Contents

776

SECTION III

NEUROSURGERY

CHAPTER III

MANAGEMENT OF GUNSHOT WOUNDS OF THE HEAD AND SPINE IN FORWARD HOSPITALS, A. E. F.

CRANIOCEREBRAL SURGERY PRIOR TO OUR ENTRANCE INTO THE WORLD WAR

Wounds of the head in the war, 1914-1918, were, generally speaking, of a more serious nature, as was true of all wounds, than was the case in former wars, due principally to the short-range firing of trench warfare, to the employment of intense artillery fire, much of which was high explosive, and to bombing from the air. Thus the relatively great number of head wounds requiring surgical intervention presented a problem seriously demanding a solution. It was evident from the first that the victims of cerebral injury were likely to constitute the last residium of the wounded needing hospitalization long after the end of hostilities.

There was a lack of unanimity of opinion, with respect to the management of these cases, which persisted throughout the war. This was more marked. however, during the first two years. Toward the latter part of this period (1916) it was considered good practice to transfer head wounds to the rear, in view of the fact that patients with head injuries bore transportation badly following operation. This, of course, meant a delay of  36 to 72 hours and longer. When to operate, which cases to operate, the anesthetic to be used, etc., gave rise to much discussion.

 De Martel ¹ and Pauchet ² were amnong the first to advocate local anesthesia in operations on the head. This presented several advantages: It did not raise the blood-pressure, either by its action per se, or as the result of the patient's struggling during the induction stage of anesthesia: and it enabled the patient to assist at the operation by coughing, which oftentimes extruded pulped brain substance, d ebris, and particles of bone, after trepanation of the skull and exposure of the dura.

The British surgeons advocated osteoplastic bone flaps with the wound at or near the (center of the flap (see figs. 1 to 4). The finger was used in palpating for foreign bodies. but gentleness in the use of the finger was emphasized. This in itself, in experienced hands. was not a serious drawback to the operation, but as many men, with limited experience in this branch of surgery, were called upon to care for these cases, lack of gentleness was very often the cause of much additional damage to cerebral tissue, and there was a high mortality. The French employed trepanation of the skull, and some surgeons habitually removed shell fragments and bullets under the fluoroscope. This method had several distinct disadvantages: It was not only necessary to remove the metallic foreign body, but pieces of indriven bone, hair, pulpel brain, and filth as well: it served as it temptation to the surgeon simply to get the foreign body itself

777

and to content himself with a more or less incomplete toilet of the tract. Furthermore, seeking the metallic fragment in the brain under the fluoroscope caused needless and, oftentimes, much additional damage to the brain.

FIG. 1.- This and Figures 2 to 4, inclusive, the technique of the osteoplastic method with the wound near the center of the flap

FIG. 2

In selected cases, however, after a careful toilet of the tract and when the foreign body could not be located definitely or removed by a powerful magnet, this method was the only solution in removing the shell fragment or bullet. The

FIG. 3

FIG. 4

French insisted also on the importance of removing all foreign bodies at the first operation: if they were at all accessible, and the resulting late brain abscesses in cases of retained missiles have verified the wisdom of this contention.

778

As stated, it had been the custom in the British Army to route all head cases to the base, and large numbers of them had passed through the hands of Colonels Sargent and Holmes at General Hospital No. 13, at Boulogne, where it was customary in the case of penetrating wounds to turn down an osteo-plastic flap, including the wounded area, to remove the foreign bodies, and to replace the flap, draining from either or both lower angles. The wound itself was often closed from the inside. With this method the mortality was high and secondary infection was relatively common.

During the Pasehendale battles in the summer and autumn of 1917 a new program was put into operation whereby the head injuries were routed to one of the casualty clearing stations at Proven and operated upon before being sent

FIG. 5.- Sketch illustrating the method of suction of the tract of a penetrating wound while searching for foreign bodies

to the base. To this station Cushing ³ and some of his assistants from United States Army Base Hospital No. 5, serving with the British Army were attached and a new method of procedure was adopted. In simple terms, it consisted in approaching the tract in the brain through the wound. This was done by excising the scalp wound down to the skull and employing the tripod or the Isle of Man incisions further to expose the skull. A piece of bone around the hole in the skull was then removed en bloc. A soft rubber catheter was passed into the tract in the brain to locate foreign bodies, a syringe being used to pro-duce suction. At the completion of the operation the scalp was tightly closed by two layers of fine interrupted silk sutures. This method lowered the mortality, prevented secondary infection, and lessened the possibility of hernia cerebri. This modification of the technique formerly employed had several very important advantages: It did not produce any more damage to cerebral tissue and it tended to prevent secondary infection from the outside. The wound, if it tended to break down, always opened at the junction of the three triangular flaps.

