex nerve and in those instances in which partial brachial plexus lesions are associated with deltoid paralysis and the loss of humeral abduction, arthrodesis of the shoulder joint gives perhaps more satisfactory results than any form of tendon transplantation if the scapular muscles have been preserved. The usefulness of this operation, however, depends entirely upon the preservation of scapular rotation. e

THE MUSCULOCUTANEOUS NERVE

GENERAL ANATOMY

The musculocutaneous nerve receives its fibers from the outer cord of the brachial plexus, in common with the outer head of the median below the clavicle. In the region of the neck of the humerus it leaves the neurovascular bundle and passes laterally between the two heads of the coracobrachialis to the under surface of the biceps, where it divides into a number of branches supplying this muscle and the brachialis anticus. Its sensory portion continues down the arm, and, emerging from between the biceps and the brachialis anticus, penetrates the deep fascia and pursues a subcutaneous course down the lateral side of the forearm, supplying the integument. As the musculocutaneous nerve enters its canal between the heads of the coracobrachialis, it gives off a branch to this muscle, which, though running directly with the musculocutaneous nerve and usually incorporated within its sheath, derives its fibers from the seventh cervical and is, to all intents and purposes, a special nerve as the musculocutaneous is derived solely from the fifth and sixth cervical.

Anatomic irregularities of the musculocutaneous nerve are too varied to be considered in detail; they consist principally of various communications with the median nerve. Occasionally a portion of the median nerve may follow the course of the musculocutaneous, and after this nerve gives off branches to the biceps and brachialis anticus the divergent fibers will again join the medial trunk; in other instances the branches to the biceps and brachialis anticus may spring directly from the median trunk. It is not uncommon to find various anomalous communications between the musculocutaneous and median nerves in various degrees. A consideration of these abnormalities will assist in explaining certain types of combined lesions which are at times inclined to be rather perplexing. The possibility of anatomic irregularities should always be considered when the clinical findings point to a combined partial musculo- cutaneous and median lesion.

SURGERY

Lesions of the musculocutaneous nerve may occur in any part of its course; in the upper portion of the arm, or in the axilla, it may be associated with concomitant lesions of the neurovascular bundle, though isolated lesions are

e The writer has recently observed a case of infantile paralysis in which the deltoid was completely paralyzed. though fairly effective abduction of the humerus by the short head of the biceps was possible. The possibility was immediately suggested of changing the point of origin of the short head of this muscle to a more lateral position by inserting it into the acromlon process under the deltoid; this might also be reinforced by transplanting the long head of the triceps Into this same position. Experimental work on the cadaver demonstrates the possibility of this procedure from an anatomic standpoint. Traction on these transposed muscles in a line corresponding to what would be their action after transposition results in forcible abduction of the arm. He has not had occasion to attempt this transplant clinically, but believes that it offers a possible means of humeral abduction in deltoid paralysis when bicipital and tricipital function is preserved.


994

by no means rare. If the nerve is injured in the lower third of the arm the motor fibers usually escape injury, its sensory portion alone being involved.

EXPOSURE

The exposure of the musculocutaneous nerve in the upper portion of the arm may be made as in lower plexus lesions. A long incision follows the medial border of the coracobrachialis, beginning a few centimeters below the clavicle and extending downward to the middle of the arm. The superficial fascia is divided, exposing the coracobraehialis. The line of cleavage between the pectoralis major and deltoid is deepened and the deep pectoral fascia exposed, care being taken to preserve the cephalic vein, which runs in the cleft between these muscles. It is usually necessary to divide the tendon of insertion of the pectoralis major to obtain access to the origin of the nerve. (See figs.165,166,167.) The pectoralis major is retracted medially and the pectoral fascia covering the neurovascular bundle in the axilla divided, exposing the median nerve, which may be identified bv electrical stimulation, though this is rarely necessary. If the nerve be followed upward, it will be found to divide into an inner and outer head, between which lies the axillary artery. Springing from the lateral side of the outer head will he found the musculocutaneous nerve, which abruptly leaves the neurovascular bundle afnd enters its canal between the two heads of the coracobrachialis.

When the lesion is below the branch to the coracobrachialis, the musculocutaneous nerve may be exposed without invading the axilla. A longitudinal incision is made over the upper portion of the biceps, and the line of cleavage between its two heads found and deepened. In the recess, the musculocutaneous nerve will be found lying upon the coracobrachlialis. The nerve may now be followed along the under surface of the biceps by extending the line of cleavage between the long and short head downward through the substance of the muscle. In this way the motor branches may be adequately exposed. The motor branches spring from the lateral portion of the nerve trunk; the sensory from its medial side. A recognition of this fact will serve to facilitate physiologic approximation.

DEFECTS

Defects in the musculocutaneous nerve may be overcome to some extent by the relaxation obtained in flexing and adducting the arm over the chest. Defects not correctable in this manner, nor by stretching, will require grafting or a viable transplant. In closing the incision, if the tendon of the pectoralis major has been divided it should be carefully sutured with 20-day chromicized gut, and the arm maintained in flexion and adduction by strapping the hand to the opposite shoulder.

EXPOSURE OF TIHE SENSORY PORTION

Lesions involving the musculocutaneous nerve in the lower third of the arm are not accompanied by motor disability in this region the nerve is entirely sensory. Occasionally, however, certain irritative lesions necessitate its exposure, when it may be sectioned or subjected to alcoholic injection. The


995

sensory portion of the nerve is occasionally utilized as material for autogenous grafts. Its exposure is made through an incision along, the lateral surface of the arm corresponding to the medial border of the brachioradialis. The nerve will be found emerging from beneath the lateral border of the biceps, after which it pierces the deep fascia to become subcutaneous in the forearm. The biceps may be retracted meldially, and the nerve followed upward beneath this muscle to the junction of the middle and lower third of the humerus, where it may be sectioned without injury to motor branches.

IRREPARABLE DEFECTS

Paralysis of the muscles supplied by the musculocutaneous nerve is not particularly disabling, as their function consists almost entirely in forearm flexion, which may he very satisfactorily accomplished in biceps paralysis by the brachioradialis, which muscle is supplied by the musculospiral nerve, though flexion by this muscle is usuallv associated with some degree of supination, which is corrected by action of the pronator radii teres. also capable of some forearm flexion. If an irreparable musculocutaneous lesion is combined with a permanent paralysis of the brachioradialis, pronator teres flexion of the elbow alone will be of little practical assistance, as its flexor power is weak and always associated with marked pronation. Under such conditions, apparently the only recourse is in arthrodesis of the elbow joint, immobilizing it in a position of flexion compatible with the needs of the individual.

MUSCULOSPIRAL NERVE

GENERAL ANATOMY

The musculospiral nerve is a continuation of the posterior cord of the brachial plexus, which in turn is formed by the posterior trunks of the anterior primary divisions of the fifth, sixth, seventh, and eighth cervical nerves. In the early part of its course it occupies a position dorsal to the axillary artery in the neurovascular bundle and pursues a somewhat spiral course posteriorly around the humeral shaft, and gains the lateral surface of the arm by penetrating the external intermuscular septum. In the lower part of the arm. it passes from a lateral to a ventral position at the elbow and there divides into two terminal branchles. the radial and posterior interosseous nerves.

BRANCHES

In the axilla and on the inner side of the arm, which constitutes its medial portion. the musculospiral nerve gives off three branches: The internal cutaneous: a muscular branch to the long head of the triceps; and a muscular branch to the medial head of the triceps, sometimes called the ulnar collateral. In its posterior or intratricipital portion it gives off two external cutaneous branches and muscular branches to the lateral and medial heands of the triceps. On the external surface of the arm, lateral portion, branches are given off to the brachioradialis, the extensor carpi radialis longior, and occasionally a small branch to, the brachchialis anticus.


996

Terminal branches.-The radial nerve passes down the radial side of the forearm, under cover of the brachioradialis muscle, to the dorsum of the hand; it is purely sensory and supplies sensation to the dorsum of the lower third of the forearm, hand, and fingers.

The posterior interosseous nerve carries chieflv motor fibers; passing under the brachioradialis muscle it pierces the supinator brevis, and in its passage through this muscle winds around the shaft of the radius to appear on the dorsum of the forearm. Before entering the supinator canal it gives off branches to the extensor carpi radialis brevior and the supinator brevis. Shortly after it emerges from the lower border of the supinator it fans out into a number of branches which supply the extensor communis digitorum, extensor carpi ulnaris, and extensor minimi digiti. A little lower in the forearm other branches are given off to the extensor ossis metacarpi pollicis, extensor longus and brevis pollicis. and extensor indicis.

SURGERY

From a standpoint of surgical accessibility, we shall consider the musculospiral nerve as having medial, dorsal, and lateroventral portions, with two terminal branches, the posterior interosseous and radial nerves.

MEDIAL PORTION

The medial portion of the musculospiral constitutes that part of the nerve trunk which lies medial to the humerus, extending from its origin in the axilla to where it crosses the long head of the triceps. In the upper part of its medial course the musculospiral nerve lies posterior to the axillary artery; in the lower part of the axilla it leaves the neurovascular bundle and, passing over the tendons of the teres major and latissimus dorsi, it begins its dorsal or intratricipital course by passing anterior to the long head of the triceps. From the medial portion of the musculospiral three branches are given off from the medial side of the nerve trunk-an internal cutaneous branch and two motor branches, one of which supplies the long head of the triceps, the other, known as the ulnar collateral, supplies the medial head of the triceps. These branches may be followed some distance up the nerve trunk, where they will be found to unite with two distinct bundles. The internal cutaneous branch unites with the bundle which in the dorsal portion of the nerve's course forms the external cutaneous. In other words, the sensory fibers contained in the internal and external cutaneous branches spring from a common bundle and separate from the parent trunk either combined or as separate branches. The identification of one sensory branch, if followed intraneurally, will serve to identify the other by their intraneural union. They may be identified distally by a careful dissection which will demonstrate their cutaneous termination. The two motor branches may be given off from the musculospiral trunk individually or, as occasionally happens, all the motor branches to the triceps will be found leaving the nerve trunk as a single large branch, later subdividing and forming the individual branches to its respective heads. More often the tricipital branches are loosely incorporated within the


997

musculospiral sheath, though they are distinctly individual and may possess a well-developed sheath of their own. The bundle containing the tricipital branches occupies a very definite position upon the medial side of the nerve, ventral to the cutaneous sensory bundle. The larger or lateral bundle constitutes that portion of the nerve which gives off branches to the supinator and extensors, ultimately terminating in the radial and posterior interosseous nerves. Lesions of the musculospiral nerve in its medial portion are often

FIG. 168.- Exposure of medial portion of musculospiral nerve in the axilla and upper portion of the arm. Tendon of pectoralis major divided. Median and ulnar nerves with brachial artery, retracted medially, exposing musculospiral nerve and superior profunda artery, as they cross the tendon of the latissimus dorsi and long head of the triceps entering their posterior humeral course. The branches of the medial portion of the musculospiral are shown springing from the medial aspect of the musculospiral trunk. A, Median nerve, outer and inner head; B, musculospiral nerve; C, axillary artery; D, pectoralis major, reflected; E, musculospiral medial branches; F, superior profunda artery; G, brachial artery; H, latissimus dorsi tendon; I, coracobrachialis; J, deltoid; K, Pectoralis minor tendon

combined lesions, corresponding to terminal plexus types, and frequently are associated with vascular injury.

The method of exposure of the musculospiral nerve in its upper medial portion within the axilla is the same as that used in infraclavicular plexus lesions. Its lower medial portion is exposed through a straight incision, extending from the base of the axilla to the middle of the humerus, in a line corresponding to the medial edge of the coracobrachialis. The deep fascia is divided, exposing the neurovascular bundle; by ligating a few branches of the brachial artery and their accompanying veins, the contents of the neurovascular


998

bundle may be retracted medially, and the musculospiral nerve located as it crosses the latissimus dorsi tendon, accompanied by the superior profunda artery; or it may be located before it leaves the neurovascular bundle, where it lies behind the brachial artery. Its cutaneous and motor branches should now be identified as they leave the nerve trunk from its medial surface. The motor branches are identified readily by following them to a muscle termination, while the internal cutaneous branch will be found to penetrate the deep

FIG. 169.-Exposure of median portion and internal part of posterior portion of musculespiral trunk through medial incision. Contents of neurovascular bundle retracted medially; superior profunda artery ligated; complete exposure of tricipital branches; intraneural dissection, showing their origin from medial side of the musculospiral trunk. A, Deltoid; B, biceps; C, coracobrachialis; D, superior profunda artery, ligated; E, pectoralis minor tendon: F, pectoralis major, reflected; G, musculospiral nerve; H, median nerve; I, internal cutaneous branch: J, latissimnus dorsi tendon; K, tricipital branches

fascia to assume a cutaneous position. Retraction of the long head of the triceps will permit very satisfactory exposure of the nerve on the medial side of its intratricipital course, where it will be found passing lateral to the medial head of the triceps as it enters the musculospiral groove on the posterior surface of the humerus. Division of a few of the internal fibers of the medial head ofthe triceps will permit adequate exposure of the nerve through most of its posterior course. This medial incision is also frequently required to gain access to the nerve in its dorsal position.


999

DORSAL PORTION

The dorsal or intratricipital portion of the musculospiral is that part of the nerve which lies posterior to the humerus and is covered by the triceps. The nerve enters the posterior surface of the arm by passing anterior to the long head of the triceps shortly after it has left the neurovascular bundle. In its dorsal course it first lies on the ventral surface of the long head of the triceps; then occupying a position in direct contact with the humerus it passes through the musculospiral groove, just above the origin of the medial head of the triceps and below the origin of the lateral head of the triceps, which latter covers the nerve. Having pursued a somewhat spiral course around the humeral shaft, the nerve emerges from its dorsal position by penetrating the external intermuscular septum to appear on the lateral aspect of the arm. In its dorsal course, the nerve is accompanied by the superior profunda artery and its accompanying vein, which lie lateral to the nerve. The superior profunda artery, upon reaching the external intermuscular septum, divides into two branches, the smaller of which penetrates the intermuscular septum and accompanies the musculospiral nerve; the larger branch follows along the posterior surface of the intermuscular septum to the elbow. In its dorsal portion the musculospiral nerve gives off branches to the lateral and medial heads of the triceps. These branches almost invariably lie on the medial side of the nerve, and may be followed up the nerve trunk as individual branches or as a single bundle into the axilla. The external cutaneous nerve, though given off with the internal cutaneous from the medial side of the musculospiral nerve, crosses the parent trunk and follows its lateral surface with the profunda artery where it divides into a superior and an inferior cutaneous branch. These cutaneous branches do not penetrate the external intermuscular septum, though the inferior branch passes over it and becomes cutaneous near the elbow.

The nerve in this region, occupying a position in direct contact with the humerus, is frequently injured in fractures of the middle third. It may be completely crushed by the trauma producing the fracture, though probably it is more often traumatized and stretched by the bone fragments at the time of injury or during efforts at fracture reduction. Occasionally it is found compressed or completely buried in the callus of an old fracture.

Exposure of the musculospiral nerve in its dorsal position may be accomplished by several methods:

A. Dorsal longitudinal incision.-An incision beginning about 7 centimeters below the acromial process is carried down the middle of the posterior surface of the arm to the junction of its lower and middle thirds, in line with the olecranon. The deep fascia is incised; and the cleavage between the long and lateral heads of the triceps found and deepened until the aponeurosis of their ventral surface is encountered. This aponeurosis is then carefully divided to prevent injury to the musculospiral nerve and its branches, which lie directly on its ventral surface. In separating the long and lateral headsof the triceps upward, the posterior fibers of the deltoid are retracted laterally, some of which may require division. Retraction of the tricipital heads will give a satisfactory though rather limited exposure of the musculospiral groove


1000

and its contents. When extensive scar tissue Is present, the dorsal incision should be supplemented by a lateral incision, exposing the nerve as it emerges on the lateral surface of the arm, after penetrating the external intermuscular septum. The nerve is identified at this point and the lateral head of the triceps separated from its attachment to the septum and retracted dorsally exposing the lateral portion of the musculospiral groove.  Occasionally it will be found necessary to divide the lateral head of the triceps and in such instances the division should he made as high as possible to avoid injury to its nerve supply. A piece of tape passed under the lateral head of the triceps between these two incisions will often permit sufficient retraction of the muscle from the humerus to expose the nerve throughout its course in the musculospiral groove. In attempting retraction or division of the lateral head of the triceps through the combined dorsol lateral incisions, the greatest care should be used to prevent injury to the tricipital branches, which at first should be identified and isolated in the upper portion of their course. The operator frequently experiences difficulty in this dissection, due to injury of the superior profunda artery and veins. It is expedient, therefore, early in the exposure, to secure primary ligation of the artery in the upper part of the wound and ligation of the veins dorsal to the external intermuscular septum.            
               
B. Exposure of the dorsal portion of the musculospiral by a combined medial lateral incision.-The medial portion of the musculospiral groove may be readily approached through the low incision described for exposure of the

FIG. 170.- Showing course of musculospiral nerve, and relation of branches to triceps, as it passes behind humerus in musculospiral groove. Preservation of branches in transposing nerve to anterior position. (Ney, Annals of Surgery, 1921)

musculospiral in the lower part of its medial portion, by retracting medially the neurovascular bundle groove may be readily approached through the low incision described for exposure of the musculospiral in the lower part of the medial portion, by retracting medially the neurovascular bundle, and dorsally the long head of the triceps. This approach will also facilitate control of the profunda artery and the identification and isolation of the tricipital branches. This medial exposure is supplemented with the lateral incision, through which the musculospiral nerve is identified as it penetrates the external intermuscular septum, and the attachment of the lateral tricipital fibers freed from this structure. In this procedure the long and lateral heads of the triceps are retracted from the humerus, exposing the nerve throughout its intratricipital course. The advantage of this method is obvious; it gives a complete exposure of the nerve and all its tricipital


1001

branches it permits of satisfactory heinostasis. in that the profunda artery may be secured above and the veins below; and in the presence of extensive continuity defects, it permits transposition of the nerve to an antebrachial position by allowing mobilization of the nerve trunk, without endangering its branches and without additional incisions.

The proximity of the musculospiral nerve to the humerus in its, dorsal course frequently calls for the preparation of a new nerve bed. When this is necessitated, because of extensive callus or sear tissue, the nerve may be separated from direct contact with the bone by changing its course from the musculospiral groove to a lower position where it is made to pass behind the medial head of the triceps. This procedure requires mobilization of the nerve through its lateral portion, and low separation of the lateral head of the triceps from the external intermuscular septum, permitting the nerve to follow a lower dorsal course between the medial and lateral heads of the triceps. When extensive bone callus or scar tissue involves the posterior surface of the humerus and the tricipital heads, recourse should be had to an antebrachial transposition, in which the nerve is directed anterior to the humerus, between the biceps and the brachialis anticus.

VENTROLATERAL PORTION

The ventrolateral portion of the mnusculospiral nerve constitutes that portion of the nerve which lies lateral and ventral to the humerus on the external aspect of the arm, beginning at the external intermuscular septum and terminating in front of the external humeral condyle, where it bifurcates into its two terminal branches. In its ventrolateral course, in passing down the lower third of the arm, the nerve is deeply placed between the brachioradialis and brachialis anticus. In the region of the external humeral condyle, the nerve is covered by the braclioradialis and rests upon the brachialis anticus. This portion of the nerve gives off a branch to the brachioradialis shortly after entering the lateral surface of the arm. A little lower, a branch is given off to the extensor carpi radialis longior. These motor branches emerge from the lateral side of the musculospiral trunk, the branch to the brachioradialis being the most ventral of the lateral group of motor fibers. Immediately behind this branch lies the bundle containing the fibers to the carpal extensors. The branch to the extensor carpi radialis brevior is occasionally given off with the branch to the long carpal extensor, though more frequently it leaves the nerve trunk from its terminal posterior interosseous division. It has, however, a long intraneural course, and if followed up the nerve will be found to spring from a bundle common to both radial extensors. Injuries to the musculospiral in this region are probably more often encountered in military surgery than injuries to any other nerve. It is a common complication of fractures of the lower third of the humerus, and in civil life the frequency of its occurrence is approximated only by lesions of the ulnar in the region of the internal epicondyle. In simple fractures the trauma to the nerve consists usually in stretching, or bruising against bone fragments; in the absence of an open wound it is rarely divided. Paralysis of the musculospiral


1002

nerve bv pressure has frequently been observed following the use of a tourniquet, and surgeons who insist, upon its use should remember this possibility and endeavor to protect the musculospiral region by properly placed pads.
           
Exposure of the lateroventral portion of the musculospiral
is made through a longitudinal incision, extending from the insertion of the deltoid to a point midway between the biceps tendon and the external condyle of the humerus. In the lower part of the incision the cephalic vein is divided between forceps. In the upper part of the incision, the lateral cutaneous branches may be met, and,

FIG. 171.- Landmarks for exposure of musculospiral nerve in its latero-ventral aspect, showing external intermuscular septum, from the ventral side of which arises fibers of the brachioradialis, and from its dorsal surface, fibers of the lateral head of triceps. Musculospiral nerve is exposed by deepening the cleft between brachioradialis and biceps which lies ventral to the latter muscle. A. Biceps; B, brachioradialis; C, triceps, lateral head; D, external intermuscular septum

FIG. 172 .– Musculospiral nerve, latero-ventral aspect, showing emergence on lateral surface of arm, after penetraging external intermuscular septum. A, Biceps; B. Brachialis anticus; C, musculospiral nerve; D, superior profunda artery E, triceps, lateral head; F. External intermuscular septum; G, brachioradialis

if necessary, sacrificed with impunity. The line of cleavage between the brachioradialis laterally and the biceps and brachialis anticus medially is identified and these muscles retracted. In the cleft, deeply placed, the musculospiral nerve is found accompanied by a small terminal branch of the superior profunda artery. Occasionally, the inferior external cutaneous branch, which


1003

is of fairly large size, will be found passing along the external intermuscular septum and should not be mistaken for the inusculospiral trunk. It will often serve as a guide, if followed upward, in locating the musculospiral. f

The musculospiral trunk should be isolated above the lesion to facilitate branch identification and preservation, even though it be necessary to invade its dorsal or intratricipital portion. In following the nerve downward, between

FIG. 173.- Musculospiral nerve, postero-lateral aspect; lateral head of triceps divided close to attachment to external intermuscular septum, exposing nerve in lateral aspect of its dorsal course in musculospiral groove. A, Musculospiral nerve; B, triceps, lateral head, divided; C, biceps; D, brachialis anticus; E, external intermuscular septum; F, brachioradialis.

FIG 174.- Musculospiral nerve at elbow, showing terminal posterior interosseous and radial divisions. Brachioradialis divided. Intraneural dissection of branches shows them springing from lateral side of musculospiral trunk. Radial nerve is given off from ventro-medial aspect of musculospiral trunk; posterior interosseous nerve arises from dorso-lateral portion of trunk. A, Musculospiral nerve; B, posterior interosseous nerve; C, radial nerve; D, brachioradialis, divided; E, brachialis anticus; F, extensor carpi radialis longior; G, branch to extensor carpi radialis brevior

the brachioradialis and the brachialis anticus, extreme care should be used in preserving motor branches. In the upper part of this region, if the circumference of the musculospiral nerve be divided into a lateral, posterior, and medial sector, the lateral sector will be found to contain motor fibers to the brachioradialis and carpal extensors, the posterior sector will contain those I

f The writer has on one occasion reoperated a case of musculospiral paralysis which failed to regenerate, and found the proximal end of the musculospiral anastomosed to the distal segment of this branch; in the original operation, the surgeon had apparently mistaken the cutaneous branch for the musculospiral trunk.


