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







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






Chapter IV



Arterial Aneurysms and Arteriovenous Fistulas General Considerations

Daniel C. Elkin, M. D. and  Harris B. Shumacker, Jr., M.D. *

For descriptive purposes, battle-incurred vascular injuries may be divided into:
1. Wounds in which the blood vessel is completely severed or in which vasospasm exists to such an extent as to render the blood supply so impoverished that death of the part or useless fibrotic tissue results.

2. Wounds in which the damaged vessel is only partially severed, producing a false aneurysm or arteriovenous fistula.

3. Wounds which result in activation of previously existing blood vessel disorders, including vascular tumors.

  In numerous instances of complete vascular severance, death inevitably occurred promptly on the battlefield because of excessive loss of blood. In patients who survived, ligation of the vessel above the point of injury was often the only feasible procedure. Early repair of the vessel while theoretically possible and always desirable was not usually practicable under existing conditions. Sometimes, particularly when small arteries were injured, clot-formation and retraction of the severed ends occurred, the process probably being facilitated by the lowered vascular tension associated with wounding. Late complications were not usual in this type of vascular injury. They were also the exception in injuries superimposed on preexisting cardiovascular or peripheral vascular disease, very few of which were observed because physical examination at induction centers was sufficiently thorough to prevent the induction of individuals with such conditions. The situation was quite different, however, when vascular injury resulted in incomplete severance of the affected vessel. Late complications were then relatively frequent.

  The number of aneurysms and arteriovenous fistulas of traumatic origin observed in vascular centers in the Zone of Interior in World War II far exceeded any similar series ever recorded. This increase may be attributed to the introduction of higher velocity projectiles. In addition to the ordinary

*Professor of Surgery and Chairman of the Department. Indiana University Medical Center. Formerly Lt. Colonel, MC, AUS.


wounds caused by machineguns, rifle bullets, and shell fragments, a great many multiple injuries were produced by fragmentation of land mines, grenades, and aerial bombs. These last-named produced a great many small individual wounds--thus increasing the incidence of trauma to the blood vessels.

  Numerous additional factors figured in the development of arterial aneurysms and arteriovenous fistulas after vascular injuries in World War II. The majority of these would likewise be operative in noncombat-incurred injuries. They included (1) the particular artery involved;  (2) the size and other characteristics of the injury to the vascular wall;  (3) the size and direction of the channel which served as an exit for the hemorrhage, that is, the path of the missile;  (4) the nature of the surrounding tissue, which determined the possibilities of controlling hemorrhage;  (5) the level of arterial tension;  (6) the degree of pressure applied to control hemorrhage;  (7) the adequacy of splinting; (8) the method of transportation; and  (9) the speed with which definitive treatment was supplied.

  The wide variety of possible causes of vascular injuries, combined with the inconspicuousness of many combat-incurred wounds caused by minute fragments of exploding missiles, made it inevitable that in the press of evacuating large numbers of casualties in theaters of operations some vascular injuries were initially overlooked, and the subsequent discovery of arterial aneurysms and arteriovenous fistulas entirely unexpected. Openings in blood vessels produced by small missiles frequently caused no early symptoms and it was natural that these should be overlooked, or their significance not appreciated. This was particularly true if more extensive injuries distracted the surgeon's attention from them. It was also understandable that under combat conditions an incomplete, or even an incorrect, diagnosis might be carried by a wounded soldier for a considerable period of time.

  One reason that these lesions were often overlooked was failure to bear in mind the fact that blood vessels are usually accompanied by nerves. The manifestations of nerve injuries were frequently so striking that they overshadowed other damage in the same area.

  During World War II it was repeatedly demonstrated that arterial aneurysms and arteriovenous fistulas could easily be missed unless every wound was examined from the standpoint of possible vascular injury, and unless auscultation was part of the routine examination. Furthermore, arteriovenous fistulas present from the onset of injury could be overlooked unless examination was repeated after edema had subsided and hemorrhage into the tissues absorbed.


  To understand the development of traumatic arterial aneurysms and arteriovenous fistulas it is necessary to discuss what may occur as a result of a penetrating wound of the blood vessels.