779

This technique lowered the mortality from between 50 and 60 percent to28.5 percent in one of the early series. It was generally accepted and practiced in the American and the other Allied Armies.

CLASSIFICATION OF HEAD INJURIES

The classification of head injuries which follows is that adapted by Cushing in his critical study of the cases which had passed through his hands at a British casualty clearing station during three months in 1917.⁴

Grade I.- This group comprised wounds of the scalp, with both cranium and dura intact, though occasionally complicated by an underlying cerebral contusion. Of 22 cases observed, one was fatal, a mortality of 4.5 percent.

Grade II.- Wounds producing local fractures of variable types, with the dura intact, were placed in the second grade. They were subgraded further

FIG. 6.- Grade 11; Wounds producing local fractures of variable types, with the dura intact. Type A, without depression of external table; type B with depression of external table

into Type A, when there was depression of the external table (fig. 6). In the 54 wounds graded thus, local contusions of the brain, or extra dural extravasation were fairly common. Among the 54 cases of this grade observed 5 deaths occurred, or a mortality of 9.2 percent.

Grade III.-Local depressed fractures of various types, with the dura

FIG. 7.- Grade III: Local depressed fractures of various types, with the dura puntured

punctured, were placed in this grade (fig. 7 ). Among the18 cases observed, because of the inevitable local contusions positive neurological signs usually were present. Two deaths occured, giving a mortality of 1l.8 percent.

Grade IV.- In this grade wounds, usually of the gutter type, with detached bone fragments driven into the brain, were placed (fig. 8). Twenty-five cases were observed. Local contusion was severe, and extrusion of the

FIG. 8.- Grade IV: Wounds, usually of gutter type, with detached bone fragments driven into the brain
 
brain were placed (fig.8). Twenty-five cases were observed. Local contusion was severe, and extrusion of the brain almost inevitable. Fungus cerebri and encephalitis were common sequels. Six deaths occurred among the cases of this grade, giving a mortality of 24 percent.

780

Grade V.- This grade comprised wounds of the penetrating type, with lodgement both of projectile and bone fragments (fig. 9). The brain frequently was found extruding and there was much contusion along the tract. In such wounds, symptoms depended on the size and course of the missile. Common sequels noted were early compression an late abscess. Among the 41 cases of this grade 15 deaths occurred, 36.6 percent.

FIG. 9.- Grade V: Wounds of penetrating type, with lodgement both of projectile and bone fragments

Grade VI- Wounds of this grade comprised those in which the ventricles were penetrated (Type A) by bone fragments or (Type B) by missiles (fig. 10). Cerebral lesions in this grade were the same as in the wounds of the two immediately preceding grades. The escape of cerebrospinal fluid is constant; hemorrhage into, or subsequent infection of, the ventricles is common. In 14 eases in which the ventricles were penetrated or traversed by bone

FIG. 10.- Grade VI: Wounds with ventricles penetrated or traversed (a) by bone fragments, (b) by projectile

FIG. 11.- Grade VII; Wounds of craniocerebral type involving (a) orbitonasal, (b) auripetrosal region
 
fragments, 6 deaths occurred (42.8 percent): in 16 cases in which the projectile penetrated or traversed the ventricles, the mortality was 100 percent.

Grade VII.-Wounds of this grade were of the craniocerebral type involving (A) the orbitonasal, or (B) the auripetrosal region (fig. 11). In these wounds the brain is commonly exposed and extruding: the fractures are radiating: nasal or petrosal cavities are opened meningitis is common. Among 15 cases observed 11 deaths occurred (73.3 percent).