1004

motor fibers which eventually forii the posterior interosseous nerve, and the medial sector will contain sensory fibers ultimately forming the radial terminal division.

In the region of the external condyle, the musculospiral nerve lies upon the brachialis anticus and under cover of the brachioradialis. Strong lateral retraction of the latter muscle will expose the nerve where its two terminal branches may be identified-the posterior interosseous, arising from its lateral and posterior sector; and the radial, arising from its medial sector.

INTEROSSEOUS NERVES

The posterior interosseous nerve, containing mostly motor fibers, innervates the supinator, extensor carpi ulnaris, and extensors to the fingers and thumb, and it constitutes the terminal motor portion of the musculospiral trunk. After its separation from the radial nerve on the undersurface of the brachioradialis, it passes under the long and short carpal extensors and swings around the shaft of the radius to the dorsum of the forearm, by passing through a canal in the substance of the supinator brevis muscle, in a line almost at right angles to the direction of the supinator fibers. After emerging from the lower border of the supinator on the dorsal surface of the forearm, the posterior interosseous nerve fans out into a leash of branches supplying the extensors of the fingers and thumb.            
           
Lesions of the posterior interosseous nerve below the upper third of the forearm are rarely amenable to suture, due to the breaking up of the nerve into a series of small branches. Suture may be effected, as a rule, only when the nerve is found injured in the substance of the supinator or proximal to that muscle. Occasionally, however, it will be found possible to complete a satisfactory neurolysis in the region where the nerve breaks up below the supinator. When it is found impossible to restore the nerve supply to the digital extensors, it is advisable to complete the operation by tendon transplantation.           
           
Exposure of the posterior interosseous nerve
.- A longitudinal incision is made down the posterior surface of the forearm in the midline, beginning at a point two centimeters medial to the external condyle and ending in the middle third of the forearm. The deep fascia is divided and the line of cleavage between the extensor carpi radialis brevior and the extensor communis digitorum identified. These muscles are separated up to their common origin at the external condyle and retracted, exposing the fascia covering the supinator brevis muscle below. The course of the posterior interosseous nerve through the supinator is at almost right angles to its fibers, and its position may often be identified by palpation. Failing in this, separation of the muscle fibers in their line of cleavage will usually serve to locate the nerve, which is greatly flattened as it passes around the shaft of the radius. Exposure of the lower border of the supinator will also aid in its identification, as it emerges from this muscle.

The posterior interosseous nerve may be completely exposed by dividing the superficial fibers of the supinator covering the nerve; by flexing the forearm and retracting the radial carpal extensors and the brachioradialis, it may be followed to its junction with the radial. Extreme care, however.


1005

should be used in avoiding the branches supplying the supinator, which are usually given off just before the nerve enters this muscle or during its passage through the muscle. Lesions within the supinator region of the posterior interosseous as a rule mav be satisfactorily exposed through the dorsal forearm incision; but when the lesion is located in its presupinator portion, it is best exposed by separating the long radial extensor from the brachioradialis

FIG. 175.- A, Supinator brevis exposed by separating extensor carpi radialis brevior and extensor longus digitorum. B, Intrasupinator portion of posterior interosseous nerve exposed by dividing superficial fibers of supinator brevis. Branches to supinator shown leaving the nerve in the substance of the muscle. At lower border of supinator, posterior interosseous fans out into a leash of branches, below which repair is dlifficult. A. Extensor carpi radialis brevior; B. supinator brevis; C, extensor longus dligitorurn; D, posterior interosseous nerve: E. branches to supinator; F, posterior interosseous, below supinator

through a slightly more anterior incision, corresponding to a continuation of the lateral musculospiral incision to the dorsum of the forearm. In exposing that portion of the nerve lying between its origin and the supinator (presupinator portion), great care must be utilized to prevent injuring the branch to the short carpal extensor, which is given off from the anterior surface of the nerve. It is usually of sufficient size, to permit of separate suture. and efforts, should always be made toward its conservation or repair.


1006

RADIAL NERVE

The radial nerve constitutes the terininal sensory division of the musculospiral trunk; it first lies under the brachioradialis and upon the brachialis anticus, medial to the posterior interosseous. It progresses down the ventral and radial surface of the forearm under cover of the brachioradialis muscle. At the junction of the upper and middle third of the forearm it is joined by the radial artery, which lies on its medial side and accompanies it through the middle third of the forearm. At the beginning of the lower third it leaves the artery and passes under the tendon of the brachioradialis to enter the posterior surface of the forearm, where it supplies sensation to the dorsum of the wrist, hand, and fingers.

SURGERY

The radial nerve, being entirely sensory, is seldom exposed for repair. It is, however, occasionally used as a graft for filling continuity defects in other nerves, because of its fairly large size and the minimal degree of disability resulting from its sacrifice; when used for grafts, its uppermost portion is usually selected.
           
Exposure of the radial nerve
may be accomplished through an incision beginning just above the bend of the elbow, midway between the external humeral condyle and the biceps tendon, and carried longitudinally down the radial surface of the forearm as far as necessary, paralleling the medial border of the brachioradialis. The deep fascia is divided an(l the medial edge of the brachioradialis exposed and retracted. The radial nerve may be identified passing along the under surface of this muscle, and may be followed up to a point where it joins the posterior interosseous to form the musculospiral trunk.

CONTINUITY DEFECTS

The correction of defects in the continuity of the musculospiral nerve differs in no way from those methods used in overcoming defects in other nerves except in transposition, when the musculospiral nerve is changed from its spiral course around the posterior aspect of the humerus to the more direct ventral course anterior to the humerus, where it is made to pass between the biceps and the brachialis anticus, following a course similar to that taken by the musculocutaneous nerve.

TECHNIQUE OF MUSCULOSPIRAL TRANSPOSITION

The musculospiral nerve is exposed in its medial portion (see technique of low exposure of the medial portion of the musculospiral nerve, p. 997); the branches to the triceps are identified by retraction of the triceps and mobilized from the musculospiral trunk.

The musculospiral is exposed and freely mobilized from the external intermuscular septum to its terminal branches, using care to prevent injury to branches in this region. The lateral head of the triceps is separated from its attachment to the external intermuscular septum and carefully separated from the humerus along the course of the musculospiral groove. A piece of tape is passed through the tunnel between the lateral and medial incisions, and the


1007

triceps by this means retracted from the bone, completely exposing the nerve. It is important that the ends of the nerve be marked with identification sutures before the nerve trunk has been rotated during mobilization.

A second tunnel is made connecting the lateroventral incision with the medial, by identifying the medial and lateral borders of the biceps and by blunt dissection, separating this muscle from the underlying brachialis anticus, following the line of cleavage between these muscles and using care to prevent injury to the musculocutaneous nerve.

The musculospiral nerve is now drawn from its intratricipital course and made to pass in as direct a line as possible from the neurovascular bundle below the axilla to its position medial to the external humeral con(ldle below the biceps. After transposition, the surgeon may avail himself of flexion-relaxation to assist in overcoming any part of a defect not corrected by the transposition. It is particularly important to prevent torsion of the nerve trunk during suture, and the early placing of accurate identification sutures is essential as a guide for restoring physiologic alignment.

PHYSIOL.OGIC APPROXIMATION OF THE MUSCULOSPIRAL AND ITS TERMINAL DIVISIONS.

The musculospiral nerve, by virtue of its motor branches, lends itself to physiologic approximation more readily than any other nerve. While the prevention of torsion is perhaps best accomplished through effectively placed identification sutures, there are certain instances, particularly in secondary sutures, where this means of identification can not be relied upon.

In effecting suture of the medial portion of the musculospiral nerve, identification of tricipital branches will immediately demonstrate the medial side of the nerve. In its posterior position, the tricipital branches likewise indicate the medial side of the nerve, though the lower branches to the triceps occupy a more dorsal position. In the region of the external intermuscular septum, the nerve occupies a rather fixed position, which is not readily disturbed by trauma: in this position, the motor fibers lie in the lateral portion of the trunk, the sensory fibers to the medial side. The nerve in its course around the humerus, first contains its motor fibers in its medial side. Later, in the musculospiral groove, they become posterior, the sensory fibers being in contact with the humerus. In the region of the external intermuscular septum the nerve presents a different arrangement, due to a slight natural torsion in its spiral course. In its lateroventral position, the motor bundles lie on the lateral and dorrsal surface of the nerve, whence springs its motor branches and which ultimately terminate in the posterior interosseous nerve. The sensory portion of the trunk, having a more medial and anterior position, terminates in the radial nerve. By keeping these facts in mind. torsion may be prevented to a great degree in lateroventral sutures of the musculospiral; the motor portion of the trunk being readily identified by the position of its motor branches, also by the posterior interosseous nerve, springing from its lateral and somewhat dorsal quadrant; the sensory portion of the nerve being identified by the absence of branches and by the position of the radial nerve. To make topographical matching still more accurate, it should be remembered that the branch to the brahiolradialis occupies the most ventral motor position. More


1008

laterally in the nerve trunk will be found the bundle sending fibers to the long and short radial extensors. That portion of the nerve immediately behind these funiculi ultimately forms the posterior interosseous trunk.

Torsion in the posterior interosseous nerve is not so serious an eventuality as in mixed nerves, because most of the nerve trunk is composed of motor fibers. In the supinator region, the nerve is greatly flattened and its fixation so secure that reasonable care should prevent torsion. If doubt exists in suturing the musculospiral nerve as to the location of its sensory segment, recourse should be had to the electro-anatomic method, in which the sheath of the upper segment of the nerve is opened and the bundles subjected to very weak faradic stimulation. Stimulation of the sensory bundles will elicit a sensation of tingling or pain localized in their cutaneous distribution, while stimulation of a motor bundle will demonstrate the presence of myosensory fibers. (See general technique, p. 951.)

SECONDARY SUTURE OF THE MUSCULOSPIRAL NERVE.

Surgery of the musculospiral nerve has perhaps given more satisfactory end results than that of any other nerve, which may be attributed to the following facts:

1. The greater number of injuries occur in its lateroventral portion or in the lateral part of its dorsal portion, which leaves a comparatively short distance for the neuraxons to regenerate to reach their ultimate muscle termination. This feature, facilitating an earlier restoration of neuromuscular junctions saves the muscles from that extreme degree of degeneration encountered when denervated for a long period of time--a condition which can not be avoided in nerves injured at a great distance from the muscles they supply.
2. The extensor muscles, supplied by the musculospiral are relatively large in bulk and the chance of receiving a greater number of regenerated nerve fibers is thereby enhanced.
3. The extensor function of the wrist is accomplished by three extensor muscles, supplemented by the common extensors of the fingers, so that defects in one or more muscles may not seriously affect extensor function.
4. Extensor paralysis of the wrist is more generally and earlier recognized, and lends itself more readily to splinting than any other types of paralysis: it is thereby often saved from prolonged muscle stretching.

A careful examination of regenerating musculospiral lesions will demonstrate that, though the functional end results of musculospiral sutures are more satisfactory than any other nerve, from a standpoint of individual muscle function, defective regeneration is present in the great majority of musculospiral sutures, and that those muscles which have a relatively small bulk and which are innervated at some distance from the point of lesion, present the same degree of defective regeneration as observed in other nerves. After musculospiral suture, radial carpal extension and brachioradialis action are commonly restored. Less frequently, however, do we observe a return of function in the ulnar carpal extensor, and in the extensor communis digitorum, while in the extensors of the thumb, index, and little fingers function is frequently not regained. Undoubtedly, they enter into that extreme degree of muscle degeneration common to muscles of small bulk, long denervated.


1009

INDICATIONS FOR REOPERATION

Before the musculospiral nerve is subjected to secondary surgical intervention, certain facts regarding the failure of regeneration should be definitely determined:

1. The absence of neuraxon regeneration beyond the point of suture, or defective or retarded regeneration through the distal segment.-Tinel's sign, elicited with the same degree of intensity below the suture line, as at the suture line, indicates successful neuraxon regeneration. The downward progression at the rate of 1 inch per month, of the point at which formication can be elicited indicates that regeneration is not retarded, at least in so far as sensory fibers are concerned. A complete absence or at greatly diminished reaction of formication below the suture line indicates absence or defective neuraxon regeneration. After three months, a failure to elicit formication below the suture line suggests the absence of regeneration, particularly if this reaction is intense at the suture line, though it mav mean only a delay in regeneration. A decision. therefore, should not be definitely made until a lapse of six months, at which time absent or greatly diminished formication in the distal segment of the nerve should call for secondary surgical intervention.

2. Torsion of the nerve trunk- resnlting in the physiologic rnisplacement of fibers is another common cause for defective musculospiral regeneration. The writer has reoperated a case in which at the original operation, the operator failed to identify the posterior interosseous terminal division of the nerve and united the musculospiral trunk wholly to its radial terminal division, the lesion being in the region of the internal condyle. In this case, the long carpal extensor lranch was individually divided and not repaired. leaving, a complete musculospiral paralysis below the supinator. Frequently, however, torsion has occurred during the primmmrv suture and the radial sensory fibers have passed down the posterior interosseous nerve and the motor fibers to the radial nerve: obviously, the misdirected fibers are physiologically lost. This misdirection of fibers may occur to any degree and the advisability of secondary operation for its correction depends entirely upon the extent of misplacement and the degree of motor function regained. We must again rely upon Tinel's sign to determine the position of sensory fibers, and the musculospiral nerve in its terminal divisions readily lends itself to convey this information. Percussion of the nerve trunk below the suture line may give an intense reaction of formication, but when percussion of the radial nerve (which normally should convey these sensory fibers) elicits no tingling or only a slight reaction, the sensory fibers are absent and have probably been misdirected, if sufficient time has been allowed for them to reach the radial trunk. Percussion of the posterior interosseous on the dorsum of the forearm in the region of the supinator will normally give little or no tingling, as it contains mostly motor fibers. When there has been any great misdirection of radial fibers through the posterior interosseous nerve, its percussion will elicit intense tingling which is localized on the dorsum of the hand. This localization of tingling is important as it demonstrates that the sensory fibers in the posterior interosseous nerve are definitely radial in origin, it being remembered that in percussing the musculospiral or the posterior interosseous nerve, the examiner may elicit reactions in the external cutaneous


1010

branch of the musculospiral, in which event the tingling sensation would be localized along the dorsal portion of the forearn and not the hand. Frequently, the presence of sensory fibers in the extensor muscles on the back of the fore-arm may be denmonstrated by pressure on their atrophied bellies, this pressure eliciting a strong formication.

Regenerative defects, due to misdirection of fibers are unfortunate, in that the muscles have been allowed to reach an extreme degree of degeneration because of their greatly delayed reinnervation. If muscle degeneration is extreme, recourse must be had to tendon transplantation for correction of the motor defect. If, however, these museles have been preserved and still respond to galvanic stimulation, secondary suture should be attempted for the purpose of correcting torsion. In the secondary operation, the surgeon must not be guided by anatomically placed identificationl sutures, but the motor side of each segment of the nerve trunk should be identified by the position of its branches or by the electro-anatoinic method, and the nerve resutured in a manner which will approximate motor segment to motor segment. (See physiologic approximation of the niusculospiral nerve, p. 1007.)

3. Defective muscolospiral regeneration due to extensive muscular deqeneration.-Percussion of the musculospiral trunk has demonstrated by Tinel's sign the progressive downgrowth of neuraxons: percussion of the radial and posterior interosseous nerves reveals the presence of radial sensory fibers in the former and their absence in the latter; therefore, neuraxon regeneration has probably not been seriously impefed or misdirected. The return of deep sensibility in the muscles suggests regeneration of myosensory fibers which normally accompany motor fibers, and their presence probably indicates the regeneration of motor fibers to the niuscles. The clinical degenerative phenomena continue to persist in the parallyzed muscles (complete loss of electrical and mechanical irritability and a total absence of voluntary motion, with extreme atrophy). Instances of this kind are by no means rare and are usually due to prolonged denervation and ischemia. In this type of regenerative failure, secondary operations upon the nerve are absolutely useless and contraindicated. In some cases, time may effect improvement, particularly if accompanied with energetic treatment to improve circulation and muscle nutrition. In most of these instances, the only hope of restoring extensor function to the wrist and fingers lies in tendon transplantation.

TENDON TRANSPLANTATION FOR MUSCULOSPIRAL PARALYSIS

Probably no other tendon transplants give results so satisfactory as those devised for the drop wrist and fingers in musculospiral paralysis.

In complete irreparable defects of the musculospiral nerve, transplantation is indicated for three conditions, namely, wrist-drop, finger-drop, and thumb-drop. In low lesions (posterior interosseous), carpal extension maybe preserved, transplantation being required only for the digitis. This maybe the case also in defective musculospiral regeneration, though more frequently the disability is confined to the thumb extensors. It is the practice of some surgeons when repairing the musculospiral nerve to transplant the pronator radii teres into the radial extensors, as this procedure will give immediate


1011
 
extension of the Wrist without Sacrificing the function of any muscle or interfering, with the natural progress of regeneration. We shall therefore consider individually the transplant necessary for the correction of each of these various conditions, though one or all may be corrected during the course of a single operation.

FOR RESTORATION OF CARPAL EXTENSION

a. Pronator radii teres into the extensors carpi radialis longior and brevior to give dorsal flexion of the wrist.
b. Divided extensor carpi ulnaris tendon into the flexor carpi ulnaris prevent radial deviation of the hand.

TECHNIQUE.

1. The incision extends along the radial border of the forearm in its middle third. The line of cleavage is found between the brachioradialis and the carpal extensors and deepened to expose the radius, where the ribbon-like tendon of the pronator teres is found partially encircling this bone; this tendon is mobilized and divided at its insertion. The tendons of the extensors carpi radialis longior and brevior are now identified, and with thev wrist held in hyperextension. the pronator tendon is passed through a slit in these tendons and anchored after proper tension has been attained.
2. To prevent radial deviation of the wrist during extension through pronator teres action on the radiocarpal extensors, the ulnar carpal extensor tendon manv be divided land transplanted into the flexor carpi ulnaris. Occasionally also in defective musculospiral regeneration, the extensor carpi ulnaris fails to regain power, and extension of the wrist is associated with marked radial deviation. This transplant is usually very satisfactory for the correction of this condition. Frequently the tendon of the flexor carpi ulnaris is used as a transplant to produice extension of the third, fourth, and fifth. fingers, and occasionally also the index; in which case the fleshy part of thle flexor carpi ulnaris tendon. which is attached to the ulnar border of this tendon. is detached and separately transplanted into the extensor carpi ulnaris tendon, or the extensor carpi ulnaris tendon may be divided some distance above its insertion and transplanted into the detached fleshv fibers of the flexor carpi ulnaris.

FOR RESTORATION OF EXTENSION OF THE DIGITS

a. Flexor carpi radialis into the extensor ossis metacarpi pollicis, extensor pollicis ,brevis. aid ext ensor pollicis longus. and occasionally into the extensors indices and comlnmunis tendons of thle index finger.
b. Palmtiaris longus into the tendon of the long extensor of the thumb, when the palmaris is present (about 20) per cent of cases) and when the flexor carpi radialis is not used for terminal flexion of the thumb.
c. Flexor carpi ulnaris into the extensor tendons of the little, ring, and middle fingers, and also the index finger when its tendon is not anastomosed to the flexor carpi radialis.

Inasmuch as isolated extension of the index finger is important, with extension of the thumb, in picking up objects, the writer prefers to use the flexor carpi radialis for both thumb and index extension. Some of our most


1012

satisfactory results in extension of the distal phalanx of the thumb have attended flexor carpi radialis transplantation into the long extensor in common with the other extensors of the thumb, though a separate transplantation of the palmaris longius into this tendon is a verv satisfactory procedure.

TECHNIQUE

a. Through a long incision, extendling from the insertion of the flexor carpi radialis tendon to the middle of the forearm on its radio-ventral aspect, the tendon of the flexor carpi radialis is exposed, divided at its insertion and freely mobilized to the upper part of the incision, care being taken to avoid injury to the radial artery and nerve lying along its radial border.
b. A second ventral incision of the same length is made alone the ulnar border of the forearm, exposing the ulnar tendon, which is divided at its attachment and freely mobilized. The tendon of the flexor carpi ulnaris is split close to its ulnar border, to which muscle fillers are attached almost to its insertion. The splitting of the ulnar flexor tendon in this manner preserves the insertion of its fleshy fibers to which portion the divided extensor carpi ulnaris tendon may be anastomosed. This separation of the flexor carpi ulnaris tendon is extended to the middle third of the forearm, where it is lost in the fleshy portion of its belly. The radial border of the muscle should. however, be mobilized some distance higher. care being taken to avoid injury to the ulnar vessels and nerve which lie below.
c. A mid-dorsal incision is now carried from the wrist joint to the middle third of the forearm, and the fat of the superficial fascia medially and laterally undermined around both radial and ulnar borders of the forearm, connecting subcutaneously, on the radial side, with the radioventral incision, and, on the ulnar border, with the ventroulnar incision. The direction of each tunnel should be arranged in a manner permitting a straight pull from the upper ventral surface of the forearm to the dorsal surface of the wrist, by the transposed extensor tendons. The mobilized flexor carpi ulnaris tendon is now passed through the tunnel to the dorsal aspect of the wrist and passed through as lit in both extensor tendons of the little finger and the common extensor tendons of the ring and middle fingers, while the fingers and wrist are held in complete extension, which extension must be maintained throughout the operation and during subsequent treatment. The undermining of the superficial fascia along the radial border of the wrist will expose the tendons of the extensor ossis metacarpi pollicis and the short extensor of the thumb. Through the dorsal incision the long extensor of the thumb and both extensors of the index finger are identified. The tendon of the flexor carpi radialis is passed through a slit in these tendons, while the thumb is in complete extension and abduction, and the index finger fully extended. When the proper tension has been obtained, the transplant is fixed to each tendon by linen sutures. When it is desired to use the palmaris longus for terminal extension of the thumb, the writer prefers mobilizing and dividing the long extensor of the thumb some distance above the wrist and transposing it to the ventral surface of the wrist for anastomosis into the divided tendon of the palmaris longus; each tendon should be sufficiently mobilized to permit a pull as nearly direct as possible from origin to insertion.


1013

After thorough hemostasis is effected, all incisions are closed, the ventral prior to the dorsal, after which the wrist, thumb, and fingers are maintained in extension by a previously made long "cock-up" splint; care should be used in preventing any hyperextension of the metacarpophalangeal joints. It is often desirable to permit slight flexion of these joints, which greatly adds to the comfort of the patient. (Subsequent treatment and reeducation follow the usual rules for tendon transplantation.)

TO SUPPLY EXTENSOR ACTION TO THE THUMB

a. Flexor carpi radialis into the extensor ossis metacarpi pollicis, extensor brevis pollicis, and extensor longus pollicis.
b. Palmaris longus into the extensor longus pollicis, when extension of the distal phalanx of the thumb alone is absent, or when it is deemed advisable to use a separate tendon for extending the distal phalanx. The palmaris longus tendon does not appear to be a suitable transplant to use for all the extensors of the thumb.