  When an artery is completely severed both ends retract. Hemorrhage may be severe at first but will stop either spontaneously or with the application of pressure, although the artery may require ligation or some attempt at repair. When the vessel is only partially severed an entirely different chain of events ensues. The wound tends to enlarge in the long axis of the vessel as the result of retraction, and hemorrhage is likely to be profuse. If the wound in the soft tissues overlying the vessel is of sufficient size, hemorrhage occurs externally and the indications for immediate control are clear. If the wound is small, the imbrication of muscle, fascia, and skin planes can prevent the escape of blood externally. Blood passes out through the rent in the arterial wall into adjacent tissues where its collection is limited in size and determined in contour by the nature of the surrounding structures, the extent of traumatic damage to these, also the size of the injured artery and of the tear in its wall, as well as by such factors as arterial pressure and effectiveness of the clotting mechanism. This results in the formation of a hematoma which is likely to become a false aneurysm.

As the clot becomes organized, it tends, as has just been intimated, to lose the characteristics of a hematoma and assume the characteristics of an arterial aneurysm. A wall composed of fibrous tissue and supported by surrounding structures becomes evident. The flattened innermost layer of fibroblasts may be suggestive of endothelial cells, and roentgenologic examination sometimes reveals deposits of calcium, but the characteristic elements which compose the wall of a true aneurysm formed by dilatation of an arterial wall are always lacking. The laminated clot continues to undergo organization, but the natural tendency toward spontaneous cure is seldom completely attained and, in the absence of surgical correction, rupture and extension are always possibilities.

  When an aneurysm results from the sudden giving way of a diseased arterial wall, the sac which forms may immediately become filled with thrombus. Like the thrombus in the sac of a traumatic aneurysm, this thrombus also tends to be compressed against the sac wall, to adhere to it completely or in part, and to become well organized with the passage of time. Frequently aneurysms which develop by the giving way of a diseased arterial wall are not filled with thrombus. This is particularly true when the aneurysm is of the fusiform type.

  In instances in which the external wound is small and the vein injured simultaneously with the artery, groundwork is laid for the formation of an arteriovenous fistula, the artery and vein communicating with each other directly or through the medium of a false sac.  It was characteristic of arteriovenous fistulas brought about by combat injuries that they resulted from penetration of shell fragments of small size, just as in civilian practice they frequently followed wounds made by knives and ice picks, injuries accidentally produced in the course of phlebotomy, or the application of transfixion sutures or of


pins for skeletal fixation. A direct shunt of blood from artery to vein could result if these vessels were injured laterally by a missile which passed between them or through them. In through-and-through wounds blood could be shunted directly from the artery to the vein only if the lateral wounds became sealed. Arteriovenous fistulas were particularly likely to develop by this mechanism if the affected vessels were enclosed in a common sheath. Femoral and carotid vessels were for this reason particularly frequent sites of arteriovenous communications. Some arteriovenous lesions followed blows with blunt instruments, no external wound being present.

False aneurysms were sometimes found in association with arteriovenous fistulas. They sometimes lay between the vein and artery and sometimes projected from the site of injury of one vessel or the other. From notes made at the time of operation and from examination of the excised specimens, an analysis was made of the presence or absence of aneurysm and the type when present in 195 patients with arteriovenous fistulas observed at the vascular center of Mayo General Hospital. In Figure 10 some of the types of fistulas and associated aneurysms studied have been diagrammatically represented. Of the 195 cases analyzed, 78 (40 percent) had no aneurysm--there was only a direct communication between artery and vein--(Fig. l0A), 117 (60 percent) had 1 or more aneurysms present (Fig. 10B through J).

  In 89 of the 117 patients with 1 or more aneurysms, the artery and vein communicated by means of an interposed saccular aneurysm, or 1 to 2 aneurysms arose from the fistula itself. (Fig. 10B through D.)  In 13 patients there were 1 or 2 separate arterial aneurysms in addition to a direct fistula; in 10 of these 13 there was a single aneurysm and in 3 a double arterial aneurysm. (Fig. 10E and F.)  In 10 patients there was an aneurysm arising from the vein and a direct fistula.(Fig. l0G.)  In 1 patient there was an arterial aneurysm and the artery and vein communicated through a second aneurysm (Fig. 10H); in another there was an aneurysm arising from the vein as well as 1 through which the 2 vessels communicated (Fig. 10I); and in 3 there were aneurysms both of the artery and vein (Fig. 10J). The aneurysms varied in size from a mass 1 cm. in diameter to an ovoid mass about 22 cm. long and 15 cm. in diameter.

  When developed in these locations true aneurysms differed in two respects from the false which arose independently:  (1) a laminated clot was likely to be present in the sac; and  (2) they were usually smaller, the tendency toward expansion being controlled by the decompression afforded by the vein.