FIG. 12.- Grade VIII: Wounds with craniocerebral perforation

Grade VIII.- Craniocerebral perforations were placed in this grade (fig. 12). Extensive cranial and cerebral damage is common to such wounds; death usually is due to intracranial hemorrhage and compression. Among 5 of these cases observed, 4 deaths took place (80 percent).

Grade IX.-Craniocerebral injuries, with massive fracture of the skull were placed in this grade (fig. 13). Such injuries involve widespread cerebral contusion; compression phenomena are common. Of 10 of these injuries one half died.

FIG. 13. - Grade IX; Craniocerebral injuries with massive fracture of skull

781

ROUTINE PRELIMINARY TREATMENT OF HEAD INJURIES AT AN EVACUATION HOSPITAL, A. E. F.

Though having received some instruction in neurosurgical diagnosis before being sent overseas the members of the hastily organized neurosurgical teams attached to the evacuation hospitals of the American Expeditionary Forces had had no experience whatsoever with war wounds in general, much less with the complicated and special procedures which the treatment of craniocerebral injuries demanded. Profiting by such instructions as were given by the senior consultant, neurosurgery, the following routine, more or less modified by individual experience, was so far as possible put into operation.

Patients admitted to the receiving room were divested of all their clothing. which was deposited in a tent set aside for that purpose. They were covered with blankets and carried into the adjoining room where their field cards were inspected. Their heads were completely shaved and a hypodermic injection of1.500 units of antitetanic serum was given in the abdominal wall, if not previously administered at the triage hospital.

The patients were then sent through the X-ray room where, in each instance, an attempt was made to determine the presence, the location, and the depth of the intracranial foreign body. By means of the fluoroscope crosses, at right angles! were made on the scalp with a lunar caustic pencil. Skiagraphs were then taken, laterally and anteroposteriorly, and delivered to the operating room, and placed in the diagnostic box for the surgeon's reference. When this was completed, patients were placed in a room near by the operating room which could accommodate :30 men, awaiting their turns for operation. Operable cases in shock were carried to a special tent where they were given hot black coffee, and morphine hypodermatically. They were covered with blankets, so arranged as to drape over the sides and ends of the bed. A lighted oil stove was then placed under the bed. An enlisted man constantly watched such stoves. From time to time the patient's condition as to pulse, temperature, and blood pressure was noted. Immediately upon recovery from shock they were operated upon. In instances of severe hemorrhage, citrated blood transfusion was employed, when possible, in addition to the treatment already described. Inoperable shock cases were placed in another special tent where most of them died. Their treatment was the same as that described for operable cases in shock, with the exception of blood transfusion, as the scarcity of blood rendered it impossible.

In the operating room the surgeon in charge employed three tables. A hurried but careful examination was made of each patient before operation.,usually on the operating table. The patient was then given morphine, three-eighths grain hypodermatically, if no morphine had been previously given within four hours. At the operating table, patient sitting with a roll under the neck, the scalp was washed with green soap and sterile water and wiped off carefully with ethyl alcohol. Patients that required suboccipital exploration, or decompression, were placed face downward on the regulation stretcher, the forehead resting on one of the slings stretched between the two handles, and the stretcher placed on the operating table. Making a mental picture of the style of incision desired, tripod, Isle-of-Man, or flap, and its possible extent

782

a block of scalp was injected. A larger needle was then passed down to the pericranium, injecting deeply within the block. Large wounds often required plastic flaps of scalp to cover cranial defects, even though large areas of intact skull were denuded thereby. Thirty cubic centimeters of a 1 percent novocain solution, to which one-sixty-fourth grain of adrenalin chloride was added, was usually sufficient for one case.

TREATMENT OF DIFFERENT GRADES OF HEAD WOUNDS

WOUNDS OF THE SCALP

All scalp wounds were considered potentially serious, even in the absence of neurological findings, until proved otherwise by operative exploration. The importance from a military standpoint of caring for these cases in the forward area can not be overestimated, as such wounds, if not complicated by fracture and cerebral injury, heal readily, permitting an early return of the soldier to duty. Many slight cases would otherwise be evacuated to the rear. Very slight wounds of the scalp were often found to overlie a penetrating wound of the skull. These cases may prove to be very serious, as the bone may be perforated without apparent fracture.