TECHNIQUE.

The procedure for transplanting the flexor carpi radialis into the extensors of the thumb or supplementing this with the palmaris longus to the long extensor of the thumb varies in no way from the method described above, except that the mid-dorsal incision may now be shortened and made nearer the radial border of the wrist.

MEDIAN NERVE

GENERAL ANATOMY

The median nerve is formed in the axilla by a union of the anterior divisions of the outer and inner cords of the brachial plexus. The outer cord formed by the fifth, sixth, and seventh cervical nerves contributes the outer head of the median, while the inner cord formed by the eighth cervical all first dorsal supplies the inner head of the median; the union of the two heads takes place usually just below the tendon of the pectoralis minor muscle. The median nerve then passes down the medial side of the arm in the neurovascular bundle in close relation to the axillary and brachial artery. In the axilla the artery will usually be found lying between the forked head of the nerve. being crossed lby the inner head before the median trunk is formed. In the lower part of the axilla and upper part of the arm the artery lies immediately behind the nerve. As it progresses down the arm the artery first lies medial to the nerve, but toward the middle of the arm the relation is reversed, the artery crossing to the lateral side. In the upper part of the arm the median nerve is overlapped by the coracobrachialis muscle and the medial border of the biceps. In the lower half of the arm the nerve lies upon the brachillis antictus. becoming superficial as it assumes a more anterior position entering the antecubital fossa. In the antecubital fossa the nerve passes under the bicipital fascia and begins its course down the forearm; passing beneath the, superficial or humeral head of the pronator radii teres, it enters its deep position in the forearm by passing under the tendinous arch of the flexor sublimis


1014

digitorum. Crossing the ulnar artery it follows the under surface of the sublimis digitorum, lying upon the flexor proftndus. As it approaches the wrist the nerve again becomes superficial, lying in this region under the tendon of the palmaris longus medial to the flexor carpi radialis and lateral to the sublimis tendons. It passes below the anterior annular ligament; and in the palm, lying under the palmar fascia, it remains lateral to the sublimis tendons, where it divides into its terminal digital branches.

BRANCHES

The only branch given off by the median nerve in the arm is in the lower portion of the lower third, where a small branch supplies the elbow joint, though, occasionally, the branches to the pronator radii teres arise in this region. There are, however, many instances in which the median in the arm communicates in various degrees with the musculocutaneous. (See musculocutaneous nerve, p. 993.)

As the median nerve enters the antecubital fossa it gives off branches, usually three in number, to the pronator radii teres. Occasionally the three branches may leave the main median trunk in a single sheath, dividing after leaving the parent nerve. The next branches given off are to the palmaris longus, flexor carpi radialis, and flexor sublimis digitorum, in the order named. Occasionally, however, at this point a single large motor branch is given off from the median, supplying all the median innervated extrinsic muscles. In such cases the branches to the palmaris longus, flexor carpi radialis, and flexor sublimis digitorum first leave this motor portion. The remainder of this trunk, carrying branches to the deep muscles, is commonly known as the anterior or volar interosseous nerve and is accompanied in its course down the arm by the anterior interosseous artery. This branch occupies a position somewhat deeper than the main trunk of the median nerve, lying in the cleft between the flexor profundus digitorum and the flexor longus pollicis, giving branches to these muscles in its earlier course and terminating in the wrist by supplying the pronator quadratus muscle and sending filaments to the radioulnar articulation and the wrist joint. The motor branches to the extrinsic muscles. however, may leave the parent trunk with many variations, superficially arising from both sides of the median trunk. When, however, these branches are followed a short distance up the parent trunk it will be found that they arise from those funiculi which occupy the medial half of the nerve, the most ventral fibers supplying the pronator radii teres. Medial and slightly posterior to the pronator funiculi lie the fibers to the palmaris longus, flexor carpi radialis. and flexor sublimis digitorum, while the funiculi to the flexor profundus digitorum and flexor longus pollicis encroach upon the posterior surface of the nerve. All these motor branches, however, are located on the medial side of the nerve and may be followed up the nerve trunk by an intraneural dissection for some distance, occasionally as high as the middle of the arm. Often the branches supplying the deep muscles (volar interosseous nerve) leave the median trunk as an individual lower lranch. Just above the wrist the median nerve gives off a sensory palmar cutaneous biranch, which passes above the annular ligament


1015

After passing into the palin the median gives off motor branches from its radial side to supply the opponens, abductor, and superficial head of the flexor brevis pollicis. In the region of the superficial palinar arch the nerve becomes enlarged and flattened, dividing into two branches, which in turn subdivide into five terminal cutaneous sensory branches. These branches. shortly after their origin andl occasionally before, give off motor branches to the first and second lumbricales muscles.

SURGERY

IN THE ARM
           
Axillary portion
.- Lesions involving the median nerve in the axilla are seldom isolated lesions; they are usually combined with injuries to contiguous elements in the neurovascular bundle, so that from a surgical stand point lesions of the median nerve in the axilla are treated as lower plexus lesions. (See lower plexus lesions, p. 984.)

In the arm.- Lesions of the me(lian nerve in the arm are seldom confined to this nerve. The close proximity of the ulnar trunk, the brachial artery. the basilic vein, and the internal cut aneous nerve makes combined lesions more frequent in the upper two-thirds of the arm, while in the lower third the ulnar and internal cutaneous are frequently spared.

Exposure of the median nerve in the arm is best accomplished with the arm in abduction and outward rotation. The incision follows the medial border of the coracobrachialis above and the biceps below. These structures somewhat overlap the nerve along its liumeral course. In the upper part of the arm the nerve will be found lvina lateral and anterior to the artery: below the mid portion it crosses the artery and lies on the medial side. The vein and artery are so commonlv involved in median injuries in this location that special care should be used in dealing with these structures to prevent undue bleeding, particularly if the lesion is associated with much scar tissue. It is always essential that these vascular structures be isolated above and below the field of scar tissue invasion. With the control of these vascular channels, the surgeon is usually master of the situation and the dissection is greatly facilitated.

In the arm the median nerve gives off no important branches: however, often it communicates by various tvpes of anomalies with the musculocutaneois. explaining certain median lesions in which, at operation. the nerve is found to be completely divided. though the clinical examination reveals incomplete median anesthesia, or the retention of motor power in the median muscles. In making a diagnosis of partial median injury, anomalous communications between the median and musculocutaneous nerves should always be borne in mind. Electrical stimulation of the median nerve in suspected partial lesions will produce contraction in the unparalyzed median muscles, if it be a true partial lesion. If this stimulation fails to elicit contraction in these muscles, stimulation of the musculocutaneous, which is easily accomplished by identifying its trunk on the dorsal surface of the biceps, will indicate the nature of the innervation, if it be due to a median and musculocutaneous anomaly. Retained


1016

function, motor or sensory, when the median trunk shows complete anatomical division, is usually due to median fibers passing down the musculocutaneous trunk and joining the median below the lesion. This refers entirely to the extrinsic muscles; the intrinsic hand muscles, particularly the superficial head of the flexor brevis pollicis may not be paralyzed in complete median paralysis, in which case it is probably wholly innervated by ulnar branches.

When median lesions in the arm are associated with irritative or marked vasomotor phenomena, particular attention should be given to complete isolation and decortication of the brachial artery (periarterial sympathetectomy, after the method of Leriche). In the writer's experience this has occasionally relieved painful nerve syndromes where all other methods have failed. In median causalgia or severe median irritative lesions without anatomic division of the nerve an internal neurolysis is also indicated if the median trunk presents any evidence of scar induration upon palpation.

The prevention of torsion of the median trunk in the arm is best accomplished by the use of accurately placed identification sutures before the nerve has been completely isolated from its bed. The anatomic method of fascicular identification, by relying upon motor branches, can be utilized only in low lesions in the region of the internal condyle, where a branch is given off to the pronator teres. This branch, however, may be followed some distance up the median trunk as a well identified funiculus. Its intraneural localization will reveal the anterior and medial portion of the nerve trunk in this region, it being the most anterior of the motor bundles. The determination of funicular topography above the branches to the pronator radii teres is possible only by the electro-anatomic method, which will serve, if properly conducted, to identify the position of the sensory bundles, which lie in the lateral aspect of the median trunk. When secondary sutures are indicated, because of torsion during the original suture, identification sutures are of no value and resuturing must be guided by localization obtained by the electro-anatomic method of identification.

IN THE ANTECUBITAL FOSSA AND UPPER TWO-THIRDS OF THE FOREAR'M

The position of the arm for exposure of the median nerve in this region is one of abduction and external rotation, with the forearm in complete supination.

The median nerve, after entering the antecubital fossa at the bend of the elbow assumes a very deep position through the upper two-thirds of the forearm, though it becomes more superficial as it approaches the wrist. Exposure of the nerve in the forearm is greatly facilitated by a clear understanding of certain anatomic relationships. The operator attempting this exposure should avail himself of an opportunity to become familiar with these points, by recourse to the dissecting room, if his visualization of this regional anatomy is at all obscured. The following anatomic points require special consideration: (1) The anatomy of the antecubital fossa; (2) the relation of the pronator radii teres to the median nerve; its insertion in the radius and its relation to the flexor carpi radialis; (3) the radial origin of the flexor sublimis digitorum and the arching fibers of this origin, which surround the nerve; (4) the relation of the median nerve in the wrist to the tendons of the palmaris longus and flexor carpi radialis.


1017
            

The antecubital fossa is a triangular space, the base above lbeing formed by an imaginary line connecting the humeral condyles. The medial boundary is formed by the lateral edge of the pronator radii teres, while the lateral border is formed by the medial edge of the brachioradialis. The biceps tendon passes through the lateral side of the fossa. Both nerve and artery rest upon the floor of the fossa which is formed by the brachialis anticus. In the lower part of the fossa, the brachial artery divides into its radial and ulnar portions. The fossa is covered by the deep fascia and in its lower part by that thickened portion of the deep fascia which constitutes the bicipital fascia. Exposure of the nerve in or just above the antecubital fossa is facillitated by identifying the medial border of the biceps and its tendon; both the nerve and artery lie to its medial side. A tape placed around the artery in the upper part of the incision will facilitate its control in case of emergency, while making a scar tissue dissection, the vein being controlled from below.
   

The pronator radii teres passes from the internal humeral condyle diagonally across the forearm to be inserted in the middle of the radius by a long flat tendon which partially encircles the bone. This muscle forms the lower medial boundary of the antecubital fossa. It is composed of two heads, a large superficial head taking origin from the internal condyle in common with the other superficial mucles. The deep head of the pronator arises from the ulna and is very much

FIG. 176.- Exposure of median nerve in lower arm and upper forearm, showing its relation to the brachial artery and biceps; both artery and nerve are covered by the bicipital fascia A Median nerve; B, brachial artery; C, branch to pronator teres; D, radial artery; E, bicipital fascia; F, internal intermuscular septum; G, ulnar nerve; H, brachial epicondyle; I, deep fascia covering superficial flexors

smaller than the superficial head. The median nerve passes in the cleft between the two heads of the pronator. The radial artery, after the division of the brachial, passes to the radial side of the forearm above the pronator teres, while the ulnar artery directed to the ulnar side, passes beneath the deep head of the pronator and is thus separated from the median nerve. The lower or medial border of the pronator has a direct fascial attachment with its adjoining muscle, the flexor carpi radialis. The insertion of the pronator into the radius is covered by the brachioradialis, and along the under surface of the latter passes the radial nerve. The median nerve in the antecubital fossa


1018

and under the superficial head of the pronator gives off its important motor branches to the extrinsic muscles. In this region great care must be used for the preservation of these branches. Their exit from the parent trunk may be, superficially, from either side. Their deep origin within the median trunk is from the medial portion of the nerve; the lateral being the sensory portion. The branches to the pronator, the flexor carpi radialis, and

FIG. 177.- A, Exposure of median nerve in the antecubital fossa; bicipital fascia divided, pronator teres mobilized from its attachment to flexor carpi radialis. B, Numeral head of pronator teres divided and retracted, exposing branches of the median nerve in this region. Deep head of pronator is seen passing below the median nerve and crossing the ulnar artery. Branches of the median nerve to the forearm muscles are shown springing from its medial side. A, Median nerve; B, brachial artery; C, branches to pronator teres; D, motor branches to superficial flexors; E, medial epicondyle; F, pronator teres, deep head; G, pronator teres, superficial head; H, bicipital fascia; I, ulnar artery; J, radial artery; K, motor branches to deep flexors; L, flexor sublimis digitorum, split; M, flexor carpi radialis

the palmaris longus are usually given off higher than the branches to the deep muscles.

Exposure of the nerve under the pronator is best accomplished by dividing its tendon near its insertion into the radius, care being taken to avoid the radial nerve and artery which pass over the tendon in this region. The radial artery must be mobilized and freed from this muscle, which will necessitate the ligation


1019

tion of several small branches. After the tendon of the pronator has been divided, this muscle may be reflected toward the medial side of the forearm, exposing its under surface and deep head which will be found passing around the median nerve. After the division of the pronator tendon, its medially adjacent muscle, the flexor carpi radialis, may be partially retracted toward the medial side of the forearm, exposing the flexor sublimis digitorum and its tendinous arch, which in reality is the keynote of the situation.
           
The flexor sublimis digiforum
.-The points to be remembered in connection with this muscle are that it covers the median nerve, which passes along its posterior surface; it arises from four heads, the first three of which are from the ulna side and more or less in common with the superficial muscles: the fourth head is from the radius, and the fibers radiating between the radial and humeral heads form an arch under which the median nerve and ulnar artery pass. The radial head should be carefully divided and the muscle gently retracted toward the medial side of the forearm with the pronator. This procedure will give adequate exposure of the median nerve throughout its deep portion.

FIG. 178.- Intraneural dissection of median branches in the forearm, showing their origin from the medial trunk. Highest branches to be given off are to the pronator, whose branches originate from the most ventral portion of the motor quadrant. The volar interosseous likewise arises from the medial side of the median trunk, though occupying a more dorsal position than the bundles to the superficial muscles. A, Median nerve; B, branch to pronator teres; C, branches to superficial fleors; D,branch to deep flexors; E, branch to pronator quadratus; F, division of brachial artery; G. ulnar artery; H, pronator teres, superficial head; I, pronator teres, deep head: J, flexor carpi radialis K, palmaris longus; L, portion of flexor sublimis digitorum; M, flexor carpi ulnaris

The median nerve above the wrist occupies a relatively superficial position, lying beneath and to the radial or lateral side of the palmaris longus tendon; in the absence of this tendon, the nerve will be found just internal to the tendon of the flexor carpi radialis. If the forearm has suffered extensive scar invasion, with destruction of tendons, etc., it may be necessary to first locate the nerve in this position at the wrist.


1020

EXPOSURE IN THE FOREARM

Exposure of the median nerve in the forearm is accomplished by a straight incision of the dlesired length, extending from the medial side of the biceps just above the antecubital fossa, in a line toward the insertion of the flexorcarpi radialis tendon. In most instances, it is necessary to extend the incision from the upper part of the antecubital fossa to the lower third of the forearm. In the upper part of this incision the median basilic vein is exposed and divided between forceps. At a point corresponding to the insertion of the biceps tendon, the sensory portion of the musculocutaneous nerve is usually encountered in the superficial fascia and its fibers preserved if possible. In the upper part of the incision the contents of the antecubital fossa are exposed by dividing the deep and bicipital fascia: the median nerve and the brachial artery are now identified as they pass along the medial side of the biceps tendon, the artery lying between the nerve and the tendon. The median nerve mav now he followed to where it passes below the superficial head of the pronator. In the antecubital fossa, great care should be exercised in avoiding injury to the motor branches to the pronator and the extrinsic muscles. If this exposure is not sufficient to identify the lower segment of the nerve, it must be approached in the lower third of the forearm and followed upward. To permit a lower exposure, the skin incision is extended to the lower third of the arm and the deep fascia divided, exposing the tendon of the flexor carpi radialis, and the median nerve located below the tendon. At the junction of the middle and lower third of the forearm, the median nerve is somewhat deeper than at the wrist, and it may be exposed by retracting the tendon of the flexor carpi radialis laterally. exposing below the aponeurotic medial edge of the flexor longus pollicis: the nerve will now be found in the cleft between the latter muscle and the flexor sublimis digitorum. The median nerve, having been exposed in the upper and lower thirds of the forearm, will require exposure in its middle third, where it is deeply placed between the sublimis and profundus digitorum muscles. Exposure of the median nerve in the middle third of the forearm is the most difficult step in the operation, and is accomplished by dividing the tendon of insertion of the pronator radii teres and reflecting this muscle with the flexor carpi radialis medially, exposing the radial head of the flexor sublimis digitorum, which in turn is divided and also retracted medially; the nerve is identified as it lies on the ventral surface of the profundus dligitorum. If precaution is not taken, serious trouble may be experienced in this exposure through injury to arterial twigs leaving the radial artery. The brachial artery, before it divides into its terminal radial and ulnar divisions in the antecubital fossi, may be secured for control by the passage of tape around the vessel, which may be tightened at will. The radial artery passes over the tendon of the pronator radii teres near its insertion in the radius and passes down the forearm undercover of the brachioradialis, accompanied by the radial nerve. In its course down the forearm, the radial artery gives off numerous twigs which must be ligated before attempting division and retraction of the pronator. After the radial artery has been mobilized, it is retracted mediallv with the brachioraidialis, exposing the flattened tendon of the pronator, which is divided, and the muscle belly reflected. The radial or lateral edge of the flexor earpi raidialis is freed


1021

and retracted with the pronator, exposing the ventral surface of the flexor sublimis digitorum. With the reflection and exposure of the under surface of the pronator, the median nerve will be found passing between its two heads, the deep head of this muscle being but little larger than the median trunk--it may be divided to permit mobilization of the nerve. The median nerve may now be followed to where it disappears below the arched fibers connecting the two heads of the flexor sublimis digitorum. The radial head of this muscle is small and fibrous, and should be divided to permit its medial reflection, which is readily accomplished after freeing its radial border: the median nerve is now completely exposed throughout the forearm. Its important motor branches are given off in the lower part of the antecubital fossa, in the middle third of the forearm beneath the pronator and the early part of its course below the sublimis. In making this exposure, great care should be used in preventing injurN to these branches. The identification of branches should proceed from above downward, resorting to electrical stimulation to confirm identification, if branches are intact. (See general anatomy of the median nerve, p. 1013.)

PHYSIOLOGIC APPROXIMATION IN THE UPPER TWO-THIRDS OF THE FOREARM

Before complete dissection of the nerve is attempted, the center of its ventral quadrant should be marked with identification sutures to be used in subsequent alignment during approximation. In approximation of the median nerve

FIG. 179.-  Median nerve lesion in middle third of in the upper forearm, it is essential forearm; nerve exposed above and below lesion and that the branches to the pronator and marked with identification sutures. In this in-long flexor muscles be identified and stance the nerve is exposed below the pronator teres by dividing the flexor sublimis digitorum, A, Branch to pronator teres; B, pronator teres; C, bicipital fascia; D, flexor carpi radialis; E, median nerve, upper segment; F, flexor sublimis digitorum, split; G, median nerve, lower segment: H, median nerve; I, brachial artery; J, radial artery; K, brachioradialis

in the upper forearm, it is essential that the branches to the pronator teres, lesions above this point will probably involve most of these branches before they have left the median trunk. It is essential, therefore, that the medial side of the nerve trunk be accurately identified in both proximal and distal ends of the divided nerve, as the bundles forming the motor branches are located in this portion of the median trunk, the fibers to the pronator being more


1022

ventral than those to the palmaris longus, flexor carpi radialis, and sublimis digitorum. Often in this region a dissection of the median trunk will reveal all its motor branches incorporated in an individual sheath, lying along the medial side of the nerve. In such instances, individual approximation of the motor segment is possible.

FIG. 180.- Transposition of median nerve to a plane superficial to superficial head of pronator radii teres. Intraneural mobilization of branches. Continuity defect may now be overcome by flexion-relaxation of elbow. A, Median nerve; B, motor branches, mobilized; C, brachial Artery; D, radial artery; E, brachioradialis; F. bicipital fascia; G, pronator teres; H, flexor fascia; I, sublimis digitorum

FIG. 181.- Median nerve transposed to overcome median defect and sutured; nerve now passes superficial to pronator radii terres and below flexor carpi radialis. Nerve sheath partly closed. A. Median nerve; B, brachial artery, C, radial artery; D. Pronator teres; E. Motor branches; F, brachialis anticus; G, biceps; H, bicipital fascia; I, brachioradialis; J, flexor carpi radialis; K, flexor sublimis digitorum


Lesions of the median trunk below the pronator seldom involve all its motor branches, though in extensive lesions many of them may suffer individual injury. The motor branches to the profundus digitorum, flexor longus pollicis, and pronator quadratus may be given off from the medial side of the nerve below the flexor sublimis digitoruni, as individual branches, but more often they


1023

leave the nerve trunk as the volar interosseous nerve in the region of the pronator, and occasionally it is possible to effect individual suture of this rather large branch, which will be found passing down the cleft between the profundus digitorum and the flexor longus pollicis, accompanied by the volar interosseous artery.

After the motor branches to the extrinsic muscles have left the median trunk, its content in motor fibers consists solely of those to the median intrinsic hand muscles, which lie in the dorsal quadrant of the nerve. Though the intrinsic hand muscles seldom recover function after a long period of denervation, the operator should exert just as great care in preventing torsion of the median trunk in this region, for a disturbance of its sensory pattern will result in defective tactile localization, which will greatly diminish the functional usefulness of the digits.

CONTINUITY DEFECTS

If the continuity defect of the nerve be too great to be overcome by primary stretching and flexion-relaxation of the elbow and wrist joints, transposition is indicated, in which the median nerve is transposed to a position above the superficial head of the pronator teres. In order to make this procedure available, the motor branches must be gently mobilized by an intraneural dissectionof the nerve trunk after its sheath is opened; failure to do so will result in their destruction. Most of these motor branches may be readily isolated from the parent trunk to some distance above the medial condyle. The median nerve in its transposed position passes along the radial border of the flexor carpiradialis above the pronator teres. A transposition of this nature will usually shorten the distance of a defect from 4 to 7 cm. during forearm flexion. If, after transposition, it still remains impossible to approximate the nerve ends, the unsectioned ends of the nerve should be sutured for secondary stretching and resort made to the two-stage operation, rather than to grafting. Failing in this, the surgeon's only recourse will probably be to grafting, for which the radial nerve, lying just under the exposed brachioradialis may be used to advantage.

MEDIAN LESIONS AT THE WRIST AND COMBINED TENDON INJURIES

Most lesions involving the median nerve at the wrist are combined with lesions of the flexor tendons. It is therefore necessary to exercise great care in these dissections, for the normal anatomic relationships are frequently lost with the destruction of the tendons. It is advisable always to first identify and isolate the median nerve to prevent secondary injury to this structure during the dissection of the tendons.

EXPOSURE AT THE WRIST

It is essential to identify the tendon of the palmaris longus, if this muscle be present; likewise the tendon of the flexor carpi radialis. The median nerve passes under the annular ligament just below the palmaris longus tendon, to the ulnar side of the flexor carpi radialis tendon and to the radial side of the


1024

sublimis tendon; identification of these structures will usually facilitate exposure of the nerve. If the lesion is low, it may be necessary to divide the annular ligament and palmar fascia, exposing the nerve in the palm where it may be readily identified as it passes along the radial side of the sublimis tendons.