  Early signs of combat-incurred arteriovenous fistula varied and depended upon the location of the lesion and its size. After injury the patient might become aware of the presence of a pulsating mass or discover the thrill characteristic of the condition. On the other hand, he might notice neither mass nor thrill, in which case the fistula could have been overlooked unless a routine search was made because of the existence of circumstances under which it might develop.


 Figure 10. Diagrammatic representation of various types of arteriovenous fistulas and associated aneurysms. Symbols: A, artery: V, Vein:, S, Sac.

  Regardless of the site of the fistula, the local clinical evidences of its presence were always the same. The most common sign was a bruit or murmur which could be heard in the region of the lesion. The murmur, in contrast to the murmur of an arterial aneurysm which is purely systolic, was always continuous and lasted through both systole and diastole, the systolic accentuation of this sound providing the distinguishing characteristic of fistulas. Although in general, the larger the fistula the louder the murmur, this was not always true; small fistulas sometimes produced loud murmurs. The murmur of a fistula, like a cardiac murmur, might be transmitted for a considerable distance from


the site of origin. Invariably the chief cause of failure to hear the murmur of an arteriovenous fistula was failure to listen for it.

  When the murmur was of sufficient intensity, a thrill might be felt.  Again, although somewhat indicative of the size of the fistula the intensity of the thrill was not completely reliable evidence of this, since it might be modified by the position of the fistula and by the structures which overlay it. The thrill of an arteriovenous fistula was less well transmitted than the murmur associated with it and was always limited to an area near the lesion.

Differential Diagnosis

  Differentiation between a false arterial aneurysm and an arteriovenous fistula was frequently difficult, but was always important since the sequelae, the general and local effects, and the treatment of the two conditions were altogether different. The points of differentiation were the same as those employed in civilian practice:  (1) As a rule an arteriovenous communication is characterized by a continuous vibratory thrill and a loud, rough continuous murmur with systolic intensification. In a false aneurysm, although a murmur is also present, there is a distinct pause between the systolic and the diastolic phase and the murmur is often heard only in systole. (2) In an arteriovenous communication the murmur is usually transmitted for some distance on either side of the lesion along the course of the vessels, while in a false aneurysm the murmur is only occasionally heard beyond the confines of the dilatation. (3) In an arteriovenous communication, although the presence of a false sac may cause a tumor of considerable size, swelling is not usually pronounced. In a false aneurysm a definite pulsating mass is usually demonstrable.  (4) The dilatation of cutaneous veins in the region of an arteriovenous fistula, and the slowing of the pulse on temporary occlusion of the lesion are additional points of differentiation since neither of these signs is associated with the presence of a false arterial aneurysm. (5) While coincident cardiac disease may, of course, be present, the heart is not affected by the presence of an arterial aneurysm. In large arteriovenous fistulas, on the other hand, and in those which persist over a period of time, cardiac dilatation with subsequent cardiac failure is likely to develop.


  The staffs of the three vascular centers established in the Zone of Interior by The Surgeon General cared for and studied large numbers of soldiers with traumatic arterial aneurysms and arteriovenous fistulas. Two hundred twenty-one arterial aneurysms and 593 arteriovenous fistulas were treated in these centers during World War II (Table 6), but it must not be assumed that these figures represent the total number of such lesions which occurred in patients in the United States or in patients evacuated from overseas theaters to the United States during the course of the war.





  While these figures do not depict the Army incidence, they do represent a large proportion of these complications. Not included are some operations that were performed overseas either because of urgent indications or in violation of the directives issued by The Surgeon General 1 which provided that Army personnel with specified types of vascular diseases and injuries would be transported to the vascular centers in the Zone of Interior.2 Furthermore, the sequelae of vascular injuries are not always immediately apparent (sometimes a considerable length of time elapses before they are detectable) and data concerning these late developments are naturally not a part of this analysis. Notwithstanding, these statistics can be assumed to be reasonably complete and certainly representative of the special sequelae of vascular injuries as they were observed in the vascular centers in the Zone of Interior.

  The disparity in numbers between 221 arterial aneurysms and 593 arteriovenous fistulas observed is immediately apparent. The most logical explanation for this unequal distribution is that arterial aneurysms more often present acute surgical emergencies than do arteriovenous fistulas. By their very nature aneurysms tend to become progressively larger and therefore subject to rupture. On many of them, then, surgery could not be delayed a sufficient length of time to permit the patient to be evacuated to the Zone of Interior.

1  (1)  Interview, M. D.  Historians with Maj Gen Norman T. Kirk, 29 Nov 51.  HD: 000.71.