Other cases, with the external table intact or with only small linear fractures apparent from the outside, may have extensive comminution of the internal table, with perforation of the dura and bone fragments in the brain. In the presence of neurological symptoms one should always drill down to the inner table, and if the symptoms are very marked the dura should be investigated. If found tense, even though intact, it should be

FIG. 14.- The in-driven fragments of inner table (natural size)

opened. In cases where the scalp wound is infected and the patient presents marked signs of cerebral injury the excised wound should be sterilized as well as possible and a block of bone removed, exposing the dura. If the dura is found to be perforated, the in-driven bone and pulped brain should be removed by the patient's coughing, by irrigating through a soft-rubber catheter, using sterile decinormal saline solution, suction, and the use of an esquillectomy forceps in removing the bone fragments. When the tract is clean, it is sterilized by injecting a small amount of dichloramine-T, with eucalyptus oil or ethyl alcohol, through the catheter on withdrawal.

A lumbar puncture should be done where symptoms of meningitis are present after injury. If the diagnosis is verified by spinal puncture, the case should not be operated upon, as such cases invariably die, operation or no operation. Figures 14, 15, and 16 reveal the findings in a case reported by Cushing, in which the external table was practically intact. Very extensive wounds of the scalp may occur without the slightest injury to the skull or the brain, but the reverse is much more common, namely, an apparently trifling though penetrating scalp wound which conceals an extensive cranio-cerebral injury.

783

Many small depressed fractures of the outer table were produced by shell fragments of spent velocity; they were usually tangential. On drilling down through such a fracture the in-ner table was often found to be intact. These cases in most instances recovered as rapidly as simple scalp wounds and, in the absence of neurological symptoms, could be returned to their organizations at the front.

When fractures of the inner table were disclosed with an intact dura, the membrane was not opened unless it proved to be tense or discolored when it was incised and the underlying brain inspected. If the brain proved to be pulped, the devitalized brain tissue was removed by irrigating gently with sterile decinormal saline solution, using the soft-rubber catheter, syringe, and bulb. If hemorrhage was present, the blood was evacuated, and any pial vessels found bleeding were

FIG. 15.- From a sketch at autopsy after removing calvarium

ligated with fine silk or preferably caught with silver clips. Naturally, when a torn dura has been disclosed, the question of advisability of incision or otherwise will not arise.

Fractures of both tables, with perforation of dura and with bone fragments in the brain, constitute a type of injury which is often complicated by the lodgment of one or more shell fragments or a bullet. The treatment of these cases will be described more fully under "Operations." Figures Nos. 17, 18, 19, and 20 are illustrations of this type of fracture. It is appropriate here to consider briefly "bursting" fractures of the skull, as they were sometimes associated

FIG. 16.- Section through the contused area, showing position of bone fragments

with local fractures. " Bursting " fractures were the result of perforating wounds, violent explosions, or falls, or of being struck by soft bodies. Some of these fractures were so extensive as to involve practically the entire

784

FIG. 17.- Trepanation block, showing behavior of thick skull to tangential wound

FIG. 18.- Bone block. Specimen on left shows interparietal suture and fissures radiating from gutter; on the right, a few fragments of internal table attached

785

skull. Numerous linear and radiating fractures occurred at the point of injury, while every fossa might show fractures. Unilateral and bilateral decompressions were performed on some of these cases, but the resulting cerebral edema was so great that recovery was rare.

Cases of this type that appeared to be hopeless but did not develop a fatal edema and were not operated upon were in some instances evacuated alive.

FIG. 19.- Example of lodged shell fragment in an obilque gutter wound

OPERATION

The prepared head, with the field of operation surrounded by sterile towels held in place by skin clips, being ready for operation, the scalp wound was excised down to the skull, and the excised tissue, forceps, and knife were placed in a basin and removed.

FIG. 20.- Small giltter fracture in thin skull; complete dislodgment of fragments

 The scalp incision.- The type most, generally used was described by Colonel Cushing as a tripod incision. Three straight incisions were made to the excised area in such a manner as to best facilitate the approximation of all edges. No general rule can be made, as the angles of the formed incisions differed with the location and the general outline of the excised area. Rat-toothed forceps were now placed on the galea, strips of gauze passed through the handles of the forceps attached to each flap, anil the flaps undermined. The strips of gauze were then fastened to the sterile sheet, serving as retractors, and the skull inspected. If the skull was intact, the wound was wiped out with

786

ethyl alcohol. The galea was then united with interrupted sutures of silk, or No. 0 or No. 1 chromic gut, and the scalp closed with silk sutures. The scalp sutures were removed in two to three days.