If difficulty is experienced in locating the upper end of the nerve, the flexor carpi radialis tendon should be followed up, exposing the line of cleavage between the tendon of the flexor sublimis digitorum and the flexor longus pollicis. Retraction of the flexor sublimis digitorum will usually reveal the nerve on the under surface of this muscle. If this exposure fails to reveal the location of the nerve, it may be necessary to locate it as it passes under the pronator radii teres and the tendinous arch of origin between the heads of the flexor sublimis digitorum, as in middle forearm exposures. The palmar cutaneous branch is given off from the anterior surface of the median, a few centimeters above the annular ligament; in the presence of much scar tissue its identification is extremely difficult and when possible its repair should be attempted. After the median nerve has been located and before it has been completely isolated from its bed, it should be carefully marked with identification sutures, as in the lower third of the forearm funicular topography can not be determined by branch identification. Occasionally, advantage may be taken of the fact that the anterior surface of the nerve is frequently marked by the comes nervi mediana artery passing along its sheath; this artery is quite large and its presence may facilitate surface identification. Electro-anatomic identification at the wrist is of but little value, as the greater bulk of the nerve in this region consists of sensory fibers, its only motor fibers being those to certain intrinsic hand muscles which are located in the dorsal part of the nerve trunk. Even though the nerve is mostly sensory in this region, it is necessary to prevent torsion during suture, for sensory fibers normally innervating one finger may be directed to another, resulting in confusion of tactile localization. The writer has observed an officer whose median nerve was sutured at the wrist; after four years there was a complete return of al forms of sensation, though touch on the third finger was always localized in the thumb, which inhibited to a considerable extent manipulative dexterity in handling small objects. 'This officer is gradually becoming accustomed to this sensory perversion, by psychological reversion and experience. This case serves to illustrate the confusion arising from the distortion of sensory fibers during suture.

REPAIR OF TENDONS IN COMBINED NERVE AND TENDON LESIONS

After the nerve has been properly identified and isolated, and before suture or the resection of scar tissue from its ends, the injured tendons must receive careful attention. They should be completely isolated from the scar tissue and every vestige of scar carefully removed with as little trauma as possible. Tendons completely divided should be resutured and when this is impossible, because of extensive destruction, they should be united to neighboring tendons having like function. The surgeon should always endeavor, in reconstructive surgery of the wrist, where distortion of the original tendon


1025

pattern is necessary, to provide tendon anastomoses of a physiologic type for both ends of each divided tendon. After the tendons have been properly cared for and perfect hemostasis assured, the ends of the divided nerve may be prepared and approximated, with due respect to physiologic alignment. Flexion of the wrist will assist in facilitating approximation, though when this is necessary in the presence of combined tendon lesions, flexion of the fingers is not advisable. Early gentle passive movements are essential to success in combined nerve and tendon lesions of the wrist, and these passive movements should be persistently encouraged throughout the healing process, until such a time as they may be replaced by active movements, endeavoring always to simulate the full range of normal coordinated hand and finger movements; the importance of these passive exercises can not be too strongly emphasized; they should be begun not later than 48 hours after the operation.

SURGERY OF THE MEDIAN NERVE IN THE HAND

Only occasionally is it possible to effect suture of the terminal digital branches of the median nerve. In the presence of extensive scar tissue it will probably be found impossible, though small incised lesions may lend themselves to successful repair. The writer has on two occasions been successful in effecting satisfactory suture of the median nerve in the palm where the injury was caused by a penetrating gunshot wound. In one instance, the suture was effected just before the nerve divided into its terminal digital branches. In the other instance, the division was about the point where the terminal branches were given off. In this case the individual branches were collected and united to the end of the trunk, apparently without torsion, as regeneration resulted in accurate tactile localization.

Occasionally the median nerve in the hand may be incorporated in scar tissue, producing an irritative syndrome, with retained sensation. Neurolysis will usually give relief. Incised wounds of the fingers occasionally present sensitive neuromas, which are a source of constant annoyance. In such instances the nerve should be exposed and sutured if possible; when approximation fails, the neuroma should be excised, the end of the branch strongly litigated and, proximal to the ligature, injected with pure alcohol.
 
EXPOSURE IN THE PALM

Exposure of the median nerve in the palm is accomplished by an incision extending from the insertion of the palmaris longus tendon into the palmar fascia to a point corresponding to the interval between the index and ring fingers. The incision is carried through the integument and fat of the superficial fascia, exposing the palmar fascia, which is divided in the line of the cutaneous incision. The long flexor tendons are identified, and at the base of the palm the median nerve will be found lying to the radial side of these tendons. Shortly after the nerve passes under the annular ligament branches are given off. which pass radially to innervate the superficial thenar muscles, and supply sensation to the thumb. It is highly essential that these branches be preserved as they control to a great extent the opposing function of the thumb.


1026

As the median nerve progresses through the palm it attains a position superficial to the flexor tendons, but passes below the superficial palmar arch. Proximal to the metacarpophalangeal joints the nerve divides into its terminal branches, where it is enlarged, flattened, and, in contrast to the fatty tissue in which it lies, has a somewhat pinkish color. The median divides into five terminal branches. supplying sensation to the first three and radial half of the fourth fingers. These branches may readily be identified, as they lie imbedded in the deep palmar fat, superficial to the flexor tendons, providing the operator has achieved adequate hemostasis, which is highly essential in such dissections and which in the writer's experience has been satisfactorily accomplished only by the use of infiltration anesthesia, containing a small amount of active adrenalin to blanch the tissues.

IRREPARABLE LESIONS

Irreparable median nerve lesions present a serious problem. Though the action of extrinsic muscles be retained and finger and thumb movements fairly well preserved, the loss of sensation in the first three fingers is a serious handicap and every endeavor should be made to restore, if possible, some type of sensation in the median distribution to the fingers. The surgeon may have recourse to autogenous grafts and may utilize a segment removed from the radial nerve. If this seems impractical,

FIG. 182.- Branches of median nerve in hand. A, Palmar fascia, cut edge; B, annular ligament; C, median branches to thenar muscles; D, branches to lumbricales; E, thenar formation, F, digital sensory branches; G, palinaris longus tendon

because of extensive scar tissue formation, the procedure of niting the radial to the median, recommeded and practiced by Harris, may be considered. In this procedure, the distal end of the radial nerve is divided at a point where it may be brought in contact with the lower segment of the median, to which it is anastomosed, for the purpose of supplying the anesthetic digits with a radial sensory innervation. It must be remembered. however, that this sensation is localized in the normal radial sensory area. It may correct trophic disturbance and serve in the prevention of injuries to the digits, but the function of tactile localization, so necessary to manual dexterity, is lacking.


1027

In median lesions with extensive continuity defects, the severity of the lesion to the surrounding parts makes the possibility of successful grafting rather remote; the functional disability remaining after radial anastomosis renders this operation a procedure of last resort. It behooves the surgeon, therefore, to exert every effort to procure end-to-end suture of the median, using the two-stage operation, and if necessary, extending the dissection high up in the arm, or even in the axilla. If there also exists a combined ulnar lesion with paralysis of the ulnar intrinsic hand muscles, which has existed beyond one year, and the proximal end of the ulnar is in a position where it may be brought in contact with the distal end of the median, the ulnar may be anastomosed into the median. This procedure will give, perhaps, better functional results than radial anastomosis, though it precludes the possibility of subsequent ulnar regeneration with a continuation of complete anesthesia in the little finger, and interossei paralysis. After one year, it is extremely doubtful whether interossei regeneration would ever occur, and the little finger, if necessary, may be amputated, if trophic changes and traumatic insults to his anesthetic member should necessitate such a procedure.

TENDON TRANSPLANTATION FOR PARALYSIS OF THE MEDIAN NERVE

If it is possible to provide sensation to the fingers and hand, a varying degree of functional usefulness may be restored, even though there exists a complete paralysis of all median innervated muscles. The flexor profundus digitorum, supplied by the ulnar, provides flexion of the fourth and fifth fingers, and frequently the middle finger. The problem, therefore, consists in restoring flexion and opposition to the thumb, and flexion to the index finger and occasionally the middle. Frequently the ulnar supplies the superficial head of the short flexor of the thumb, or a slight change in the insertion of this muscle diminishes greatly the disability of opponens paralysis, so that in a certain number of cases opposing thumb function is fairly well preserved and tendon transplantation is required only to restore function in the long flexors of the thumb and index finger.

In those cases of median paralysis in which the thumb is displaced laterally and call not be made to assume its normal opposed position, an arthrodesis of the carpometacarpal joint in a position of opposition and abduction with an angle of about 600 between the first and second metacarpals. as recommended by Baldwin, is a very satisfactory procedure. A fibrous ankylosis probably gives better results after this artbrodesis than a solid bony ankylosis. Ney has obtained very satisfactory opponens function of the thumb by tendon transplantation, in which the short extensor of the thumb is divided at the wrist and freed to its insertion. The palmaris longus is then exposed and divided just above the annular ligament. The short extensor of the thumb is now passed through a tunnel under the subcutaneous fat of the thenar eminence to the palm of the hand, and after being passed under the annular ligament is transplanted into the tendon of the palmaris longus, or, the latter muscle being absent, into the flexor carpi radialis. If the flexor carpiradialis and palmaris longus are paralyzed, it may be inserted into the extensor


1028

carpi radialis tendon or the flexor carpi ulnaris tendon, which are being utilized to restore flexion to the distal phalanx of the thumb and the index finger. This procedure for obtaining opponens action finds its greatest usefulness in combined median and ulnar lesions with complete intrinsic hand muscle paralysis, under which subject the technique is described in detail.

FOR PARALYSIS OF THE LONG FLEXOR OF THE THUMB AND THE FLEXORS OF THE INDEX FINGER

An incision is made about 5 cm. long from the base of the metacarpo-carpal joint of the thumb along the radial border of the forearm, following the radial edge of the flexor carpi radialis. Undermining this incision medially, the deep flexor tendons are exposed and the long flexor of the thumb located just below the tendon of the flexor carpi radialis. Retraction of the edges of the wound laterally will expose the tendon of the extensor ossis metacarpi pollicis which is divided near its insertion into the base of the first metacarpal bone. The radial artery and its several branches lie in the space between these tendons and may require ligation. After the tendon of the extensor ossis metacarpi pollicis has been freed and divided it is anastomosed to the divided end of the flexor longus pollicis, under sufficient tension to insure flexion of the distal phalanx of the thumb. The flexor tendons to the index finger may now be identified and freed. An incision on the dorsum of the wrist over the tendon of the extensor carpi radialis is made from its insertion some distance up the forearm. The tendon of the radial carpal extensor is divided, carefully mobilized, and passed through a subcutaneous tunnel from the dorsum of the forearm to the incision on the ventral surface and transplanted into both flexor tendons of the index finger. During this procedure the wrist is held in flexion and the index finger about half flexed. It is essential in the postoperative treatment that early passive movements be instituted to prevent adhesions, which would tend to keep the index finger in permanent flexion. In our experience these procedures have given perhaps the best results, though very satisfactory results have followed a transplantation of the index flexor tendons into the flexor carpi ulnaris.

ULNAR NERVE

GENERAL ANATOMY

The ulnar nerve arises from the inner cord of the brachial plexus in common with the inner head of the median and the internal and lesser internal cutaneous nerves, most of its fibers being derived from the anterior divisions of the eighth cervical and first thoracic roots. The nerve lies medial to the inner head of the median and lateral to the internal cutaneous. In passing through the axilla and the upper two-thirds of the arm, the ulnar occupies a position medial to the artery and the median nerve in the neurovascular bundle. In the lower part of the middle third of the arm the nerve leaves the neurovascular bundle, and penetrating the internal intermuscular septum follows a posterior course to a position posterior to the internal humeral condyle, lying in close relation to the posterior surface of the internal intermuscular septum, and crossing the fibers of the medial head of the triceps, which|


1029

arise from this portion of the septum. In this region the ulnar nerve is accompanied by the inferior profunda artery, which usually lies on its ventral surface. The profunda artery in the region of the internal humeral condyle anastomoses freely, as dloes its vene comites, with the posterior ulnar recurrent artery and veins. Behind the medial condyle the ulnar nerve occupies a very superficial position and is covered by a thick, aponeurotic fascia, which extends from the condvle to the olecranon. The nerve enters the forearm by passing under an aponeurotic arch, connecting the humeral and ulnar leads of the flexor carpi ulnaris, and continues its course down the forearm, covered by the belly of this muscle and lying on the profundus digitorun. In the upper third of the forearm the ulnar nerve is separated some distance from the ulnar artery, while in the middle and lower thirds they lie in direct contact. In the middle and lower thirds of the forearm the ulnar nerve lies dorsal to the tendon of the flexor carpi ulnaris; at the wrist it passes through a special compartment in the anterior annular ligament to the radial side of the pisiform bone to enter the palm, where it divides into its terminal superficial and deep palmar branches.

BRANCHES

The only branches of the ulnar nerve given off in the arm are in the region of the internal humeral condyle, where several small twigs are distributed to the elbow joint. About 2 cm. below the medial condyle two branches are given off to the two heads of the flexor carpi ulnaris; at a slightly lower level another set of branches is given off to the ulnar side of the flexor profundus digitorum. These motor branches arise from the medial side of the dorsal quadrant of the nerve; they pursue a relatively long intraneural course and may be followed up the ulnar trunk for some distance above the medial condyle by an intraneural dissection. In the middle third of the forearm the ulnar gives off a relatively insignificant palmar cutaneous branch, which distributes branches to the ulnar artery and follows the artery to the annular ligament, over which it passes to supply a small area at the base of the hypothenar eminence. At the junction of the middle and lower third of the forearm there springs from the medial side of the ulnar trunk a large dorsal cutaneous branch, which at times nearly approximates the ulnar trunk in size This lorsal cutaneous branch, after passing under the tendon of the flexor carpi ulnaris to the dorsal aspect of the wrist, becomes cutaneous and supplies sensation to the ulnar dorsal border of the hand, and to the dorsum of the little finger as far as its terminal phalanx and the ring finger to its middle phalanx.
           
Terminal palmar branches
.-After passing under the annular ligament the ulnar nerve terminates in a superficial and deep branch. The superficial branch, beyond giving a few twigs to the palmaris brevis muscle, is strictly sensory and supplies sensation to the ulnar side of the palm and divides into an outer and inner digital branch, supplying sensation to the little finger and adjacent side of the ring finger. The deep branch is principally motor; after its separation from the ulnar trunk, it gives off branches to the muscles of the little finger and enters its deep course in the palm, accompanied by the ulnar


1030

artery, by passing between the abductor and flexor brevis minimi digiti muscles, penetrating the opponens minimi digiti. It follows the course of the deep palmar areh beneath the flexor tendons, giving off branches to the dorsal and palmar interossei muscles and the two inner lumbricales, and terminates by sending motor branches to the adductor pollicis and the inner head of the flexor brevis pollicis.

SURGERY

Lesions of the ulnar nerve in the axilla are usually combined lesions of the terminal plexus type, and are dealt with as such.

In the upper arm the ulnar nerve lies within the neurovascular bundle, and any traumatism sufficient to injure the ulnar will probably result in injury to the median. As the ulnar nerve leaves the axilla under the lower border of the pectoralis major, it passes across the tendon of the latissimus dorsi, in close relation to the musculospiral, and in this region both occasionally suffer simultaneous injury.

EXPOSURE IN THE ARM

Exposure of the ulnar nerve in the upper part of the arm is obtained as with the median, through a long incision, paralleling the neurovascular bundle. This incision differs, however, in that at the junction of the middle and lower third of the arm it diverges posteriorly, following the internal intermuscular septum to the medial humeral condyle. The deep fascia is now opened and the medial border of the coracobrachialis above and the biceps below retracted, exposing the neurovascular bundle, which lies in the cleft between these muscles and the long head of the triceps. At the lower part of the axilla the neurovascular bundle is covered by the pectoralis major and, to obtain an axillary exposure, the pectoral tendon will require division. The neurovascular bundle is opened, exposing its contents and the ulnar nerve identified above the lesion. About the junction of the middle and lower third of the arm the lower segment of the ulnar nerve is identified posterior to the internal intermuscular septum, after it has left the neurovascular bundle. Before disturbing the circumferential relationship of the nerve by complete isolation, in both upper and lower positions, the center of its lateral quadrant should be carefully marked with accurately placed identification sutures. The dissection of the nerve may now progress to the point of lesion. If the identification sutures previously placed are found to be some distance from the lesion, as the dissection progresses they should be supplemented with additional sutures, carefully observing the surface markings of the nerve. Much difficult y is experienced in the dissection of the neurovascular bundle, particularly when it is invaded with sear tissue, due to bleeding from numerous venous and arterial radicals. This difficulty may in a large measure be obviated by isolation of the basilic and brachial veins below, and the application of permanent or temporary constriction. Arterial hemorrhage may be controlled by isolating the brachial artery above and applying a tape which may be adjusted to control its circulation while branches are being ligated or accidental wounds in the arterial wall repaired.

Defects in the continuity of the ulnar trunk in the arm, if of a magnitude not correctable by primary nerve stretching, will call for transposition of the


1031

nerve from behind the medial condvle of the humerus to a more ventral position, after which advantage nmav be taken of elbow flexion-relaxation. When this procedure fails to overcome the defect, the ulnar nerve should be freely mobilized in the axilla and recourse taken to the two-stage operation.

SURGERY IN TIIE REGION OF THE INTERNAL HUMERAL CONDYLE

The ulnar nerve, as it passes behind the internal humeral condyle is exposed to frequent trauma. Its involvement in war injuries is second only to lesions of the musculospiral. Its lateral surface, being in direct contact with the bone, and its medial surface being very superficial, renders it particularly susceptible to trauma.

PROGRESSIVE PERIFERAL ULNAR PARALYSIS

Severe injury to the ulnar nerve usually results in complete immediate paralysis, while frequently repeated slight trauma will result in a gradual progressive type of peripheral paralysis, in which sensation is slowly lost and atrophy of the intrinsic hand muscles supervenes. The ulnar nerve, because of its exposed position behind the medial huuneral condyle, is commonly the recipient of frequent mild degrees of trauma, though of suflicient intensity to institute a gradual compression paralysis, due to the production of intraneural scar tissue. Occasionally this type of ulnar paralysis will develop years after a trauma to the medial condvle or the elbow joint. The original trauma may escape memory, but is probably responsible for the institution of some anatomic change in the bone proximal to the nerve, or for a thickening of the aponeurotic fascia covering the nerve, producing compression or intensifying the effect of ordinarv trauma to this region.

In lesions of this type, the ulnar nerve shouldl be transposed from its exposed position behind the humeral condyle to the less traumatized region anterior to the condyle. This procedure will usually suffice to correct most progressive peripheral lesions, though when the nerve is found to be indurated, its sheath should be opened and the bundles subjected to internal neurolysis. The nerve sheath is allowed to remain open for its decompressive effect.

EXPOSURE

The surgery of the ulnar nerve in the region of the internal condyle necessitates exposure in its lower humeral and upper forearm course. With the arm fully abducted and in external rotation, the forearm flexed and in extreme supination, a curved incision is carried from the junction of the middle and lower third of the arm, beginning along the medial border of the biceps, following the course of the nerve downward behind the internal condvle to the forearm where it follows its anterior ulnar border to the junction of its upper and middle third. In the line of this incision, the deep fascia is divided and the nerve identified distal and proximal to the lesion. In its lower humeral course. the nerve will be found, after having penetrated the internal intermuscular septum, following the posterior border of this structure, lying upon the internal head of the triceps or covered by its superficial fibers. It is accompanied by the inferior profunda artery which may usually be separated from the nerve


1032

to the internal condyle, where ligation will be required, as in his region the artery and its venae comites anastomose freely with the ulnar recurrent artery and vein. The preliminary ligation of these vascular structures, both above and below the internal condyle, where they lie in direct contact with the nerve, will greatly enhance the dissection. The ulnar collateral branch of the musculospiral nerve, supplying the medial head of the triceps, follows a course along the posterior border of the internal intermuscular septum in close relation to the ulnar nerve and care should be used in avoiding its injury. At the internal medial condyle, the ulnar nerve is protected by a thick aponeurotic membrane stretching from the medial condyle to the olecranon, as they are occasionally adherent. Before the relationship of the nerve has been disturbed in its bed, it should be carefully marked by sheath identification sutures to prevent torsion during subsequent suture. The nerve, after being uncovered in its course posterior to the medial condyle, will be found entering the foramen bypassing under the tendinous arch, connecting the two heads of the flexor carpi ulnaris. The fibers of this arch are divided and the line of cleavage between the two heads of the muscle separated for a short distance, exposing the nerve in its passage below this muscle, where it lies upon the muscular fibers of the flexor profundus. In the region of the condyle and below, important branches will be found innervating the flexor carpi ulnaris and the ulnar half of the profundus digitorum; these motor branches spring from the medial and posterior quadrant of the nerve. The branches to the flexor carpi ulnaris leave the ulnar trunk somewhat higher than those to the profundus-usually about the level of the condyle. The fibers innervating the flexor carpi ulnaris may leave the nerve trunk as a single bundle, subsequently dividing to supply its two heads, or these branches may leave the parent trunk after their division. If the ulnar sheath is opened, however, they will be found united into a single branch or bundle, and may be followed some distance above the internal condyle. The lower branch, innervating the ulnar half of the flexor profundus digitorum, may likewise leave the nerve trunk as a single branch, or after its division. Within the nerve trunk, however, it may be followed as a single bundle some distance above the condyle, where it will be found to join with the bundle containing the fibers to the flexor carpi ulnaris. After these two bundles have combined, their intraneural course may frequently be followed as high as the middle third of the arm. The intraneural dissection of these branches is extremely important in transposing the nerve to a position anterior to the humeral condyle, in the correction of continuity defects-it is our only means of preserving these branches during such a procedure. Occasionally, in lesions in the region of the medial condyle, these motor branches may be found divided; and if they can not be identified and repaired, provision should be made for ulnar profundus paralysis by dividing the two medial profundus tendons above the wrist and uniting them with the two lateral profundus tendons which are supplied by the median nerve.

TRANSPOSITION OF THE ULNAR NERVE

Transposing the ulnar nerve from its position behind the internal humeral condyle to the anterior surface of the elbow is a valuable procedure in over-


1033

coming defects in the continuity of the nerve; it is also indicated in progressive peripheral ulnar paralysis in which the nerve is subjected to trauma when occupying its normal position. In overcoming defects, transposition of the ulnar nerve will shorten its course and render it susceptible to flexion-relaxation

FIG. 183 - Ulnar nerve, showing scar tissue as found at operation. Nerve exposed by splitting the two heads of flexor carpi ulnaris; exposure of distal and proximal segments. Identification sutures placed. A, Medial epicondyle; B. olecranon; C, ulnar nerve, proximal segment; D, flexor carpi ulnaris, ulnar head; E, ulnar nerve, distal segment; F flexor carpi ulnaris, humeral head; G, internal intermuscular septum

FIG. 184.– Ulnar nerve exposed above medial humeral condyle preparatory for transportation anterior to the condyle. Intraneural exposure and mobilization of branches to ulnar half of flexor sublimis digitorum and flexor carpi ulnaris; mobilization of branches should be carried at least 3 cms. Above condyle. A, Medial epicondyle; B, flexor carpi ulnaris, humeral head; C, ulnar nerve; D, wall of ulnar canal; E, branch to flexor carpi ulnaris; F, branch to flexor profundus digitorum; G, ulnar nerve, distal segment.

of the elbow, when defects in the forearm of three or four inches may be readily overcome.