  (2)  Interview, M. D.  Historians with Maj Gen Paul R. Hawley, 18 Apr 50.  HD: 000.71.

(1)  Ltr, Surg ETO, 19 Mar 44, sub:  List of medical conditions which make it advisable to return personnel for hospitalization in Zone of Interior.  HD: .008 Policy (ETO Professional Service File).

(2)  Cir Ltr 101, Off Surg ETO, 30 Jul 44, sub:  Care of battle casualties.


  While statistics are not available, experience of surgeons in foreign theaters points to the fact that numerous patients with aneurysms were operated on because of rupture or to forestall rupture.

  In most patients with arteriovenous fistulas, surgery could be delayed until the patient could be transported to the Zone of Interior, for an arteriovenous fistula has no tendency to bleed externally. The very nature of this lesion, which is actually a hemorrhage from artery to vein, acts as a safety valve to prevent loss of blood and seldom presents an acute surgical emergency. These figures, therefore, do not reflect the frequency with which such lesions occurred and cannot be used for comparative statistics. The fact that there were 593 arteriovenous fistulas and only 221 arterial aneurysms in this analysis is not an indication that the former occurred, roughly, three times more frequently.

Site and Regional Frequency

  Concerning the sites of the arterial aneurysms and arteriovenous fistulas observed in the vascular centers in the Zone of Interior during World War II, no extended comment is necessary. They were identified in practically every blood vessel in the body (Table 6) but were most frequent in the axillary, brachial, and subclavian vessels in the upper extremity; in the femoral, popliteal, and tibial vessels in the lower extremity; and in the carotid and vertebral vessels in the head and neck. The largest number in any single vessel occurred in the femoral artery.  Aneurysms and arteriovenous fistulas of the vessels of the trunk were uncommon; wounds of the aorta, the renal artery, and other large and deep-seated vessels of the abdomen usually precipitated immediate death.


  As a rule the diagnosis of aneurysms and arteriovenous fistulas resulting from battle injuries offered few difficulties; the chief problem in such cases was to demonstrate physical signs of their presence. If the lesion lay deep under heavy musculature, the characteristic pulsating mass was sometimes neither visible nor palpable but other signs which clarified the clinical picture were usually present. When a proper survey of the whole body was carried out after wounding, the diagnosis of combat-incurred arterial aneurysms and arteriovenous fistulas offered no greater difficulties in military practice than do similar lesions encountered in civilian practice.

  Occasionally, especially when associated nerve injuries existed, the question arose whether the lesion was an arterial aneurysm or an abscess, and in at least one such case actual incision of the mass was undertaken before the error was realized.  Occasionally, too, an arterial aneurysm or arteriovenous fistula was suspected to exist when it did not, though the following case in which this difficulty was particularly confusing is believed to be unique:


  Case 1. A 30-year-old soldier was wounded in France on 2 August 1944 by shell fragments.  He received penetrating wounds of the soft tissue in the left parietal region of the scalp and in the left trapezius area.  There was little bleeding at the time of wounding. The wounds were debrided in an evacuation hospital and on 5 August he was admitted to a general hospital where an abnormal pulsation and bruit were noted in the left infraclavicular region.  Secondary closure of the wounds was carried out on 15 August.  Healing was satisfactory.A diagnosis of arteriovenous fistula was made and the patient transferred to the vascular center at Mayo General Hospital, Zone of Interior.

  When first seen at this center, 24 October 1944, the patient denied numbness, paralysis, pain, excessive sweating, knew of no color or temperature changes, and had not observed any dilated veins, edema, or trophic lesions.  He thought he had noted some swelling in the left clavicular area.  He stated his health prior to wounding had always been good.

  Examination revealed a number of dilated veins in the left pectoral and left infraclavicular areas, but otherwise inspection of the left clavicular area revealed nothing unusual.  There was no visible mass, and no abnormal pulsation could be detected. There was, however, a strong pulsation palpable just beneath the midportion of the left calvicle, which was not present on the contralateral side.  No thrill could be palpated but a loud systolic bruit was heard in this area.  The unusually well-developed muscles in both shoulders and arms made it impossible to compress the subclavian artery above the clavicle or the axillary artery beneath it, and tests for collateral circulation therefore could not be carried out.

  No color changes, edema, abnormal sweating, trophic, or neurologic changes were present in either hands or feet.  The peripheral arteries were neither thickened nor tortuous.  The radial, ulnar, and brachial pulses were normal and equal in the two extremities, and the venous filling time was the same (3 seconds) in both hands.  Skin temperature studies revealed no significant unilateral differences.