The three-legged or Isle-of-Man incision was the incision used in larger wounds. The technique was the same as that described in the tripod incision, except that each of the three incisions had a knee (fig. 22).

FIG. 21.- Tripod incision for small irregular wound of vault. Dotted lines indicate area of reflection of flaps. (Cushing)

Flap incisions ere employed in wounds of the temporal and suboccipital regions, especially in cases that required drainage. Occasionally straight incisions were used in the temporal region. Large osteoplastic flap incisions were employed in searching for a shell fragment or bullet, intracranial, but opposite to the wound of entrance.

The craniotomy.-The instruments required for trepanation of the skull are: A cranial perforator, a half-inch burr, a dural separator, and a pair of Montenovesi

FIG. 22.-Three-legged (Isle of Man) incision for larger wound of cranial vault. (Cushing)
 
or De Vilbis linear cutting forceps. The cranial perforator was used to perforate the bone down to the inner table or through the inner table at a smallpoint. This was followed by the burr. The dural separator then was used to elevate fragments of the inner table at the bottom of the opening made by the burr, and rotating it between the thumb and forefinger, the dura was separated well beyond the margins of the opening made by the drill. The linear cutting forceps then followed the burr.

787

Trepanations were triangular, quadrangular, or pentagonal, drilling 3, 4, or 5 holes. Pentagonal trepanation was performed when by so doing the defect might be smaller. Quadrangular and rectangular trepanations were usually employed in larger injuries. Figure 23 illustrates a quadrangular trepanation.

The osteoplastic flaps used were those common to surgery of civil life. The enlargement of an already existing defect in the occipital and lower frontal regions where the bone is thick and as small a defect as possible is desired, was done by the use of rongeurs. If the injury was over a sinus, trepanation was alwavs done.

FIG. 23.- Quadrangular trepanation

The infracranial procedure.- The perforation in the dura was not enlarged, unless the opening was very small. Pieces of indriven bone, hair, or felt from the inside of the helmet, if found in the opening, Were removed. A soft-rubber catheter was then passed through the opening in the dura and into the track in the brain, and bone fragments located in this manner were removed by the use of an equillectomy forceps. Pulped brain and small pieces of bone were were removed from time to time during the progress of the operation by the patient's coughing, by irrigating gently through the catheter with sterile decinormal saline solution, and by gentle suction, using glass syringe and bulb. As larger pieces were located by the catheter, they were removed. A shell fragment or bullet, when located, was removed by the equillectomy forceps, and the tract again very gently explored with the catheter, searching for more bone fragments.
Figure 24 illustrates the use of the catheter in locating foreign bodies.

FIG. 24.- Diagram to show the insertion of soft rubber catheter in locating foreign bodies

788

Only in cases where more than one tract existed it the brain, with shell fragments at different levels and wildely separated one from another, was the finger employed to locate them, and then with the utmost care and gentleness to avoid doing more damage than already existed (figs. 25 and 26). Opinions differed as to whether or not foreign bodies, difficult of access, should be removed. Foreign bodies in the brain should always be removed, if at all possible, as the chances of infection are very much increased, especially if bone fragments and, possibly, hair and filth lie below them.

FIG.25- Split shell fragments with separate tracts and fragments at varying depths. (Cushing)
        
 It was rarely found to be necessary to remove a shell fragment or a bullet under the flouroscope. This should never be done, unless the foreign body can not be removed by the usual method and no magnet is at hand, as more or less additional damage always results.