The incision for ulnar transposition follows the course of the nerve from the middle of the arm to the middle of the forearm. The deep fascia is divided in the line of the skin incision and the nerve exposed throughout the length


1034

of this incision, care being taken to place a sufficient number of identification sutures to assure alignment of the nerve, if suture is required. The anterior skin flap containing the fat of the superficial fascia is reflected, exposing the medial border of the biceps and its tendon. The ulnar nerve is now followed upward to the middle third of the arm where it penetrates the internal intermuscular septum; the septum is divided and the nerve made to pass along its ventral border, following the course of the neurovascular bundle to the ante-cubital fossa. To transpose the nerve below the medial condyle so that it will follow a straight line down the arm and forearm, its motor branches to the flexor carpi ulnaris and the profundus digitorum must be mobilized some distance up the nerve trunk to prevent stretching and tearing in transposing the nerve. The nerve in its transposed position passes over those muscles arising from the medial condyle. If the cutaneous structures in this region have been well preserved and contain sufficient subcutaneous fat, the writer

FIG. 185.- Ulnar nerve transposed; defect overcome by transposition and flexion-relaxation of elbow; branches preserved through mobilization

prefers transplanting the nerve in this superficial position, where it is covered merely by the fat of the superficial fascia. Some operators, however, prefer directing the nerve through a muscular tunnel below the common heads of origin of the superficial muscles springing from the medial condyle, the bed of the nerve being the flexor profundus digitorum. This may be accomplished in two ways, depending upon the ability to preserve the motor branches to the flexor carpi ulnaris and profundus digitorum. If the motor branches have been sacrificed, the nerve is passed through a tunnel between the sublimis and profundus digitorum, made by passing forceps between these muscles, widening the tunnel and drawing the free end of the nerve through. The second method consists in dividing the superficial muscles close to their origin to the medial condyle, placing the nerve in the cleft and resuturing the muscles above it. This has the disadvantage of possibly interfering with the nerve supply of normal muscles, and also of inflicting additional trauma to the


1035

structures surrounding the nerve. While the tunneling procedure is less apt to injure the nerve supply of adjacent muscles, it undoubtedly invites an excess of sear tissue by virtue of the trauma inflicted.

EXPOSURE IN THE MIDDLE AND LOWER THIRDS OF THE FOREARM

An incision extending from the radial side of the pisiform bone toward the medial condyle will follow the course of the ulnar nerve in the forearm. After the deep fascia has been divided, the tendon of the flexor carpi ulnaris is identified. It will be remembered that the lateral or radial border of the flexor carpi ulnaris is tendinous for some distance up the forearm, while from its medial or ulnar border muscular fibers radiate almost as far down as the wrist joint.. The ulnar nerve, accompanied by the ulnar artery and its venae comites. lies beneath the radial border of the flexor carpi ulnaris tendon and to the ulnar side of the flexor sublimis digitorum. Following upward the cleft between these muscles, the ulnar nerve will be found resting upon the flexor profundus digitorum and covered by its thin sheath. It is advisable to first identifv the nerve in the region of the wrist, after which it may be followed upward by separating the loose attachment between the flexor carpi ulnaris and the flexor sublimis digitorum: the identification of the line of cleavage between these muscles is facilitated by following up the radial tendinuos edge of the flexor carpi ulnaris. If dificulty is encountered in exposing the proximal end of the nerve, it may be readily identified at the medial condyle, which it leaves to pass between the two heads of the flexor carpi ulnaris. These two heads may be separated for some distance and the nerve followed as it passes below the belly of the humeral head, which latter may the conveniently elevated by tape retraction, its radial border having been previously separated from the sublimis and palmaris longus. In the upper part of the middle third of the forearm, the ulnar artery joins the nerve and follows it throughout the remainder of its course. In the middle third of the forearm a small palmar cutaneous branch is given off, which follows the ulnar artery down the forearm and passes over the annular ligament: this branch is unimportant and may be sacrificed with impunity. About the lower third of the forearm, though at times much higher, a large dorsal cutaneous branch arises from the medial side of the ulnar trunk and passes under the tendon of the flexor carpi ulnaris to supply the integument of the ulnar dorsal side of the hand and fingers. This branch is occasionally of rather large size and must not be confused with the main ulnar trunk. It follows a relatively long intraneural course and its preservation is important, as it supplies sensation to the much traumatized ulnar border of the hand and little finger.

Lesions of the ulnar nerve in the lower two-thirds of the forearm are commonly associated with more or less tendon injury, whieb should be repaired before the nerve is sutured. If sublimis or profundus tendons on the ulnar side are injured, they should be freed of all sear tissue and reunited if possible. If suture of the divided tendons is not possible, they should be anastomosed to intact sublimis or profundus tendons, and their bellies above freed of scar and likewise united. Occasionally the palmaris longus may be transplanted into the distal end of the divided sublimis tendons.


1036

SURGERY OF TIHE ULNAR NERVE IN THE PALM

The ulnar nerve enters the palm by passing through the annular ligament in a compartment of its own along the radial border of the pisiform bone, and in the palm divides into its two terminal branches-a superficial branch, mainly sensory, and a deep branch, principally motor to the ulnar intrinsic hand muscles. These terminal ulnar branches do not lend themselves favorably to surgical repair, though as with the median, occasional instances will be found where extremely localized injuries involving the nerve with small defects will permit repair.

EXPOSURE

Exposure is made through an incision extending from the radial border of the pisiform bone to the interspace between the fourth and fifth fingers. The nerve is localized at the wrist and followed into the hand by dividing the annular ligament, the palmar fascia and the palmaris brevis, exposing the flexor brevis minimi digiti and the opponens of the little finger. By retracting the ulnar border of the incision, the deep branch may be followed and exposed as it passes in a cleft between the flexor brevis and abductor, which muscles require careful identification as they are important landmarks. If the interspace between these muscles is spread, exposing the opponens, the

FIG. 186.- Branches of ulnar nerve in hand. A, Digital cutaneous branch, fourth and fifth fingers; B, digital branch, fifth finger; C, palmaris brevis; D, motor branch to intrinsic muscles; E, ulnar nerve; F, ulnar artery; G, palmar fascia; H, palmaris longus tendon; I, annular ligament

nerve may be followed to a point where it penetrates the latter muscle. After penetrating the opponens, the deep palmar branch passes below the flexor tendons and in this region is practically inaccessible. The superficial palmar branch may be identified by its superficial position and by the distribution of its digital branches. Occasionally it is possible to effect individual suture of these digital branches, their arrangement being similar to the median digital branches.

In attempting repair of the small cutaneous or motor branches of the median or ulnar nerve in the hand, the operator should proceed with extreme care and a full appreciation of the difficulties involved, in that the disability resulting from the inflicted surgical trauma may be greater and overshadow the original disability.


1037

DEFECTS IN CONTINUITY

Because of the passage of the ulnar nerve posterior to the internal humeral condyle, it does not lend itself to flexion-relaxation of the elbow joint. In fact, flexion of the elbow joint tends to stretch the nerve, which normally is relaxed by forearm extension. It therefore becomes necessary in most ulnar defects to resort to transposition of the nerve from its posterior position behind the internal condyle to an anterior position, by which the operator utilizes the elongating effect of directing the nerve from an angular to a straight course, and which also permits an extensive relaxation of the nerve during forearm flexion.

Defects in the forearm up to 3 or 4 inches may be overcome by utilizing transposition of the ulnar nerve with flexion-relaxation of the elbow. Defects of greater magnitude in the forearm require extensive mobilization of the nerve, which should be continued to the wrist below and to the upper third of the arm above, when another inch may be added by ulnar flexion of the wrist and abduction of the arm. If the surgeon, having availed himself of these procedures, finds difficulty in obtaining end-to-end approximation, he must then resort to the two-stage operation, described under general technique. An endeavor should always be made to preserve the motor branches at the elbow, though they may be sacrificed if their extensive intraneural mobilization fails to give sufficient relaxation to the nerve. If it becomes necessary to sacrifice the motor branches to the flexor carpi ulnaris and ulnar half of the profundus digitoruin, the function of the profundus tendons may be maintained by transplanting them at the wrist into the two remaining active profundus tendons.

The surgeon is rarely justified in sacrificing the dorsal cutaneous branch of the ulnar for relaxation purposes. If it is found to interfere with ulnar mobilization, it should be freed as far as possible along the ulnar border of the forearm. Sacrifice of the dorsal cutaneous branch involves a permanent anesthesia of the ulnar border of the hand and little finger, which may in itself necessitate amputation of these parts, because of the constant irritation of trophic or traumatic sores.

In the arm, large defects of the ulnar nerve necessitate exposure well up into the axilla and well down into the forearm. Flexion-relaxation is inhibited in upper defects of the nerve by the motor branches at the elbow, and to obtain its full effect, sacrifice of these branches is usually necessary. In transposing the ulnar nerve for the correction of defects, the surgeon should assure himself that the nerve is freely mobilized in the region of the internal intermuscular septum. Ulnar defects, which can not be overcome by stretching, transposition, and flexion-relaxation of the elbow, wrist, and shoulder are probably beyond repair, though in such cases the surgeon should not neglect the possibilities of a successful graft, but grafting may be used only as a procedure of last resort.

SECONDARY SUTURES

Surgery of the ulnar nerve gives perhaps more unsatisfactory results than any other nerve, primarily because of failure in regeneration of the ulnar intrinsic hand muscles. In the writer's experience, regeneration of the inter-


1038

ossei muscles is absent, after four years, in 94 per cent of ulnar sutures. These delicate, highly specialized intrinsic hand muscles reach an extreme degree of muscle degeneration in a comparatively short period. In several cases of ulnar paralysis, resulting from compression lesions, in which there has been a satisfactory restoration of ulnar sensation following neurolysis, the ulnar intrinsic hand muscles have remained paralyzed where the uncorrected lesion had existed beyond one year. The writer has observed a number of cases in which some function had returned in the muscles of the hypothenar eminence and the adductor pollicis, though no action could be demonstrated in the interossei.

The restoration of function in the ulnar portion of the profundus digitorum and the flexor carpi ulnaris occurs in approximately 60 percent of cases, following suture. In most cases of ulnar suture in which there has been some degree of axis cylinder regeneration, some return of sensation is found in the ulnar border of the hand and little finger. This sensation, in most instances, after a period of two years is nondiscriminative in character, all types of sensation being described as tingling. Usually after the third year, the power to discriminate between various sensations gradually returns to a greater or lesser degree, but usually sufficient to afford protection to this region of the hand.

In view of these facts, a failure of interossei regeneration should not be considered as defective regeneration beyond our usual experience, and this defect does not call for secondary surgical intervention. The failure of sensory restoration in the ulnar region is indicative of defective regeneration and does not call for intervention. In considering defective regeneration in the ulnar nerve, sensation deserves the greatest attention. Any hope of regeneration, to a satisfactory degree, in the intrinsic hand muscles may be practically abandoned for an uncorrected paralysis of one year's standing. Paralysis of the flexor carpi ulnaris and ulnar half of the profundus digitorum is productive of no great disability and the slight disability existing from their paralysis may readily be overcome by tendon transplantation.

DETERMINATION OF REGENERATION

The examiner must depend entirely upon Tinel's sign for information during the early stages of ulnar regeneration following suture. After six months, if formication is not elicited in the distal segment of the nerve. it should be considered as defective neuraxon regeneration, and secondary surgical intervention is indicated.

Certain precautions must be taken in the presence of much scar tissue in the region of the suture, in eliciting this reaction. The tugging of a scar upon a neuroma or suture line may elicit the reaction of formication in a region below the regenerated nerve fibers. When Tinel's sign is elicited throughout the distal nerve trunk and this reaction is intense, it may be prognosticated that sensory fibers are regenerating satisfactorily, and the patient assured of so men degree of sensory restoration. A persistent loss of sensation in the ulnar region with a sensitive suture line, giving evidence on palpation of neuroma formation, suggests defective suture and secondary exposure is indicated.


1039

The writer has observed scveral instances in which there was no evidence of regeneration following a neurolysis of the ulnar nerve. Such failures, frequently commented upon by other writers, need not, necessarily, call for resection of the nerve and suture, unless an internal neurolysis has demonstrated a complete blocking of the scarred trunk, with obliteration of motor bundles. These defective end results, following neurolysis, are usually due to imperfect surgical judgment at the time of operation, when an external neurolysis was probably the extent of the operation; the real compressive factors evidently were overlooked.

In complete persisting failure of regeneration, and in irreparable defects, where the anesthetic area is constantly being subjected to trauma, and suffering trophic disturbances, amputation of the little finger offers the only solution. Occasionally this amputation must also include the ulnar metacarpal region. A very deforming fibrous griffe may be allowed to develop in the little finger, greatly diminishing the usefulness of the hand, which also may justify amputation.

COMBINED LESIONS OF MEDIAN AND ULNAR NERVES

The close anatomic relationship existing between the median and ulnar nerves in the middle and upper thirds of the arm makes their combined lesions|comparatively frequent. Less often are combined lesions of these nerves experienced below the lower third of the arm, where they pursue a divergent course, though combined lesions are by no means uncommon in extensive gunshot injuries of the forearm. The importance of combined median and ulnar lesions is emphasized by the seriousness of the resulting diasbility, in which all of the extrinsic and intrinsic flexion power of the fingers is lost, as well as the power of pronation; this paralysis, combined with complete anesthesia of the palm and fingers, renders the extremity practically useless.

The surgical repair of combined median and ulnar lesions differs in no respect from the individual repair of these nerves, each of which should be treated surgically as an individual unit. The necessity for a separate consideration of combined median and ulnar lesions is advisable, however, because of supplementary surgical procedures indicated in a total absence of or defective nerve and muscle regeneration.

In the great majority of instances following a successful end-to-end approximation in both median and ulnar nerves (if due respect has been given, the prevention of torsion during suture) there will be a restoration of voluntary motor power in the pronators, wrist flexors, and extrinsic digital flexors, with a gradual return of sensation; the total anesthesia of the fingers is replaced at first by a nondiscriminative type of anesthesia, which later, after two or three years, is slowly supplemented with the power of tactile localization and discrimination. This degree of functional restoration following the suture of these nerves is probably as complete as may reasonably be expected. The residual disability is confined to a loss of function in the intrinsic hand muscles, expressed by: (1) Loss of opponens function of the thumb; (2) inability to flex the metacarpophalangeal joints; (3) loss of extension of the interphalangeal joints with hyperextension of the metacarpophalangeal joints.


1040

These disabilities produce a deformity commonly characterized as the "claw hand" of complete intrinsic hand-muscle paralysis. The long flexors of the fingers are unable to conjointly flex the metacarpophalangeal joints and terminal phalanges; attempts at flexion are followed by a rolling up of the terminal phalanges, in which the nails, instead of the palmar surface of the fingers meet, the palm, due to the fact that flexion of the metacarpophalangeal joints does not occur synchronously with phalanx flexion, and when flexion does occur at the metacarpophalangeal joints it is only after the interphalangeal joints are completely flexed.

The thumb disability consists in the loss of opponens action, and though flexion of the distal phalanx is restored through the long flexor of the thumb, this member rests against the radial border of the hand, and normal thumb and index finger opposition ("pinch action") is replaced by a very defective and ineffectual approximation, in which the nail of the thumb meets the radial side of the index finger, objects being held between the thumb nail and this finger very much as a small boy holds a marble for shooting.

The extensor communis digitorum is capable of extending the terminal phalanges when the long flexors are completely paralyzed, but it seems to lack this power when they are active. Its principal action is extension at the metacarpophalangeal joints. In intrinsic hand-muscle paralysis the fingers can not be completely extended; during attempts at extension the terminal phalanges remain partially flexed, while the metacarpophalangeal joints are hyperextended. The interossei, by virtue of their insertion into the extensor tendons beyond the metacarpophalangeal joints, accomplish terminal phalangeal extension; in this they are assisted by the lumbricales, which also by virtue of their origin from the tendons of the profundus digitorum and their insertion with the interossei into the extensor tendons, near the center of the first phalanx, prevent hyperextension of the metacarpophalangeal joints, as the lumbricales are flexors of these joints and antagonists to the extensor communis digitorum. The function, therefore, which we wish to restore in intrinsic hand-muscle paralysis is the opposing action of the thumb, and flexion of the metacarpophalangeal joints, with extension of the terminal phalanges, metacarpophalangeal flexion is particularly important in the index finger.

For the restoration of opponens function, or rather the maintenance of the thumb in the opponens position, Major Baldwin attempted arthrodesis of the first metacarpophalangeal joint and fixed it in an opposed position at an angle of 60 degrees, which is practically the normal angle existing between the first and second metacarpals during pinching action. This operation, though restoring the opponens position of the thumb and greatly improving its function. has a distinct disadvantage; it does not permit abduction and extension, the thumb remaining continuously opposed over the palm, in which position it is like the drop thumb of extensor paralysis and is frequently in the way.

The writer devised and practiced a tendon transplant for the restoration of opponens function in the thumb, which duplicates this action perfectly, at the same time allowing its complete extension. The procedure consists in directing the short extensor tendon of the thumb through a tunnel under the fat of the


1041

superficial fascia; from the palm it is directed under the annular ligament to the wrist where it is anastomosed to the palmaris longus. The end results of this procedure have been observed in a series of cases for a period of more than three years. In each instance the opposing action of the thumb has remained excellent and has gradually increased in power with the development of the palinaris longus muscle.

TENDON TRANSPLANTATION FOR RESTORING OPPONENS ACTION TO THE THUMB IN INTRINSIC HAND-MUSCLE PARALYSIS

This procedure is conducted entirely under local anesthesia, which assists materially in hemostasis; it also permits the testing out of the transplant during operation, by voluntary contraction of the palmaris longus, the patient having been previously instructed in the contraction of this muscle.

TECHNIQUE

a. A midventral wrist incision is made over the course of the palmaris longus tendon, extending from the base of the palm just below the annular ligament, upward for a distance of about 7 cm. The palmaris longus tendon is identified and freed, and its attachment to the palmar fascia divided, after it has passed over the annular ligament. The median nerve lies immediately below the palmaris tendon and is usually exposed with the retraction of this tendon. A pair of blunt-curved forceps is now passed under the annular ligament at this point, and made to emerge below its palmar border, by penetrating the palmar fascia, the forceps being spread to enlarge the opening. The opening in the palmar fascia is enlarged sufficiently, by excising its edges, to prevent subsequent constriction of the transposed extensor tendon, which is passed through this opening.

b. dorsal thumb incision from the base of the proximal phalanx extends to the radiocarpal articulation, bisecting the "anatomical snuffbox." To the palmar side of this incision the short extensor tendon is located as it lies in a separate sheath in juxtaposition with the dorsal border of the tendon of the extensor ossis metacarpi pollicis. The long extensor tendon, which lies to the dorsal side of the incision, must not be confused with the short extensor tendon-traction on the tendon of the long extensor produces extension of the distal phalanx. The sheath of the short flexor is opened and the tendon followed downward to the metacarpophalangeal joint of the thumb. This short flexor tendon is divided about 10 cm. above its insertion, where it lies in close ipproximation with the tendon of the extensor ossis.

c. A tunnel is now maed between the lower ends of the two incisions, undermining the fat of the superficial fascia in a line connecting the annular ligament and the metacarpophalangeal joint of the thumb. Through this tunnel a pair of forceps is passed from the ventral to the dorsal incision and the divided end of the extensor brevis pollicis tendon caught and drawn through the tunnel to the opening in the palmar fascia. Its end is then grasped by the previously placed subannular forceps, drawn through the opening in the


1042

palmar fascia under the annular ligament to the wrist. where it is anastomosed to the palmaris longus tendon, with sufficient tension to produce marked abduction and opposing rotation of the thumb. When the palmaris longus is absent, as it is in approximately 20 per cent of cases, the transposed extensor tendon may be passed through a slit in the flexor carpi radialis tendon

FIG. 187.- Tendon transplant for restoring opponens position and function to the thumb in intrinsic hand muscle paralysis; exposure of palmaris longus tendon

FIG. 188.- Tendon transplant for restoring opponent position and function to the thumb in intricate hand muscle paralysis. Tendon of palmaris longus divided; opening made in palmar fascia at upper part of annular ligament.

and sutured, without division of the tendon of the latter muscle. (In the writer's experience, more satisfactory end results have followed transplantation into the palmaris longus.) The thumb now lies across the palm in an opposed and abducted position and with the wrist slightly flexed: the skin incisions are closed, and the hand dressed with the thumb and fingers grasping an unrolled


1043

bandage, which tends to maintain the thumb in the desired position. The fist, with the wrist flexed, is firmly supported with suitable bandages or splinting to insure the maintenance of this position.

The subsequent treatment is conducted along lines common to the after-treatment of all tendon transplantations. Reeducation is, as a rule, not

FIG. l89.- Tendon transplant for restoring opponens position and function to the thumb in intrinsic hand muscle paralysis; exposure of short extensor of the thumb

FIG. 190.- Tendon transplant for restoring opponens position and function to the thumb intrinsic hand muscle paralysis. Passage of the divided short extensor tendon through a subcutaneous tunnel in the thenar eminence and through opening in palmar fascia, where it passes below the annular ligament and is anastomosed to the palmaris longus tendon, holding the thumb in the opponens position

difficult, and a few demonstrations will usually be found sufficient to institute effective opponens action of the thumb in individuals of ordinary intelligence. This transplant is synergetic, in that forceful prehension of the thumb is accompanied normally by contraction of the palmaris longus.


1044

TENDON TRANSPLANTATION FOR RESTORING METACARPOPHALANGEAL FLEXION AND EXTENSION OF THE TERMINAL PHALANGES (LUMBRICALES AND INTEROSSEI FUNCTION)

For the restoration of inetacarpoplhalangeal flexion, Stiles devised an ingenious operation in which he alters the insertion of the flexor sublimis digitorum tendons, rendering them flexors of the metacarpophalangeal joints by suturing them to the exterior communis tendons distal to the knuckle, which makes them also extensors of the interphalangeal joints, thus causing the flexor sublimis digitorum to subserve the function of both interossei and lumbricales. This procedure may be carried out on each finger or it may be limited to the index.

TECHNIQUE

a. Through a mid-dorsal incision, the skin and fascia are divided, and the common extensor exposed. The insertions of the interossei and lumbricales into its lateral aspect are identified and freed.
b. An incision is made along the flexor aspect of the digit from the transverse palmar crease to the distal interphalangeal joint, exposing the flexor sheath, which must not be opened through the line of incision. About the base of the metacarpophalangeal joint, the sublimis tendon divides to permit the passage of the profundus tendon. The lateral and medial slips of the tendon then pass forward to be attached into the sides of the middle phalanx. Near their insertion a small incision is made through the sheath of each tendon slip anl the tendon divided. Next the tendon is identified opposite the metacarpophalangeal joint, where it has just begun to split; an opening is made in the sheath at this point, and each tendon slip pulled out of its sheath. Stiles emphasizes the importance, in making these openings in the sheath, of not disturbing the small bands which hold the profundus tendon in place opposite each phalanx, "otherwise the tendon will stand forward under the skin like a bowstring when the patient attempts to flex the finger, and will lose its pull on the terminal phalanx." Each half of the sublimis tendon is now passed through a subcutaneous tunnel on each side of the digit to the dorsal incision and threaded through an opening made in the expanded portion of the extensor sheath which receives the insertion of the interossei and lumbricales. The two tendons are sutured in this position with linen, with the knuckle flexed at right angles and the interphalangeal joints straight. Any excess of tendon is clit away alud the raw ends buried by a suture through the two sides of the coinniunis. Both wounds are closed and the fingers fixed by pads and bandages in a position of metacarpophalangeal flexion and terminal phalanx extension.

IRREPARABLE DEFECTS OF THE MEDIAN AND ULNAR NERVES

An irreparable defect of the median and ulnar nerves is attended with such a degree of disability to the extremity that it may be considered as permanently useless, unless some flexion function can be restored to the fingers through tendon transplantation. The accompanying anesthesia, however, seriously diminishes the functional usefulness of the hand, even though digital flexion is regained through tendon transplantation. Before resignation is


1045

FIG. 191.- Diagrammatic explanation of viable neuroplastic transplant for filling of median defect in irreparable lesion of both median and ulnar nerves. A Showing large irreparable defect of both median and ulnuir nerves: the upper segments of both nerves are marked with identification sutures. B, Proximal end of both nerves united by end-to-end suture. Ulnar nerve divided to permit degeneration of fibers, the central end of which is injected with alcohol to prevent regeneration of ulnar fibers in that portion of ulnar trunk which is to be used later as a viable transplant. C, Second stage of operation, showing transplant which now contains regenerated median fibers turned down and approximated to the end of the distal segment of the median, overcoming the defect, the ulnar nerve being sacrificed to repair the median


1046

made to an irreparable median and ulnar defect (the two-stage operation for nerve repair and grafts having failed), the surgeon should consider the possibility of preserving some type of sensation in the fingers by nerve anastomosis. Harris suggested anastomosing the radial nerve to the distal stump of the median. This procedure may be followed by some restoration of sensation in the median sensory area of the hand and fingers; this sensation, however, is attended with dorsal localization, and although it protects the fingers by replacing anesthesia with a defectively localized sensibility, it contributes little to functional usefulness.

SACRIFICE OF THE ULNAR NERVE AS A VIABLE NEUROPLASTIC TRANSPLANT FOR THE REPAIR OF A MEDIAN DEFECT

When all hope has been definitely abandoned of effecting approximation in both median and ulnar nerves, the disability may be diminished by using the ulnar nerve as a viable neuroplastic transplant for bridging the defect in the median, in the hope of restoring at least median sensation, and possibly function, in the extrinsic muscles of the hand. If this motor restoration is possible, opponens function and metacarpophalangeal flexion may be restored by tendon transplantation, and a very useful hand provided; if it is found impossible to restore motor function, flexion of the digits may be restored to some degree by tendon transplantation and some usefulness, at least, regained.

TECHNIQUE
           
First stage.-The proximal ends of both median and ulnar nerves are exposed, the scar tissue and neuroma resected, after which they are carefully approximated, following the usual technique of end-to-end suture, extreme care being taken to prevent torsion of the nerve trunks in this approximation. This procedure results in the formation, after anastomosis, of a loop uniting the proximal ends of these nerves. Without disturbing the bed of the ulnar nerve any more than necessary, it is exposed some distance above the anastomosis and divided to permit Wallerian degeneration of the nerve fibers in that portion of the ulnar trunk which at the second operation will be turned down to meet the distal end of the median as a transplant. This transplant should be slightly longer than the defect--if the defect be 5 inches, the ulnar nerve should be divided 5½ or 6 inches above its anastomosis to the median. The upper end of the divided ulnar trunk is injected with alcohol or treated after the manner of amputation neuromas, to prevent regeneration of ulnar fibers down the transplant. It is essential in this procedure to conserve in every way the nutrition of that portion of the ulnar trunk which is being used as a transplant; hence the necessity of not disturbing its bed. The median fibers, in regenerating, will now pass around the loop and follow the transplant upward, and if sufficient time be allowed they will form a neuroma at the upper end of the transplant where the ulnar nerve was divided. Inasmuch as regeneration occurs, approximately, at the rate of 1 inch per month, the second operation should be postponed until the median fibers have completely traversed the transplant; in a 5-inch transplant the second operation should be planned at the end of the sixth month, allowing a month of grace.


1047

Second stage.- Both proximal and distal segments of the median nerve are now exposed; in the lower segment the center of the ventral quadrant is marked by identification sutures before disturbing its relations; in the upper segment the loop of anastomosis between the median nerve and the transplant is now carefully exposed and the transplant followed upward to its end, where identification sutures are placed in the center of its exposed ventral quadrant. The transplant is freed by careful dissection and turned down to meet the end of the distal segment of the median; the scar tissue is now resected from both ends of the nerve and approximation effected, with strict care toward the prevention of torsion. The selection of a satisfactory bed for the transplant is important.

In the above procedure the defect in the median nerve is filled by a viable transplant through which median fibers have regenerated and is far more likely to prove successful than nonviable grafts.

When the ulnar nerve has been sacrificed and the ulnar portion of the hand and little finger is the subject of traumatic and trophic sores amputation is indicated. If, however, the skin of the ulnar portion of the hand remains in good condition through proper protection, restoration of sensation can be obtained in certain instances by anastomosing the radial nerve to the ulnar.

TENDON TRANSPLANTATION FOR COMPLETE FLEXOR PARALYSIS

If passive movements of the wrist and fingers are unrestricted, tendon transplantation may be utilized to restore, to some extent, digital flexion; though flexion may be restored to a certain degree through transplantation of extensors, the loss of the intrinsic hand muscles greatly limits the usefulness of the hand beyond the ability to hold or carry objects.

TECHNIQUE

 a. Splitting the tendon of the extensor carpi radialis longior, one slip of which is transplanted into the sublimis and profundus tendons of the index finger, a second slip is transplanted into the same tendons of the ring finger-through a long incision, extending from the base of the thumb to the upper third of the dorsum of the forearm, the long ribbonlike tendon of the extensor carpi radialis longior is identified between the extensor brevis, to its ulnar side; and the brachioradialis, to the radial side. At the wrist, these tendons are crossed by the bellies of the extensor ossis metacarpi pollicis, extensor brevis pohlicis, and the tendon of the long flexor of the thumb. The tendon of the long radial extensor is divided at its insertion into the second metacarpus. Its sheath is opened high up and the tendon withdrawn and split. The tendon of the brachioradialis is now an identified and divided, and its borders freed from fascial attachments some distance up the forearm; in mobilizing the brachioradialis caution should be used in the isolation of the radial nerve and artery, which lie on its under surface. Several muscular branches of the radial artery will require ligation.

b. Transplantation of the brachioradialis tendon into both flexor tendons of the ring and little fingers--a ventral incision is now made, extending from the base of the thumb, near the insertion of the flexor carpi radialis tendon, along the radial border of the flexor surface of the forearm to the junction of its upper


1048

and middle thirds. The skin and fat of the superficial fascia is undermined between the two incisions, permitting the passage of the extensor and brachioradialis tendons to the ventral aspect of the forearm in a straight couse. Medial retraction of the ventral incision permits exposure of the long flexor tendons, each of which should be identified and arranged in pairs, sublimis and profundus for each finger. T

The long flexor of the thumb is identified and separately isolated. With the fingers completely flexed, the lateral slip of the divided long extensor tendon is passed through a slit in the sublimis and profundus flexors of the index finger; the medial half of the radial extensor tendon is anatomosed into both flexor tendons of the ring finger. The tendon of the brachioradial is is now brought forward in as direct a line as possible and inserted in alike manner into the flexor tendons of the ring and little fingers.

c. Transplanting extensor ossis metacarpi pollicis into the flexor lungus pollicis---the extensor ossis metacarpi pollicis tendon is divided at its insertion and transplanted into the long flexor of the thumb, with the thumb flexed over the previously flexed fingers, making a fist. After the proper tension has been placed on these transplanted tendons, so that the fingers are flexed equally, and with the wrist in partial flexion, the tendons are anchored.

Both dorsal and ventral skin incisions are now closed while the hand is maintained in complete digital and partial wrist flexion. This position is maintained by suitable pads and bandages, after a small unrolled bandage placed in the hand. (After-treatment and reeducation are conducted along the usual lines.)

THE SCIATIC TRUNK AND ITS TERMINAL DIVISIONS
 
GENERAL ANATOMY  

The sciatic trunk is formed by the ventral and dorsal divisions of the fourth and fifth lumbar, and second and third sacral nerves. The ventral divisions go to make up the medial or tibial portion of the sciatic trunk, while the dorsal form its lateral or peroneal portion.

The tibial and peroneal portions of the sciatic trunk are from a physiologic standpoint totally differentiated, though anatomically they are usually incorporated into a single nerve trunk for some distance down the thigh. The extent of their union, however, varies, and occasionally the sciatic trunk is replaced by two distinct nerves, emerging from the great sacrosciatic foramen. More often however, their union continues to the upper part of the popliteal space, where the sciatic trunk divides into its terminal divisions; the medial or tibial portia becoming the internal popliteal nerve, while its lateral or peroneal forms the external popliteal nerve. The separation in the thigh of the component tibial and peroneal portions of the sciatic trunk is always more or less evident, varying from a complete separation, in which each nerve has a distinct and separate sheath, to the type in which inspection alone fails to demonstrate the anatomical division and in which palpation is required to give the surgeon a clue as to the line of cleavage. These nerves, when combined in a single sheath, are anatomically separated by a septal prolongation of the sheath, which is usually evident on cross section. As the trunk proceeds down the thigh, the demarcation between its component parts becomes more conspicuous on its surface, and it dividing septal sheath shows greater development.


1049

Though the tihial portion of the sciatic trunk is usually described as lying medial, and the peroneal portion lateral, the line of division does not strictly follow a ventrodorsal plane; the tibial portion lies somewhat more ventral than dorsal, while the peroneal portion occupies a dorsolateral position.

After emerging from the pelvis to the dorsum of the thigh, through the great sacrosciatic foramen under the pyriformis muscle, the sciatic trunk passes down the posterior aspect of the thigh to the popliteal space. In its upper part it lies on the external rotators of the thigh, and below these upon the dorsal surface of the adductor magnus. In its gluteal portion, the nerve is covered by the gluteus maximus; in the thigh it runs parallel with and to the lateral side of the flattened tendon of the semimembranosus, and is crossed and covered in most of this portion by the thick belly of the ischial head of the biceps.

NERVE TO HAMSTRING MUSCLES

The nerve to the hamstring muscles, while commonly incorporated within the sheath of the tibial portion of the sciatic trunk, should not be regarded as a collateral branch of the tibial nerve, but rather as one of the three separate elements, which, being bound together by a common sheath, constitute the great sciatic trunk, namely, from within outward, nerve to the hamstrings, tibial nerve, and peroneal nerve. The nerve to the hamstrings may or may not be incorporated within the sheath of the sciatic trunk, but always lies to the medial side of the tibial portion of the sciatic; when incorporated within its sheath it is usually anatomically separated by a well-developed sheath of its own and may be readily separated from the sciatic trunk. The short or humeral head of the biceps, however, is supplied about, the middle of the thigh a branch which leaves the lateral surface of the peroneal portion of the sciatic. This branch has a distinct intraneural course and may or may not he regarded as a collateral branch of the peroneal nerve. The nerve to the hamstrings. applies the semitendinosus, semimembranosus, long head of the biceps, and a portion of the adductor magnus.

TERMINAL BRANCHES

The tibial nerve (internal popliteal), which formerly occupied a ventro-medial position, upon leaving the sciatic trunk in the lower third of the thigh. Enters the popliteal space, where it lies in a pad of fat, superficial to the popliteal vessels. It passes longitudinally through the middle of the popliteal space, from which it emerges by passing ventral to the union of the two heads of the gastrocnemius. It terminates at the lower border of the popliteus muscle, by passing through an arch in the soleus muscle, accompanied by the popliteal artery and vein, where it becomes the posterior tibial nerve.

In the popliteal space, motor branches arise to supply the gastrocnemius. plantaris, soleus, and popliteus muscles: these branches are given off from the dorsal surface of the nerve. At the upper part of the popliteal space, a large branch diverges from each side of the nerve trunk to the two heads of the gastrocnemius. Though usually given off as separate branches, their intraneural origin is from a bundle common to all of the above-mentioned muscles: it has a rather long intraneural course, lying on the dorsal aspect of this portion


1050

of the sciatic trunk. Occasionally, the nerve to the soleus leaves the parent trunk in common with the branch to the lateral head of the gastrocnemius. The branches to the plantaris and popliteus are usually given off individually but their intraneural course is common with the gastrocnemius and soleus fibers. Below the branches to the gastrocnemius, a sensory branch is given off, the communicans tibialis, which passes downward in the sulcus between the two heads of the gastrocnemius, becoming superficial in the calf by piercing the deep fascia where it joins the peroneal communicating to form the external or short saphenous nerve, which supplies sensation to the lower third of the leg on its outer side, and the outer side of the foot and little toe. The tibial nerve also gives off branches to the knee joint and to the posterior tibial vessels.
           
The posterior tibial nerve
, the terminal portion in the leg of the tibial (internal popliteal) nerve, passes down the leg, occupying a deep position between the deep and superficial muscles of the calf, in a special compartment in the intermuscular septum between these muscles, accompanied by the posterior tibial artery and its venve comites, in relation to which it holds a superficial or dorsal position. In the lower part of the leg, it lies medial and ventral to the tendo Achillis; after passing behind the internal malleolus it enters the sole of the foot, to terminate as the internal and external plantar nerves. Immediately after the posterior tibial nerve is formed by passing through the arch of the soleus muscle, it breaks up into a number of branches which supply the lower portion of the soleus, the tibialis posticus, the flexor longus digitorum and the flexor longus hallucis.

The peroneal nerve (external popliteal), considerably smaller than the tibial nerve, after leaving the common sciatic trunk, at the apex of the popliteal space, follows the lateral border of the fossa and the medial margin of the biceps tendon. Passing over the plantaris and lateral head of the gastrocnemius. it winds around the neck of the fibula, where it enters a canal in the origin of the peroneus longus muscle, and breaks up into its terminal branches.

In the popliteal space, the peroneal nerve gives off the following collateral sensory branches, from the dorsal portion of its trunk: The communicating peroneal which passes medially across the lateral head of the gastroenemius.pierces the deep fascia on the back part of the leg and unites with the communicating tibial to form the short or external saphenous nerve; the sural branches, which may arise in common with the communicating peroneal or at a lower origin to supply the skin over the back and outer side of the calf and leg. Intraneurally, the fibers forming these nerves originate from a single bundle which occupies a long intraneural course on the dorsal surface of the peroneal portion of the sciatic trunk.

The terminal branches of the peroneal nerve are given off after it swing around the neck of the fibula and passes through a canal in the peroneus longus muscle. They are the musculocutaneous, anterior tibial, and tibial recurrent. The anterior tibial branch springs from the lower portion of the flattened perneal trunk and has a long intraneural course, composed of a single bundle which in the upper part of the popliteal fossa lies in the ventral portion of the nerve. The musculocutaneous branch arises from the ventral portion of the trunk and likewise has a long, intraneural course, though composed of two bundles, one


1051

of which contains motor fibers to the peronei muscles, the other containing the sensory fibers.            

The musculocutaneous nerve
, lying at first between the peroneus longus muscle and fibula, passes obliquely downward and forward; it is deeply placed and located in a fibrous canal in the septum between the peronei and extensor longis digitorum; it supplies the peroneus longus and brevis with motor branches. In the middle of the leg it penetrates the deep fascia and divides into an internal and external branch, supplying the skin of the front of the leg and dorsum of the foot and toes.
           
The anterior tibial nerve
, originating from the pleroneal between the recurrent tibial and musculocutaneous, passes downward and forward beneath the extensor longus digitorum to the ventral surface of the interosseus membrane, where it passes down the leg accompanied by the anterior tibial artery and its venae comites, lying between the tibialis anticus and the extensor longus hallucis. At the ankle it is crossed by the tendon of the extensor longus hallucis, as it passes to the dorsum of the foot, where it divides into its external and internal terminal branches. The anterior tibial nerve supplies motor fibers to the tibialis anticus, extensor longus hallucis, extensor longus digitorum and peroneus tertius. Its external branch supplies the extensor brevis digitorum, and its internal branch the integument between the first and second toe and a small area on the dorsutm of the foot adjacent to these toes.

The tibial recurrent nerve, the smallest, most anterior and highest of the three terminal branches of the tibial nerve, passes between the origin of the peroneus longus and the fibula, giving motor branches to the upper portion of the tibialis anticus and articular branches to the knee joint.

SURGERY OF THE SCIATIC TRUNK

In war surgery injuries of the sciatic trunk are far more common than any other nerve lesion in the lower extremity. While severance of the entire trunk does occasionally occur, this incident is rare. The nerve to the hamstrings seldom shows complete paralysis, and very frequently the peroneal portion of the sciatic trunk alone suffers injury. This is probably due to the position of the peroneal in the sciatic trunk, where it occupies a dorsolateral position and tends to protect the tibial portion of the nerve. The components of the sciatic trunk may suffer injury in any part of their course in the buttocks, thigh. popliteal space, or leg. Throughout its entire course the tibial nerve is afforded greater protection than the peroneal. The peroneal is particularly liable to injury in the region of the head of the fibula, where it occupies a very superficial position.

The surgery of the sciatic trunk resolves itself into the surgery of its component elements, and not infrequently exposure will reveal the entire trunk extensivelv involved in scar tissue when clinically the lesion is confined only to its peroneal portion. The surgeon must therefore be governed by clinical indications rather than by the gross pathologic appearance of the nerve at operation, and thereby avoid the sacrifice of intact sciatic elements. Surgery of the sciatic trunk resolves itself into the exposure of three regions, namely: For gluteal lesions, upper thigh lesions, and lower thigh lesions. In


1052

the gluteal region the nerve is exposed by reflecting the gluteus maximus; in the upper thigh, by retracting the ischial head of the biceps medially; while in lower thigh lesions this muscle is retracted laterally. It is essential therefore that the surgeon be familiar with the technique involved in exposing the sciatic trunk in these regions; in addition to this, he must be able to appreciate and differentiate the component parts of the trunk, and, in suture, to effect perfect matching and alignment of its respective elements. The approximation of the peroneal to the tibial portion through torsion of the trunk would obviously be a catastrophe, and though neuraxon regeneration might progress unimpeded, the disturbance of physiologic pattern would give very confusing and unsatisfactory results. The sciatic trunk must be considered not as an individual nerve but as three distinct elements closely approximated in a common sheath, namely, from within outward, the nerve to the hamstrings, the tibial, and the peroneal. The nerve to the hamstrings seldom requires surgical intervention, while the two remaining elements, the tibial and peroneal portions, are commonly involved, though the two latter are so separated by a septal prolongation of the sciatic sheath that identification is rendered possible and perfect alignment facilitated by observing and maintaining the position of this septum in suture. In this respect, the sciatic trunk offers in the majority of cases a most satisfactory means of effecting physiologic approximation.

EXPOSURE IN GLUTEAL REGION

Exposure of the sciatic trunk in the gluteal region is most satisfactorily accomplished and less trauma inflicted by dividing the tendon of the gluteus maximus at its insertion into the femur and reflecting it medially. This exposure not only assists in preserving the nerve supply to the gluteus maximus, but also is effective in avoiding the troublesome hemorrhage associated with across section of its fibers. The line of incision is curved, with its base outward, and extends from the posterior iliac spine outward and downward to the upper border of the great trochanter of the femur, and, slightly posterior, it passes downward over the insertion of the gluteus maximus tendon; it then curves backward, following the gluteal fold to the midline of the thigh, down which it may be extended over the course of the sciatic nerve, as far as the individual exposure necessitates. The deep fascia is exposed and divided in the line of incision. Along the upper border of the incision the fibers of the gluteus maximus are split in their line of cleavage and the tendon exposed and divided close to its insertion in the femur, permitting a medial reflection of the lower half of this muscle. In reflecting the gluteus maximus, care should be used to prevent injury to branches of the inferior gluteal vessels. To avoid obscuring the field, these vessels, when possible, should be caught and divided between forceps. The sciatic trunk is now exposed as it emerges from the great sacrosciatic foramen below the pyriformis muscle, and again care must be used to prevent injury to the inferior gluteal nerve which emerges from under the pyriformis along the dorsal surface of the sciatic trunk, accompanied by branches of the inferior gluteal artery and veins. To the medial side of the great sciatic trunk lies the small sciatic nerve, which may be differentiated from the nerve to the hamstrings by following its course to the lower border


1053

of the gluteus maximus, where it becomes superficial. The nerve to the ham-strings in this region may be either incorporated within the sheath of the sciatic trunk or it may pass along its medial side as a separate nerve. Electrical stimulation will readily serve to identify this motor branch when doubt exists. In this region the sciatic trunk is surrounded by many vascular branches, necessitating great caution; the reckless or blind application of a hemostat in this region to control an unseen bleeder may result in injury to important motor branches. The sciatic trunk may be followed beneath the pyriformis muscle, and for a short distance within the foramen, by upward retraction of the pyriformis. Even better exposure may be had by dividing its tendon and reflecting the muscle medially. After all vessels have been carefully ligated to avoid, as much as possible, the subsequent oozing. the tendon of the gluteus maximus is sutured with strong linen. Drainage should always be instituted for a period of 48 hours to care for oozing, which is usually abundant following this exposure.

EXPOSURE IN THE THIGH

As the sciatic trunk emerges from below the lower border of the gluteus maximus, it passes down the middle of the dorsum of the thigh,

FIG. 192.- Exposure of sciatic trunk and branches to the hamstrings resting upon the adductor in gluteal region; insertion of gluteus maximus divided and the magnus, in close proximity to muscle reflected, exposing the sciatic nerve emerging from the greater sacrosciatic foramen below the pyriformis muscle. A, the ribbon-like tendon of the Gluteus maximus, lower portion reflected; B, nerve to hamstrings: C, semimembranosus tendon; D, biceps, long head; E, pyriformis; F, sciatic trunk; G, adductor magnus

resting upon the adductor magnus, in close proximity to the ribbon-like tendon of the semimembranosus.  The nerve trunk is fairly superficial at the lower border of the gluteus maximus and at the popliteal space, where it is covered only by an abundance of surrounding fat and the deep fascia. In the middle of the thigh, however, the sciatic trunk is diagonally crossed by the ischial head of the biceps from within outward, the nerve being deeply placed beneath the fleshy belly of this muscle.