In searching for shell fragments where no tract existed from the side of the brain approached in the operation, i. e., in cases where osteoplastic flaps were turned down opposite to the wound of entrance for the removal of a shell fragment or bullet in the opposite hemisphere, a telephone probe was used. A telephone probe is an ordinary silver probe, 8 or 9 incites in length, to which one of the wires of an ordinary telephone receiver is attached. The other wire is attached to an empty brass cartridge shell. Taking care that the metal cartridge shell does not come in contact with any metal fillings, it is placed in the mouth of the patient. The probe is then used to search for the, foreign body. When it, comes in contact with the steel fragment, a spluttering is heard, as in the presence of overcharged electricity. This proved to be a very useful instrument in searching for shell fragments as already described, in the cerebellum, the posterior fossa, and the lateral ventricle.

Puncture of the lateral ventricle was done where bulging existed after turning down a large osteoplastic flap in the search for larger

FIG. 26.- Split shell fragments in temporal lobe. (Cushing)

shell fragments opposite to the wound of entrance. In these cases the original tract was first cleansed as deeply as possible and ethyl alcohol, or dichloramine-T with eucalyptus oil, injected.             In cases where the shell fragments entered the brain through the orbit, the destroyed eye was enucleated, the indriven pieces of bony orbit removed, and the pulped brain cleansed from the tract in the brain. The deep bone and shell fragments were removed by the esquillectomy forceps as these fragments were located by the catheter. When the tract was clean, it was injected with

789

dichloramine-T with eucalyptus oil, or ethyl alcohol through the catheter. Ocassionally this type of wound was approached through a supraorbital incision, enucleating the eye at the completion of the operation. 

Cases with a large shell fragment that had passed through the brain from above and embedded itself in the roof of the mouth were treated from above as already described for penetrating wounds of the brain. The embedded shell fragment was then removed through the mouth, using large foreign body forceps.

Bullets or larger shell fragments that passed through the anterior portion of the frontal lobe and lay on the basilar process of the occipital bone in front of the spine were removed through the original tract.

Shell fragments that penetrated the middle fossa from below were removed by first rendering approach possible. The zygoma was resected and the opening in the skull d brided by using small rongeur forceps. The catheter was then inserted anld pieces of bonie and the shell fragment. were removed when located.

Perforating wounds of the skull were sometimes associated with bursting fractures. The treatment consisted of trepanation of the wounds of entrance and exit, cleansing the tract from both ends of all pulped cerebral tissue and pieces of indriven bone, some of which were found nearer to the wound of exit that to that of entrance. Perforating wounds of the temporal region often were associated with blindness due to a severing, of either the optic nerves or the chiasm.

BRAIN ABSCESS

Operations on brain abscesses due to war wounds gave a high mortality. Meningitis resulted because the abscess, when opened. was usually opened through the uncontaminated subdural space. When abscesses were opened at a point in the skull directly over the site of the injury through a relatively small opening, without disturbing the adhesions to the inner table and opening the dura carefully it was possible. in sone cases, to open directly into abscess. These cases did not develop meningitis, because no connection was established with the uncontaminated subdural space. Brain abscesses that developed under a scalp wound in which the skull was found apparently intact were best treated in this manner. Neglected cases, or cases in which the foreign body could not be removed, or was not removed, at the first operation. could not be treated as described for the cases with no fracture, or an undiscovered fracture of the inner table, when the abscess developed at, some distance from the original wound. In such instances, it was necessary to turn down a flap in order to locate the abscess. Abscesses of this type were drained through one of the openings in the skull made by the drill, using a soft rubber-tissue or gutta-percha drain.

SPINAL INJURIES

War wounds of the spine were particularly distressing. These injuries were so frequently associated with chest and abdominal wounds of a serious nature that one scarcely knew where to begin, if to begin at all. In the forward hospitals, cases in which a transverse lesion was suspected were not.

790

operated upon. These, complicated by serious wounds of the chest and abdomen, were considered inoperable. Of the operable cases, those of the bony spine, compression of the cord, and partial lesion of thc cord, were the only ones which held out, a little hope of benefit from surgical interference.

Fractures of one or more spinous processes and laminae were common in wounds entering from the back. Wounds of the spine and cord in which the shell fragment or bullet entered from the front rarely caused fractures of the vertebrae in perforating the bodies, unless the shell fragment was large, when the case was hopeless and inoperalile. The most difficult cases to deal with were those of partial lesion of the cord in which the shell fragment or splinter entered from the front, penetrating the cord or perforating it and remaining in situ. Occasionally one end of the shell splinter would be lodged in the body of the vertebra and the other in the cord. At other times the shell fragment might be free in the canal.