1054

Exposure of the sciatic trunk in the thigh is accomplished through a mid-dorsal longitudinal incision, extending from a few centimeters above the gluteal fold to the popliteal space. The small sciatic nerve lies in close relation to the deep fascia in this region and should be identified and preserved if possible. While it is purely sensory and not of great importance, its identification and retraction will serve to prevent irritative symptoms, following the accidental strangulation by suture, during closure of the deep fascia. If the lesion be an upper thigh lesion, the lateral border of the ischial head of the biceps is identified and retracted medially, exposing the flattened tendon of the semi-membranosus, which must not be mistaken for the sciatic trunk, lying immediately along its lateral border, surrounded by an abundance of fat. The lower border of the gluteus maximus is undermined and retracted upward to afford higher exposure and if necessary a few of its fiber may be divided. In lower thigh lesions, the medial border of the ischial bead of the biceps is identified and retracted laterally exposing the nerve below. Medial or lateral retraction of the ischial head of the biceps will usually permit

FIG. 193.- Exposure of the sciatic in the middle and lower thirds of the a satisfactory exposure of thigh by lateral retraction of the short head of the biceps. Branches to the the sciatic trunk in any short head of biceps is shown emerging from the lateral side of the peroneal division of the sciatic trunk about the middle of the thigh. A, Semi-membranosus; B, tibial nerve; C, biceps, long head; D, branch to biceps, gluteus maximus. After short head; E, peroneal nerve; F, sciatic trunk

a satisfactory exposure of the sciatic trunk in any part of its course below the gluteus maximus.  After identifying the nerve trunk,  attention must be given to its component parts. The nerves to the ham-strings, passing along its medial border, must be identified and carefully retracted medially to prevent their injury during dissection of the trunk The sciatic trunk in the thigh may present itself as a single large nerve, or the division of its component tibial and peroneal portions may be conspicuous by their complete separation, or by markings on the sheath of the trunk along their line of cleavage. If inspection of the trunk fails to reveal


1055

any line of demarcation between its respective tibial and peroneal portions, palpation, or rolling the nerve between the fingers, will usually serve in effecting this identification. It should be remembered that the component portions of the sciatic trunk are separated by a more or less well-defined septal prolongation of its sheath; that its tibial elements occupy a medial position, somewhat more ventral than dorsal: that the peroneal portion , somewhat more ventral than dorsal; that the peroneal portion, which is the smaller of the 

FIG. 194.– Diagrammatic cross section of sciatic trunk, showing its tibial and peroneal components. The septal division between these portions of thee sciatic trunk does not, lie in a strictly ventro-dorsal plane. The peroneal or smaller division of the trunk occupies a ventro-medial position. A, A type of sciatic nerve trunk in which the line of division between its component parts is evident by a groove on the surface of the nerve. B, Type in which separation of the component parts is not so evident. (The above diagram represents the right sciatic trunk as exposed through a dorsal incision; the upper part of the diagram represents the dorsal surface of the sciatic trunk)

two, is located in the dorsolateral aspect of the trunk. The line of cleavage, therefore, will not be found in a strictly ventrodorsal plane, but passing from the center of its dorsomedial quadrant to its ventrolateral quadranit. Rotation of the sciatic trunk outward will serve to rotate the plane of the intraneural septum to a ventrodorsal position. Occasionally, it will be necessary to find the line of cleavage between the peroneal and tibial portions at their division in the popliteal space, from which point they may be carefully separated up

FIG. 195.- Method of alignment in physiologic approximation of the sciatic trunk, the intraneural septum between the peroneal and tibial porttions of the trunk serving as a guide to alignment

the sciatic trunk. In effecting such a separation, the septal division should be closelv followed. There is no intercommunication of fibers between the component elements of the sciatic trunk, and this separation mav be accomplished, if executed with sufficient care, without destruction to nerve elements. The septum dividing the sciatic trunk into its component portions serves admiraly as a guide in effecting physiologic alignment during suture. Identification sutures in the sheath of the nerve should be placed in a position to assure per-


1056

feet alignment of the intraneural septum, and if it becomes necessary to disturb the relationship between one portion of thle trunk andi this septum, surface markings should be so placed that identitication of this region is assured before
the original relationship has been disturbed.

EXPOSURE OF THE TIBIAL NERVE (INTERNAL POPLITEAL) IN THE POPLITEAL SPACE

The skin incision extends longitudinally through the middle of the popliteal space, from a few centimeters above its apex to a point, a few centimeters below the union of the two heads of the gastrocnemius. The deep fascia is divided, exposing the fat of the popliteal fossa. The tibial nerve should be identified as it emerges at the upper part of the fossa from below the medial border of the biceps; its distal portion is identified as it passes below the union of the two heads of the gastrocnemius. Having located the nerve at these two points, it may be followed through the popliteal space by dividing the overlying fatty tissue. Several vascular branches will be met and divided between clamps to prevent obscuring the field. A careful exposure in this way will usually avoid injury to the popliteal vessels. The small saphenous vein joins the popliteal vein about the middle of the fossa and will require ligation. Near the apex of the popliteal space the tibial nerve gives of diverging motor branches to the two heads of the gastrocnemius. These

FIG. 196.- Exposure of the external and internal popliteal nerves in the popliteal space. A,Tibial nerve; B, internal hamstring tendon; C, branch to gastrocnemius, medial head; D, branch to gastrocnemius, lateral head; E, branch to soleus; F. Gastrocnemius, medial head; G, tibial communicating nerve; H, lateral hamstring tendon; I, peroneal nerve; J, peroneal communicating; K, sural nerve; L, gastrocnemius, lateral head
.
branches originate from a common bundle on the dorsal surface of the nerve, which by intraneural dissection may be followed  some originate from a common bundle on the dorsal surface of the nerve, which by intraneural dissection may be followed some distance up this portion of the sciatic trunk. These branches by their intraneural localization identify the dorsal surface of the tibial nerve. Occasionally, one of the branches of the gastroenemius will


1057

carry fibers to the superficial fibers of the soleus: at other times the branch to the soleus is given off from the tibial trunk at a lower level. Intraneural dissection reveals this branch, however, it may reach the soleus, as Springing from the common gastroenemius bundle. A small branch to the plantaris may be given off directly from the tibial trunk or from the branch to the lateral head of the gastrocnemius. After the nerve has passed below the gastrocnemius, a small branch is given to the popliteus muscle. Shortly after the divergence of the motor nerves to the two heads of the gastrocnemius, a sensory branch, the communicating tibial, leaves the tibial trunk and passes in the sulcus between the two heads of the gastrocnemius to pierce the deep fascia and enter into the formation of the external saphenous. Branch identification in the popliteal space should at least include the branches to the gastrocnemius and soleus, and the tibial communicating. The identification of each is simplified by its course, the nerves to the gastrocnemius diverging respectively to its medial and lateral heads; the communicating tibial by its superficial course. If access to the tibial nerve is desired below the popliteal fossil, the line of cleavage between the two heads of the gastrocnemius may be split as far down this muscle as necessary without disturbing its nerve supply. With the retraction of the lateral and medial head of the gastrocnemius, the tibial nerve may be readily followed to the point where it disappears under the tendinous arch of the soleus, accompanied by the popliteal vessels.

SURGERY OF TIHE POSTERIOR TIBIAL NERVE
          
A certain number of lesions of the posterior tibial nerve are over looked, because patients frequently ignore the intrinsic paralysis and anesthesia of the sole of the foot. There are, however, certain lesions which are associated with marked irritative symptoms referred to the sole of the foot, and these, in many instances, are peculiarly resistant to the moderating effect of time and sedative treatment, and occasionally require neurolysis, alcohol injection, or even excision of the posterior tibial nerve. The writer has observed one instance in which the painful svndrome in the sole of the foot persisted after a neurolysis of the posterior tibial nerve and a subsequent alcohol injection, and was later relieved by a decortication of the popliteal artery. In the lower part of the popliteal space several branches pass from the tibial trunk to the popliteal vessels: these branches follow the vessels through the canal in the soleus and are probably active to some degree in the production of irritative phenomena. In effecting a decortication of the popliteal artery the denudation should proceed to the opening in the soleus muscle by splitting and retracting the heads of the gastrocnemius.            
           
As the tibial nerve passl's throughl the tendinous arch of the soleus to become the posterior tibial, it divides into two branches, the dorsal of which immediately breaks up into at number of branches to supply the ventral surface of the soleus, tibialis posticus, flexor longus digitorum, and flexor longus hallucis. The other branch of the nerve passes down the leg in a compartment formed by the intermuscular septum, separating the superficial from the deep calf muscles, accompanied by the posterior tibial artery anol its vena, comites. In the lower third of the calf it is more superficial, lying to the medial side of


1058

the tendo Achillis and covered by the deep fascia. In this region the nerve is composed of two bundles, which ultimately go to the formation of its terminal internal and external plantar branches.

Exposure of the posterior tibial nerve in its deep position in the upper two-thirds of the calf is best accomplished through a mid-dorsal incision, extending from the middle of the popliteal space to the lower third of the leg, where it curves somewhat to the medial border of the tendo Achillis. In exposing the deep fascia care should be used to avoid injury to the tibial communicating branch of the tibial nerve, which unites in the middle of the calf with the peroneal communicating branch to form the external saphenous nerve. Injury to this branch, which is sensor! in function, is productive of but little disturbance, but when a lesion is already associated with a certain amount of anesthesia it is not wise to easily sacrifice other sensors branches. Identification and isolation of this nerve is also important, in that with closure of the deep fascia it mav be caught in the sutures, resulting in considerable pain and possible irritative phenomena. The short saphenous vein is also encountered and should be ligated to prevent tearing and troublesome hemorrhage. The incision is now carried through the deep fascia, and the gastrocnemius bisected, from the point of union between its two heads to the tendo Achillis and widely retracted, exposing the soleus muscle below. The soleus muscle is now bisected in the same way exposing the intermuscular septum below. It will be remembered in splitting the soleus muscle that it has an aponeurosis covering its dorsal surface, which must not be mistaken for the underlying intermuscular septum. Both gastrocnenmius and soleus, now having been longitudinally bisected from their origin to the tendo, Achillis are retracted, exposing the deeply placed intermuscular septum, separating the superficial muscles from the deep. The posterior tieial nerve and its accompanying artery and vein will be found by splitting the fascia along the medial border of the flexor longus hallucis muscle. The close proximity of the posterior tibial nerve to its accompanying arterial structures renders its dissection rather tedious, because of the frequent tearing of veins. The simplest procedure to follow is: Ligation or clamping of veins in the lowest part of its course; in the upper part, compression of the posterior tibial artery, thus greatly facilitating isolation of the nerve.

In the lower third of the leg the tibial nerve may be located along the medical border of the tendo Achillis by dividing the deep fascia, extending from the tendo Achilles to the deep muscles. The nerve will be found lying between the flexor longuts digitorum and the flexor longus hallucis.

EXPOSURE OF THE PERINEAL NERVE (EXTERNAL POPLATEAL) IN THE POPLITEAL SPACE

The incision extends from the apex of the popliteal fossa, following the medial border of the external hamstring tendon; it passes around the neck of the fibula, and, for a short distance down the anterolateral surface of the leg midway between the crest of the tibia and fibula. The deep fascia is divided in the line of the incision; as it approaches the head of the fibula, this fascis


1059

will be found to be greatly thickened and great care should be used in its division to prevent injury to the nerve, which lies immediately under its deep surface and to which it may be adherent. The nerve should first be located  at the apex of the popliteal fossa, where it leaves the sciatic trunk, and followed along, the border of the biceps tendon. As it crosses the plantaris and the lateral head of the gastrocnemius the peroneal nerve becomes greatly flattened, which shape it retains in its passage around the neck of the fibula under cover of the peroneus longus, where it breaks up into its terminal divisions.

In the popliteal fossa the peroneal nerve gives off sensory branches. The peroneal comunicating is the superior branch and crosses the head of the gastrocnemius muscle toward the posterior aspect of the calf, where it perforates the deep fascia and joins with the tibial communicating to form the external saphenous nerve. A short distance below the peroneal communicating the sural branch or branches are given off, supplying sensation to the external surface of the leg. Occasionally the sural branches leave the tibial trunk with the peroneal com unicating,. These sensory branches bring from the dorsal aspect of the tibial trunk and have a long, combined intraneural course.

FIG. 197.- Exposure of external popliteal ind its terminal division as it swings around the neck of the fibula, the insertion of the peroneus longus having been divided to expose the terminal branches. A, Lateral hamstring tendon; B, head of fibula; C, origin of peroneus, divided; D, peroneal communicating; F, sural nerve; G, peroneal nerve; H, anterior tibial nerve; I, tibial recurrent ; J, musculocutaneous nerve, K, branches to peronei; L, musculotaneous, sensory portion

In the upper part of the popliteal fossa the peroneal nerve is composed of four bundles, the most dorsal of which contains the sensory fibers of the peroneal communicating and sural nerves. Below this sensory bundle, two bundles are found, one of which contains motor fibers to the peronei muscles, the other containing sensory fibers to the musculocutaneous nerve. Below these, in the ventral part of the nerve trunk, lies a bundle which contains fibers for the anterior tibial nerve. As the nerve proceeds along the lateral border of the popliteal fossa and the dorsal cutaneous bundle is given off in the formation of


1060

the peroneal communicating and sural nerves, it undergoes a degree of torsion and flattening in passing over the lateral head of the gastrocunemius, so that the three remaining bundles lie side in a mediolateral plane: the most medial bundle now being, that which ultimately forms the anterior tibial nerve, the two lateral bungles, soon to unite, form the musculocutaneous nerve--the more lateral bundle containing the motor fibers to the peroneal muscles, the medial bundle, the sensory fibers to the musculocutaneous. As the flattened trunk passes around the head of the fibula, the lateral hundlle forms the musculocutaneous nerve and the dorsal. or lower, bundle forms the anterior tibial and tibial recurrent nerves.

Exposure of the terminal branches is accomplished by carefully dividing the fibers of the peroneus longrus at its origin, as thle nerve is followed around the bone. Occasionally it is possible to effect suture of the individual branches as in the posterior interosseous nerve, where it fans out; beyond this point, reconstruction becomes increasingly difficult.

Physiologic approximation of the peroneal nerve is greatly facilitated by the position of its branches and the flattened or ribbonlike shape of the nerve in the lower part of the fossa and as it passes around the neck of the fihula. In the upper part of the popliteal fossa, the dorsal surface of the nerve is definitely indicated hiy the position of its sensory branches. In this position the nerve is oval in shape, its longest axis being in a ventro-dorsal plane. As it passes over the lateral head of the gastrocnemius, the dorsal quadrant of the nerve is rotated to a lateral position. which in passing around the neck of the fibula becomes ventral and superior, and from this ventral border the musculocutaneous nerve is given off. The dorsal part of the nerve, which is now inferior, becomes the anterior tibial, which gives off the tibial recurrent nerve.  Though most writers describe the tibial recurrent nerve as one of the terminal peroneal branches, its intraneural course is common with the bundle containing anterior tibial fibers. With a visualization of the natural torsion occurring in the peroneal trunk in the popliteal fossa and as it winds around the neck of the fibula, and with the assistance afforded by the flattened shape of the nerve trunk, the surgeon should have little difficulty in effecting physiologic approximation.

CONTINUITY DEFECTS

Relatively large defects in the sciatic trunk and its terminal external and internal popliteal portions may be overcome by flexion relaxation of the knee, providing these nerves are extensively mobilized, particularly in the popliteal fossa. There is no type of transposition which can he used in the lower extremity to shorten the course of these nerves. Defects, therefore, uncorrected by primary stretching and flexion relaxation, if at all correctable, will probably yield only to secondary stretching, i. e., the two-stage operation, in which the unsectioned nerve ends are united, after full advantage is taken of primary stretching and flexion relaxation, by xtensive liberation of the nerve trunk in both directions. In the writer's experience, several cases were encountered in which approximation seemed hopeless at the primary operation, and grafting was resorted to. In each instance regeneration failed, but in a subsequent


1061

operation end-to-end suture was effected by the two-stage method. We believe that the two-stage operation will probably be found efficacious in approximating any correctable defect, and that grafts should be resorted to only after the two-stage operation has failed, certainly never before it has been attempted. When end-to-end approximation has been despaired of, and in cases of incorrectable regenerative failure, the disability, in so far as ankle instability and drop-foot are concerned, may be fairly well corrected by arthrodesis or tendon transplantation. (See irreparable defects, p. 9.59.)

SECONDARY SUTURE

Defective regeneration in sciatic lesions, as in other nerves, may be due either to defective neuraxon, or defective muscle regeneration: it is utmost importance, in considering the question of secondary operation, to determine the degree of responsibility of each. In approximately 85 per cent of complete sciatic trunk sutures, plantar flexion of the foot returns to some degree through regeneration in the gastrocnemius-soleus group. In about 40 percent of cases, some voluntary power, after a four-year period, is observed in the peroneus longus and tibialis posticus. In about 30 per cent, some flexion of the toes is observed, while return of voluntary movement in the tibialis anticus and extensor of the toes is found in less than 20 percent. In no case has the writer observed a regeneration of the intrinsic foot muscles, following sciatic suture or suture of one of its terminal branches. In approximately 80 percent of suture cases there has been return of sensation, to some degree, in the foot, though rarely, after four years, has the writer observed a complete disappearance of anesthesia, though in most instances trophic disturbances and pressure sores have ultimately healed. In the majority of instances in which anesthesia his disappeared, discriminative sensation has not been recovered, the anesthetic zone having been replaced by a type of nondiscriminatorv sensation, which is usually interpretedl by the patient as tingling; there seems, however, to be a tendency toward a gradual sensory improvement. A few of these patients find it possible to distinguish between hot and cold water when taking a bath, but are unable to localize touch, or discriminate between a dull and sharp point. A study of these regenerative defects throws important light upon both muscle and nerve regeneration.

Neuraxon regeneration is demonstrated by Tinel's sign and in most instances is found to be present in some degree. When Tinel's sign is totally absent below the suture line, or is greatly diminished in intensity, after six months, a defective suture is probatbly accountable. This is particularly true when pressure over the suture line elicits a strong reaction; in such instances secondary suture is indicated.

In a number of cases, percussion of the peroneal nerve over the head of the fibula has elicited tingling, localized on the plantar surface of the foot. This indicates that the tibial portion of the nerve during suture was probably united to the peroneal, in that tibial sensory fibers are present in the peroneal trunk. In other instances, the writer has observed plantar flexion of the foot when the patient was attempting dorsollexion, which was perhaps due to the


1062

presence of peroneal motor fibers having reached the gastrocnemius. The presence of some regeneration in the gastrocnemius in about 80 per cent of cases can be accounted for in two ways: First, this muscle, because of its large size and high innervation, receives a certain number of regenerated motor fibers; second, because of the drop-foot, it is seldom overstretched as are the extensors of the anterolateral group. The failure of recovery in the tibialis posticus and flexors of the toes indicates that overstretching is not entirely responsible for these motor defects, which in all probability are due to an extreme degree of muscular degeneration and atrophy following prolonged denervation. The failure of muscle regeneration in the intrinsic foot muscle can probably be accounted for by extreme muscular degenerative changes, due to their small volume and distal position. We have not observed a single instance of regeneration in the intrinsic foot muscles, following a paralysis lasting longer thran 18 months. Even in cases of spontaneous regeneration. or after a neurolysis, with probably no disturbance of nerve pattern and in which sensation was fully recovered, the intrinsic foot muscles failed to show any signs of regeneration and their extreme atrophy persisted.            
           
The frequent failure of regeneration in the extensors of the anterolateral group, namely, tibialis anticus, extensor longus hallucis, and extensor longus digitorum, requires special comment. A number of cases have been observed, following suture and neurolysis of the peroneal nerve, in which the anesthetic areas showed much improvement, without a restoration of motor function in these muscles, while voluntary power was observed to have returned in the peroneus longus. Instances of this kind can be explained only on the ground of muscle stretching due to foot-drop, which the peroneus longus escapes, the anatomical couIrse of its tendon behind the external malleolus. In several such cases, lengthening of the gastrocnemius tendon, with splinting of the foot in dorsoflexion, has in a short time from one to three month, resulted in return of voluntary power in the tibialis anticus and to a less degree in the extensors of the toes.         
           
The defect in peroneal lesions presenting a return of voluntary power in the peroneus longus, without a corresponding recovery in the tibialis anticus, may with reason be attributed to the evil effects of overstretching. A failure of regeneration in the peroneus longus, however, indicates that extreme degree of atrophy and miuscle degeneration which is so commonly found in the legg following prolonged sciatic paralysis: this regenerative muscular failure is also, commonly observed in the smaller muscles on the dorsurm of the leg. It maybe roughly stated that between 40 and 50 percent of failures in the return of dorsoflexion of the foot are due to overstretching.           
           
There is a marked tendency for the gastrocnemius to develop contractures in all sciatic trunk lesions, as well as terminal peroneal lesions. The prevention of these contractures and their correction after development is of extreme importance. A patient with a sciatic or peroneal paralysis should be provided with a satisfactory foot-drop splint and the importance of its use strongly emphasized. Not only should a splint be applied to the shoe for day wear, but there should also be provided a night, splint, and its persistent use should be


1063

insisted upon. When contracture in the gastrocnemius has definitely developed, lengthening of the Achilles tendon should not be postponed and the patient trained to assume, when sitting, a position which the writer has termed in his clinic, for convenience, “Achilles corrective habit posture, " in which the patient stretches the Achilles tendon by resting the anterior part of the foot on the floor with the knee overflexed, exerting pressure by the weight of the opposite leg when crossed over the knee--the common habit of sitting with the knees crossed: the common habit of sitting with the knees crossed; i.e., the injured foot, while resting on the floor, is drawn backward in a position causing tension on the Achilles tendon, which is stretched by the superimposed weight of the opposite leg. The development of this habit corrective posture has proved of great value in stretching the Achilles tendon and relaxing the extensor muscles.

Secondary sutures should be attrempted only for the correction of defective neuraxon regeneration, or for marked degrees of torsion. In resuturing the sciatic trunk or its terminal portions, identification sutures are not to be employed for surface identification, as the nerve may have suffered torsion in the original suture. In the sciatic trunk, the intraneural septum between its component tibial and peroneal portions will serve as a reliable guide in effecting physiologic approximation. In the popliteal portion of these nerves, anatomic or branch identification will serve in the correction of defects caused by torsion. (See technique for sciatic trunk and popliteal lesions, pp. 1051-56.)

SUPPLEMENTARY PROCEDURES FOR IRREPARABLE DEFECTS AND DEFECTIVE REGENERATION IN THE SCIATIC TRUNK AND IT'S TERMINAL DIVISIONS

Though complete anesthesia of the foot follows total paralysis of the sciatic trunk the sensory disability alone does not seriously interfere with locomotion. Undoubtably, the greatest factor in the disability following complete sciatic paralysis is referable to the motor side. The loss of sensatition in the foot, however predisposes to traumatic insults, and a continued irritation of a traumatisized anesthetic area may be serious in its consquences-through pressure sores are not usually so and will readily heal when he irritation is relieved.
          
In a series if more than 300 sciatic lesions. not a single instance was observed in which it was not possible to effect approximation of at least the tibial portion of the sciatic trunk; in about 80 percent of cases followed, there has been a restoration if motor function to some degree, with at gradual disappearance of pressure sores. The more defect consists briefly in the loss of stability in the foot. In the majority of instances following sciatic suture, there has been a return of least of at some useful power in the gastrocenemius, rendering plantar flexion of the foot possible, which assists in correcting much if the original disability. It is obvious that the tibial nerve from a functional standpoint is of more importance than the peroneal, and in continuity defects of both nerves resisting correction by every other means, the surgeon is justified in sacrificing the peroneal portion of the trunk, when the defect in the tibial can be corrected by the use of the peroneal as a transplant, in an efffort to restore at least, gastrocnemius function and sensation to the sole of the foot.


1064

VIABLE NEUROPLASTIC TRANSPLANT FOR THE CORRECTION OF TIBIAL DEFECTS

First stage.- In this procedure, to assure viability of the graft, a two-stage operation is done, in which the proximal end of the sciatic trunk first receives attention. At the lower end, the line of cleavage between the peroneal and tibial portions is found and these respective elements separated for a short distance up the sciatic trunk. The ends of both nerves are now resected until normal appearing nerve bundles are found, they are then brought together and carefully approximated, making a loop of the distal end of the proximal segment of the sciatic trunk. The sciatic trunk is now followed upward for a distance a little longer than that necessary to fill the defect; at this point the line of cleavage between the peroneal and tibial portions of the nerve is identified

FIG. 198.- This and Figure 199 illustrate viable neuroplastic transplant for repair of tibial portion of sciatic trunk in irreparable lesions of both divisions. A, Exposure of proximal end of sciatic trunk; identification sutures placed; B, neuroma resected from both divisions: C, tibial and peroneal portions separated in line of cleavage to permit end-to-end approximation; D, end-to-end approximation uniting proximal end of tibial division to proximal end of peroneal division; E, peroneal division divided to permit degeneration of peroneal fibers in that portion of the trunk to be used as a transplant; upper end of peroneal portion injected with alcohol to prevent regeneration of peroneal fibers in transplant; F, peroneal transplant now becomes a viable part of the tibial portion of the nerve and presents a neuroma, showing complete migration of tibial fibers; G, transplant turned down and united to distal end of tibial division, filling the defect--peroneal portion having been sacrificed to repair the tibial. (See fig. 190)

and separated for a short distance to permit division of the peroneal portion. The purpose of this step is to permit degeneration in the anastomosed segment of the peroneal nerve, which is later to be reflected downward and attached to the distal segment of the tibial nerve as a viable transplant. This procedure will allow regeneration of the tibial fibers around the loop and tip the intact transplant. The upper end of the peroneal nerve, proximal to the point where it is divided, is now injected with alcohol, or foreign tissues interposed to prevent regeneration of peroneal fibers shown that portion which is to be used as a transplant. In this way, the viability of the transplant is assured during migration of tibial fibers.

If the transplant to be utilized is 6 inches long, the second stage of the operation is postponed for a period of seven months, allowing regeneration to proceed at the rate of 1 inch per month plus one month’s grace. If the transplant


1065

FIG. 199


1066

plant is longer or shorter, the time of the second operation will be modified accordingly.
          
Second stage.- The sciatic trunk is again exposed above, and the tibial nerve below; and a favorable bed prepared to receive the regenerated transplant. The loop at the end of the sciatic trunk is isolated with great care and the line of cleavage between the tibial and peroneal (now transplant) portions of the trunk carefully separated up to a point where the peroneal was originally divided: the transplant is now turned down. The upper segment of the tibial nerve, with its vialble transplant attached, is accurately approximated to the distal segment of the tibial nerve, after the scar tissue of both ends has been carefully resected. If the branches to the gastrocnemius are properly preserved, and their dorsal position in the tibial trunk borne in mind, and the approximation with the transplant is planned without torsion, a return of function in the gastrocnemius and sensation to the sole of the foot may be expected.           
           
Inasmuch as voluntary dorsal flexion of the foot is permanently lost, the drop-foot may be corrected by arthrodesis of the ankle joint or tendon slinging--no muscles being available to restore by tendon transplantation the lost function of the anterolateral group of muscles supplied by the peroneal nerve. Artrodesis stabilizes only the posterior portion of the foot without giving support to the toes; in certain eases it seems to be valuable but occasionally it is attended with end results entirely unsatisfactory to the patient. Tendon slinging is by far most satisfactory of these procedures for the correction of drop-foot in irreparable nerve lesions. It consists in suspending the foot by passing the upper part of the tendon of the tibialis anticus and peroneus longus through a hole in the shaft of the tibia. where thev are anchored; and supplementing this suspension with the peroneus brevis; the extensor tendons of the toes are used to suspend the anterior part of the foot.
                                    
TECHNIQUE
          
A long incision on the anterolateral surface of the leg, midway between the crest of the tibia and the fibula is carried down to the ventral aspect of the ankle joint, mid-way between the two condyles. As it reaches the annular ligament it curves to the outer side of the foot, below the external malleolus, to expose the peronei tendons. The tendon of the tibialis anticus is now isolated, and freed from muscle fibers to a point about 6 inches above the annular ligament. The tendons of the extensors longus halluctis and digitorum are treated in a similar manner. The peronei tendons are exposed by undermining the lateral border of the incision; after being mobilized. they are divided at the upper end of the incision. They are again exposed and their sheath opened on the lateral border of the foot; after having been. withdrawn from their sheath they are directed anterior to the condyle, through a passage made under the annular ligament. A hole through the crest of the tibia is now made by a gouge, the tendon of the tibialis anticus drawn through from the medial side, and the tendon of the peroneus longus passed through from the lateral side, with the foot held in dorsal flexion, at a right angle to the tibia, and neutral to inversion and eversion, the tendons are drawn taut, the angle of flexion being slightly overcorrected: they are then fixed with strong linen sutures. Experience


1067

has shown that it is advisable to supplement this procedure by also suspending the tendons of the peroneus brevis, extensor longus digitorum, and extensor longus hallucis; which is effected by suturing them to the transplanted tendons of the tibialis anticus and peroneus longus, the combined group of tendons being too large to be passed through the hole in the tibia. It is essential that the foot he kept slightly overflexed and straight, and each tendon attached individually, after its tension has been properly adjusted. After six weeks, the plaster cast may be disregarded: at shoe fitted with a drop-foot splint should, however, be worn during the day and a drop-foot bed splint used at night. Walking on the foot without a suitable drop-foot splint should not be permitted until after the third or fourth month, by which time the tendons should be firmly fixed in hardscar tissue.

FACIAL NERVE

FACIAL PARALYSIS
          
Most surgical lesions of the periphleral portion of the facial nerve are found in its course through the facial canal in the temporal bone, where it is occasionally involved in skull fractures or injured during operations for mastoid disease. Its most frequent affection, however, is that type of facial paralysis termed Bell's palsy, with a common history of exposure to cold. Its long course through the facial canal, which it almost completely fills, renders it particularly susceptible to injury or disease in this region; it may be completely severed or crushed by fragment of bone during a mastoid operation, or compressed by inflammatory exudate, hemorrhage, or congestion. Bell’s palsy is a compression paralysis due to congestion or inflammatory affection from exposure. Congestion of the facial nerve in this region is particularly serious because the nerve is unable to expand, being completely surrounded by a bony wall, and any congestion or inflammatory exudate

FIG. 200.-  Dissection of temporal bone, showing course of facial canal in its vertical and tympanic portion--wire directed through canal; anterior to the nerve lies the tympanic cavity. Note the straight course of the nerve from its position anterior and below the eminence of the lateral semicircular canal to the stylomastoid foramen. Identification of the nerve at these two points--just below the lateral semicircular canal and at the stylomastoid foramen--is the keynote to the exposure of the nerve through its vertical course. (Ney. The Laryngoscope, 1922)

is rapidly converted into a compressive agent. The nerve, after its exit from the stylomastoid foramen, is occasionally injured by stab or gunshot wounds; as it passes through the parotid gland it may be compressed by suppurative processes or neoplasms. Injuries distal to the parotid gland, unless very extensive, seldom implicate the complete nerve, usually involving only one or more of its terminal branches.

May facial palsies of the compressive type recover spontaneously, as the congestion or inflammatory exudate responsible for the paralysis subsides. After surgical division during mastoid operations, if the facial canal is no blocked


1068

by bone fragments or scar tissue, the nerve will usually regenerate. If spontaneous regeneration progresses satisfactori1y, the facial muscles should show signs of recoverv before one vear, and in each instance this period should be allowed to lapse lefore a decision of nonregeration is definitely made. During this expectant period it is essential that the facial muscles be properly splinted to prevent overstretcihing. which frequently is responsible for much delay in muscle regeneration.

NERVE ANASTOMIOSIS IN FACIAL PARALYSIS
          
Frequent attempts have been made to correct facial paralysis by anastomosing the spinal accessorv or hypoglossal nerve to the facial. In this procedure the facial nerve is exposed at its exit from the stylomastoid foramen and divided ; the hypoglossal or spinal accesssory, whichever may be selected, is next exposed and divided and its proximal end approximated to the end of the distal segment of the facial. The regenerating spinal accessory or hypoglossal fibers will now pass through the peripheral facial trunk to the facial musculature; after a period of five or six months these muscles will regain their tone providing regeneration has been satisfactory--and the face will show great improvement during repose. The facial muscles, however, fail to react to emotional expression, anl to all intents and purposes remain totally inactive until called into play by spinal accessory or hypoglossal impulses. If hypoglossofacial anastomosis has been done, the facial muscles will contract vigorously during mastication and deglutition ; after facio-accessory anastomosis, the facial muscles will react during turning of the head or elevating of the shoulder.  These facial movements, following facial nerve anastomosis, are entirely irrelevant to emotional expression, and occasionally are extremely embarrassing, in attempts to swallow or lift the shoulder will inaugurate winking or drawing up the angle of the mouth. Some writers have recommended the use of the hypoglossal nerve for anastomosis in facial paralysis in preference to the spinal accessory on the ground that anatomic proximity of the facial and hypoglossal center in the cortex might facilitate reeducation. Experience seems to indicate that the conversion of the hypoglossal center to facial function is not readily, if ever, accomplished. Ney has not observed a single instance of nerve anastomosis in which any of the coordinated movements required in facial expression have been regained, though he has had under his observation several cases which, for a period of eighteen months, received constant daily instruction under, the direction of experienced teachers. It has been possible to train certain favorable cases to a point where they would be able to effect satisfactory isolated closure of the eye, Or contraction of the muscles about the corner of the mouth, without synchronous movements in the shoulder or tongue; coordinated emotional movements however, were totally lacking in spite of all reeducative enleavors; the hypoglossal or the spinal accessory fibers remain hypoglossal or spinal accessory fibers from a functional standpoint and probably never will be able to subserve any other function. The most that may be said in favor of nerve anastomosis for the correction of facial paralysis is that, in favorable cases, there will be a restoration of tone in the facial muscle; it is usually associated with more or less embarrassing incoordinated facial movements


1069

which are incited by hypoglossal or accessory stimuli. The operation of nerve anastomosis for the correction of facial paralysis is now being generally abandoned.

The possibility of recovering coordinated emotional expression in the paralyzed facial muscles probably lies only in restoring the physiologic continuity of the facial nerve. This procedure has generally been considered impractical from a surgical standpoint, because of the difficulty involved in isolating, and repairing the facial nerve. Ney, however, has devised a technique by which the facial nerve may be satisfactorily exposed and repaired in that portion of the facial canal where it most frequently suffers trauma, namely , in its vertical or tympanic segment.

TECHNIQUE OF EXPOSURE AND REPAIR OF THE FACIAL NERVE THROUGH IT'S COURSE IN TIHE TEMPORAL BONE FROM THE GENU OF THE FACIAL CANAL TO THE STYLOMASTOID FORAMEN

The incision corresponds to that commonly used in the radical mastoid operation, except that it is extended forward, above the ear, to expose the zygomatic tubercle: below it is extended from 1 to 2 inches beyond the mastoid tip, following the anterior border of the sternomastoid muscle. Through this incision the mastoid bone is exposed and the soft parts reflected from the entire mastoid process, completely exposing the mastoid tip, during which care should be used to prevent injury to the nerve as it emerges from the stylomastoid foramen. The soft parts should be thoroughly freed from the under surface of the auditory process of the tympanic plate. The stylomastoid foramen is located about 7 millimeters in front of the tip of the mastoid; its depth from this point varies considerably, depending upon the extent of mastoid development. In infants, before the mastoid has developed. the facial nerve emerges from the stylomastoid foramen directly on the surface of the skull, where it may be injured in the application of forceps. With the development of the mastoid the stylomastoid foramen becomes covered with the mastoid tip and the auditory plate of the tympanic ring. The cartilaginous portion of the external auditory meatus is entirely separated from its bony wall, except in its anterior portion. Its attachment to the auditory process is usually firm, because of the roughened character of the bone in this region. It is important, however, that the auditory plate be well exposed.

The auditory plate of the tympanic ring is now removed by a sharp chisel or strong rongeurs. This removed should be executed with considerable care to prevent fracture, of the tympanic ring the bone being cut down rather than broken down. The removal of the tympanic plate is the key to the situation so far as locating the stylomastoid foramen is concerned. The auditory process now having been removed, the operator may proceed with the removal of the mastoid cortex and cells. The mastoid antrum is located as in the usual mastoid operation. The mastoid cells are now thoroughly removed, endeavoring to keep the cavity as shallow as possilble by trimming down the rim left by their evacuation; this is particularly important in the suprameatal region, where the cortex and cells of the posterior zygoma should be removed completely as far


1070

FIG. 201.- Primary incision and exposure of the mastoid tip, suprameatal ridge, superior, posterior, and inferior bony mestal walls. The chisel shows point at which removal of auditory plate is begun, to permit exposure of the stylomastoid foramen. (Ney. The Laryngoscope, 1922)


1071

FIG. 202.- Auditory portion of the tympanic bone, partially removed, exposing the exit of the facial nerve from the stylomastoid foramen to where it passes into the substance of the parotid gland. Mastoid cortex removed and cells evacuated-probe passed into antrum; suprameatal ridge partly removed. (Ney. The Laryngoscope, 1922)


1072

FIG. 213.- Bridge formed by the posterior meatal wall broken down over antrum,exposing the eminence of the lateral semicircular canal;.suprameatal ridige not sufficiently broken down. (Ney. The Laryngoscope, 1922)


1073

forward as the zygomnatic tuberle. Frequently it is necessary to expose the dura to gain sufficient, depth in this relgion.
             
The mastoid tip is entirely removed and the facial nerve exposed as it emerges from the stylomastoid foramen. Occasionally the localization of the nerve is attended with some difficulty; its identification is often facilitated by palpation, when it is felt as a tense cord passing from the stylomastoid foramen to the parotid gland. The difficulty usually lies in attempting to locate the nerve in the region of the mastoid tip, when in the majority of instances it is to be found anterior to the tip. The posterior meatal wall may

FIG. 204. - Facial nerve uncovered through a portion of its vertical and tympanic course, showing method of breaking down the wall with a fine, sharp chisel. (Ney. The Laryngoscope, 1922)

now be reduced to the level of the mastoid evacuation; that portion overlying the mastoid antrum is entirely removed. If an attempt is made to break down the antrum with at chisel, great care should be used, as the chisel in breaking through the bridge may traverse the antrum and enter the facial canal, or the dislodged bone driven into the Canal. There is no doubt but that, the facial nerve is frequently injured during this procedure in the radical mastoid operation. The prominence of the lateral or external semicircular canal is now identified and immedliately anterior to this structure, between it and the tympanic cavity lies the prominence of the facial canal. Occasionally, in this region the facial nerve will be found exposed through at defect in the canal wall.


1074

If during the evacuation of the mastoid cells any suppurative process is uncovered, it should be thoroughly eradicated and the exposure and repair of the facial nerve postponed until the infection is entirely controlled. If it is deemed necessary to drain the mastoid, the second operation is not to be considered until the mastoid wound has been completely healed for three months and the ear free from all discharge, or sterile to culture.

The stylomastoid foramen, having been exposed below, and the facial canal located anterior to the lateral semicircular canal, the operator may now definitely orient the course of the vertical segment, which follows a direct vertical

FIG. 205.- The sheath of the facial nerve is firmly attached to the periosteum of its canal; its attachment is severed with a cataract knife, while the nerve is gently lifted froii its bed. In this region the facial nerve is very fragile and gentleness is necessary to prevent tearing the nerve trunk. (Ney. The Laryngoscope, 1922)

line between these two points. The reduction of the posterior meant wall and mastoid cells may now be safely conducted to the level of these points, without fear of injuring the nerve. The vertical portion of the canal, which is about 1 mm. in diameter, lies, in the average case, about 3 mm. behind the posterior meatal wall; frequently the canal is exposed as a ridge of compact bone, through occasionally cells may open directly into it. After the mastoid has been reduced to about the level of the canal wall, the lateral or exposed side of the canal wall, the lateral or exposed side of the canal wall is carefully scaled off with a small sharp chisel, beginning at the stylomastoid foramen and progressing upward to the bend of the canal.


1075

FIG. 206.– The nerve removed from the facial canal throughout its vertical and tympanic course. (Ney. The Laryngoscope, 1922)


1076

Occasionally, the canal is separated from the mastoid cells by 1 or 2 mm. Of solid bone; at other times it is only a thin shell.

The vertical segment of the canal extends from the stylomastoid foramen to the "bend'" or the pyramidal segment, which lies in front of the lateral semicircular canal and constitutes that portion of the canal where its direction changes from the vertical to the horizontal; the bend is usually gradual, implicating from 2 to 6 mm. of the canal, though occasionally the direction changes

FIG. 207.- Decompression of the facial nerve by openinig its sheath. Diagrarn shows the attachment of the sheath of the facial nerve to the periosteum of the canal, which has been preserved in the dissection. The sheath of the facial nerve is opened, exposing fibers which appear as a pearly white bundle. (Ney. The Laryngoscope, 1922)

with a sharp angle. The tympanic segment of the canal constitutes its horizontal portion, which extends from the pyramidal segment to the genu and passes along the roof of the tympanum; its average length is about 8 mm. In its horizontal course, the canal undergoes a gradual steady rise in level, reaching the junction of the roof and inner wall, where it leaves the tympanic cavity to form the genu. About midway in its course it comes in close relationship with the fenestra ovale and stapes. After forming the genu, the canal gently curves toward the cochlea and tunnels the dense petrous portion of the bone throughout


1077

FIG. 208.- Plastic procedure to protect the nerve from subsequent compression by turning down a flap of temporal fascia which is passed under the nerve, separating it from immediate contact with the bone; method of anchoring the flap. (Ney. The Laryngoscope, 1922 )


1078

its course, independent of the contour of the upper surface of the bone. This, the labyrinthine section of the canial, extends from the genu to the lamini cribrosa.
          
After the vertical segment of the facial nerve has been uncovered, the operator, by the same process of careful scaling, may uncover the nerve through its pyramidal and tympanic portions, the canal wall being gradually reduced until the nerve is well exposed. In cutting away the wall of the facial canal, fragments of bone will le found firmly attached to the periosteum of the canal and should be dissected rather than torn away; the sheath of the facial nerve is intimately attached to the peliosteum of the canal, and the tearing away of attached fragments may injure the nerve, which is extremely fragile. The separation of the nerve from the canal wall is begun at the stylomastoid foramen, where it is elevated on a fine glass hook, and the attachments of the periosteum to the bone are divided by a fine-pointed cataract knife. During dissection of the facial nerve, the operator should take time to control oozing-a fine stream of warm saline is usually effective.

With the above mobilization of the nerve, the nature of the lesion will probably be apparent. If the nerve is not found to lie divided, it has probably been subjected to compression but careful inspection is required, as its ends may be united by sear tissue. If it is completely divided, it may be sutured, after the scar has been resected from its ends, with arterial silk, care being used to prevent torsion of the nerve trunk, the center of the exposed quadrant of its sheath having been previously marked by identification sutures. The needles used are Lane's half-curved cleft-palate needles, No. 3, which are of very small size and may be handled by mosquito forceps. If the nerve is found to be the subject of compression, it should be decompressed by carefully splitting its sheath with a cataract knife, exposinge its pearly white bundles; as the point of compression is reached, these bundles are often pinkish in color. If, in opening the sheath, an area of scar invasion is met (in which the bundles lose their identity) this area must be resected and the nerve approximated by end-to-end suture. If approximation is found diflicult, because of a continuity defect in the nerve, the lower segment of the nerve should be mobilized as far forward in the parotid gland as possible. This mobilization, by permitting the nerve to take a more direct course, will frequently facilitate the correction of a defect up to 1 mm. The mobilization of the facial nerve may be somewhat increased by uncovering its proximal end to the genu of the canal, when it may be transposed; the anterior part of the cartilaginous portion of the external auditory meatus is separated from its bony wall and the entire ear turned forward; the nerve when sutured is made to pass anterior to the cartilaginous portion of the meatal canal instead of assuming its original posterior position. After transposing the facial nerve to the anterior position, the cartilaginous canal is replaced and sutured. The mobilization of the proximal end of the nerve is greatly facilitated by reducing the suprameatal wall, and removing as far internally as necessary the superior surface of the petrous portion of the temporal bone, after elevating the dura until the genu of the canal is fully exposed. When this procedure fails to overcome a continuity defect, recourse may be had to grafting, when a sensory


1079

FIG. 209.- The portion of the temporal muscle denuded of its fascia, turned over the nerve. The facial nerve now lies between the flaps formed by the temporal fascia and the temporal muscle. (Ney. The Laryngoscope, 1922)


1080

nerve of approximate size may be inserted to fill the defect. It has not been found necessary to resort to a graft of this kind in any of the cases.

The preparation of a bed for the repaired nerve is a rather difficult procedure Occasionally the question may be solved by transposing the nerve to a position anterior to the cartilaginous meatus. In all decompressions and occasionally after suture, the writer has protected the nerve by turning down a flap of the temporal fascia and muscle, the muscle being used to cover the nerve, while the smooth surface of the temporal fascia is passed under the nerve to protect it from immediate contact with the bone.

The incision is closed by interrupted sutures: three small rubber drains will usually be found sufficient to care for the rather abundant oozing following the operation. These drains are allowed to remain in place for 48 hours. A 5 percent dichloramine-T solution is instilled into the middle ear, and the external auditory meatus lightly packed with gauze. If, during the course of the operation, the ossicles have been completely removed, the external auditory meatus mav be obliterated by suturing together the cartilaginous canal before the skin is closed. If this is done. a drain is placed from the mastoid incision directly into the middle ear, and the condition of the wound carefully watched.

The after-treatment consists mainly

FIG. 210.- Incision closed; points of drainage indicated. External auditory meatus lightly packed with iodoform gauze. (Ney. The Laryngoscope,1922)

The after-treatment consists mainly in preventing overstretching of the paralyzed facial muscles by the application of adhesive straps, as suggested by Yawger; or by retracting the angle of the mouth with a hook, the hook being held by a piece of string or tape, which is tied to adhesive, attached to a shaven portion of the scalp above the ear back of the hair line, and kept under sullicient tension to prevent overstretching of the paralyzed muscles during laughter.
          
The exposure anal repair of the facial nerve in the temporal bone is a practical, though very delicate, procedlure, and one requiring considerable patience. Each step of the procedure should be carefully and completely executed; a little carelessness here or there may jeopardize the entire operation. Before attempting this procedure clinically, the operator should thoroughly familiarize himself with the anatomy by making several complete facial nerve exposures in the dissecting room. It is not necessary to use any especially made instruments in this operation; it is essential, however, that the surgeon be equipped with the smallest size chisels, gouges, and osteotomes.