Injuries in which the shell fragment or bullet struck the transverse process were accompanied by early symptoms of a transverse lesion following the injury. Some of these cases recovered spontaneously without any interference, while others developed a true transverse myelitis. Shell fragments or bullets which struck the spine and were, deflected without producing fracture, caused a local contusion of the cord in some instances. Injuries of this type sometimes recovered spontaneously. The symptoms in the cases which recovered spontaneously were due to a form of concussion in which all function below they was suspended for a time. The finding of a Babinski reflex soon after the injury showed that a complete transverse lesion did not exist. Such a ease was classified as a partial lesion.

Just what should be done for the bladder in these spinal cases was never a matter for general orders. It remained a difference of opinion whether permanent drainage, intermittent catheterization or abstention from any intervention, merely allowing the bladder to fill and overflow by dribbling, was the method most likely to forestall infection. On the whole there was something to be said for each of these procedures, but the "let alone" policy was that in general favor in tbe evacuation hospitals. The main object, of course, was to avoid infection if possible, for only so were the automatic lower-cord reflexes likely to be restored and thereby an automatic and periodic spontaneous evacuation of the bladder established.

For descriptive purposes wounds of the spine may be classified as follows:(1) Wounds of the bony spine without perforation of the dura or injury to the cord: (2) wounds of the bony spine without perforation of the dura, but with injury to the cord; (3) wounds of the bony spine with perforation of the dura and injury to the cord; (4) injuries to the cord without external wounds.

Cases of the first and second categories will be jointly considered, as the dura was not opened in these cases. The external wound was excised and loose bone fragments were removed. The wound was then sutured, bringing the muscle together with No. 2 or No. 3 chromic-gut interrupted sutures, and the skull closed with heavy interrupted silk sutures. When compression of the cord existed, the depressed bone or shell fragment lying on the dura, was removed. Great care was exercised

791

so as not to produce further injury in relieving the compression. Shell fragments wedged in the fracture, or between the laminae or spines, if firmly embedded, were approached from either side by performing laminectomy. In infected cases the wounds were left wide open, Carrel-Dakin tubes inserted, and the wound packed with sterile gauze saturated with Dakin's solution. No sutures were inserted.

The treatment of cases falling in the third category will be described under operations.

Cases in the fourth category were not operated upon. Collier ⁵ has described these as spinal concussion. Cases whose spines had been exposed to the shock of violent explosions showed numerous small subpial hemorrhages.

The results obtained in operations on wounds of the spine with injury to the cord were very discouraging, on the whole, and the mortality very high. Of 32 injuries of the cord repoi ted by Cushing,⁶ 7 were cervical, 2 were thoracic,8 were lumbar, and 15 were not specified. Eight were inoperable and there were 23 deaths, or a mortality of 71.8 percent; 24 were operated upon with 15 deaths, or an operative mortality of 62.5 percent. These cases were all cared for in the forward area.

In considering records of work done in the forward area, it must be borne in mind that the surgeons were forced to labor under trying conditions, finding it very difficult at times to properly care for the wounded. It was at such times of great activity that the records were more or less incomplete. Because spine cases were usually evacuated early, if at all transportable, any following up of these cases in the forward area was thus impossible. Many of these cases undoubtedly died soon after their evacuation to the rear. It was rather common to have men with spinal cord injuries arrive dead or dying. Injuries of the spine., perlhaps, formed a much larger group than those computed from hospital records would lead one to think, as the serious woundls involving the chest and abdomen in which death occurrecl at the front, were undoubtedly in many instances, complieated by spinal injuries.

OPERATION

The utmost gentleness and most extreme care should be taken in the handling of the cord. It should never be sponged or pressed upon. For the removal of foreign bodies delicate forceps should be used. Cord debris and blood should be removed by gentle irrigation with sterile decinormal saline solution. All one can expect to do is to remove foreign bodies and pulped cord substance that is free in the spinal canal, and in this manner to remove infected material and prevent infection. If this is done, one has accomplished all that is possible. Suture of the cord is a vain and harmful procedure, as the added handling produces more injury. An injured cord can be cleansed, but not restored.

REMOVAL OF FOREIGN BODIES

The external wound was excised down to the bony spine and loose fragments of bone removed. If the wound was directly over the spine, the excision was enlarged at either end and laminectomy performed. When the wound was on either side of the spine, the skin incision was made directly over

792

the spine. In separatin'g the nmuisculal, senlitendillnous. and fascial attachments from the spines and laminae, a large periosteal elevator was used. Retractors were then placed in the wound, thoroughly exposing the bony spines. The spines were removed by large bone-cutting forceps and the laminae carefully rongured away. The spine and lamiinae of two or three vertebrae were removed in this manner.

The dura was first opened in the following manner: Two delicate silk sutures were placed in the duia on either side of the median line. Pulling up on these sutures, the dura was carefully incised with a scalpel. The opening was enlarged by using a pair of straight and slender-banded scissors. As the opening was gradually enlarged, other sutures were inserted as before and used as retractors.

On inspecting the cord, if only a contusion existed, cord debris was removed as much as possible by irrigating gently with sterile decinormal saline solution without bringing the syringe in contact with the cord. In practice when bone fragments were found, the cord was first irrigated as already described, and bone fragments remaining in the cord were carefully removed by all esquillectomy or other delicate forceps, always in direct line with that of entry.            
   
Small shell fragments embedded in the cord were removed in the same, manner. Such fragments, if buried in the cord, were approached by first making an incision carefully in the long axis of the cord, preferably in its lateral aspect, severing one of its

FIG. 27.-  Method of opening dura

posterior nerve roots and using it as a retractor and removing the fragment from the front of the cord. This was important, as often the end of the fragment presenting itself on the anterior aspect of the cord was larger than the portion deeply buried in the cord or extruding posteriorly. In this mnanner further damnage to the cord was avoided. Missiles of this type that penetrated the cord, but remained embedded in the body of the vertebra, presented the greatest problem for the operator. In order to free the cord from the foreign hodv it was necessary to sever several anterior and posterior nerve roots on one side, or on both sides, to permit lifting the cord entirely free from the firmly embedded splinter that penetrated or transfixed it. A firm hold on

793

the shell splinter embedded in the body of the vertebra could then be secured by using small curved rongeur forceps and extracting. If extraction was difficult several methods were found to be useful. Lifting the cord by its posterior roots, the shell-splinter was firmly, grasped by the rongeur and an attempt made to deflect to one side and extract. This was not difficult unless the portion embedded in the body of the vertebra was larger than appeared apparent from the portion exposed. When found to be firmly fixed rotation on its long axis was done, having the effect of a drill, and attempting from time

FIG. 28.– Exposing cord for removal of embedded shell fragment

FIG. 29.- Exposing cord for removal of embedded shell fragment, using nerve root as tractor

to time to rock it back and forth. Great care was necessary in order not to break it, leaving a portion of it projecting into the calnal. When such splinters broke off, leaving the bony canal free, they were disregarded. By perseverance and firm but gentle force the removal of such bodies was possible in most instances when it often seemed impossible. Another danger was the possibility of the rongeur slipping and striking the anterior and lateral portions of the cord, resulting in contusion. Operations on the spinal cord required greater care than the usual operations for cerebral injuries.

794

REFERENCES

(1) Pauchet, Victor: L'Anesthésié régionale. O. Doin et fils, Paris, 1914.
(2) de Martel, T.: La chirurgie cranienne sous anesthésie locale. Bulletins et mémoires de la société de chirurgie de Paris, July 24, 1918, xliv, 1364.
(3) Cushing, Barvey: Notes on Penetrating Wounds of the Brain. British Medical Journal, London, February 23, 1918, xliv, 1364.
(4) Cushing, Harvey: A Study of a Series of Wounds Involving the Brain and its Enveloping Structures. British Journal of Surgery, Bristol, 1918, v, No. 20, 558.
(5) Collier, James: Discussion on Gunshot Wounds of the Spine. British Medical Journal, London, March 25, 1916, i, 451.
(6) Hanson, Adolph MI.: A Report of Wounds Involving the Head and Spine Cared for at Evacuation Hospital No. 8, A. E. F. The Military Surgeon, 1920, xlvi, No. 4, 414.