  The result of the Kahn test was negative.  Roentgenograms revealed several metallic fragments in the left parietal scalp, the left side of the neck, and the left supraclavicular region.

A diagnosis of traumatic arterial aneurysm was made, either of the distal end of the left subclavian artery or of the proximal end of the axillary artery.  Since it had been impossible, for the reason stated, to test the collateral circulation there was naturally a feeling of in security concerning its adequacy and as a precaution a preganglionic upper dorsal sympathectomy was carried out 31 October 1944.

  The vascular operation was performed 30 December.  The entire extent of the axillary artery was visualized by dividing the fibers of the pectoralis major and minor muscles but no aneurysm was found in the artery or in any of its branches.  The axillary vein was bifid and somewhat dilated just distal to the point at which it passed beneath the clavicle, and at which the two branches joined.  The clavicle could be elevated sufficiently to permit visualization and digital exploration of the subclavian vessels beneath, but neither in this area nor above the clavicle was there evidence of aneurysm or of constriction of the vessel by scar tissue. No foreign bodies were observed.  The muscles were resutured and the wound closed in layers with fine black silk.  No bruit was audible at the close of the procedure.

  Recovery from both sympathectomy and the vascular operations was uneventful.  It was noted postoperatively that the bruit was sometimes loudly audible though at other times it was absent.  The prominent pulsation beneath the left clavicle persisted.  The patient had a full range of motion in the left shoulder and continued to have no complaints referable to the upper extremities.

  Eventually observations were made which were thought to offer an adequate explanation for these findings:  The bruit was found to be present, or absent, or altered in character, according to the position of the patient and his method of breathing.  It was present and intense when he stood with shoulders unsupported.  It was absent when he was recumbent.  When he sat with elbows resting on the arms of a chair a single short bruit was heard with


each inspiration and expiration but none was audible when respiration was held either in inspiration or expiration. When he stood or sat with his arms unsupported and hanging by his side, the bruit ceased completely during full inspiration and at the same time all arterial pulsations in the left upper extremity disappeared.  Downward traction on the shoulders had no influence upon pulsation, bruit, or radial pulse. The various positions of abduction and hyperabduction, with the extremity in the plane of the body, likewise did not influence these manifestations regardless of whether the body was bent forward or backward.  Similarly, strong backward thrusting of the shoulders, with the head turned to one side or the other, had no effect upon them.  When the patient attempted to elevate the left shoulder against downward traction on the arm, the bruit often disappeared.

Oscillometric readings were within normal range in both arms (Table 7) except during deep inspiration when the patient was seated with his arms hanging unsupported by his sides; then oscillations ceased completely in the left arm. They also ceased when he stood with his arms hanging unsupported. When he was recumbent there was no change in the readings during forced inspiration.


  The blood pressure was 128 mm. of mercury systolic and 78 diastolic in the right arm and 120 mm. of mercury systolic and 80 diastolic in the left. When the patient was standing, the venous pressure in the cephalic veins in the antecubital fossae varied from 48 to 60 cm. of water. With deep inspiration there was a rise of 5 cm. of water or more on the left side and a drop of approximately 5 cm. of water on the right.

  Regardless of position, the left hand was of somewhat better color than the right. The radial and ulnar pulses were full at both wrists. In a room in which the temperature was 22O C., the skin temperature in the fingers on the right hand varied from 17.5O to 20O Centigrade.  On the left side the range of variation was from 32.5 Oto 33.5O Centigrade. The left side of the face and the left upper extremity did not perspire; on the right side the same areas perspired normally.  Ergometric studies, repeated on several occasions, revealed no essential differences in the fatigability of the two hands.  Roentgenograms taken in a position of lordosis revealed some narrowing of the space between the clavicle and the first rib on the left with inspiration, but the interpretations were inconclusive.

From these various observations it was decided that the physical signs which this patient presented could be explained by compression of the subclavian vessels on the left between the first rib and the clavicle. When he stood with arms unsupported, partial compression of the vessels occurred and there was a resulting loud systolic bruit. When he was recumbent, compression did not occur and there was no bruit. Full inspiration in the


standing or sitting position caused sufficient compression of the vessels to shut off completely all flow of blood to the left upper extremity. With normal respiration this phenomenon did not occur.

Since the patient had no complaints and had ample circulation under ordinary conditions, further operative interference was not thought justified and he was returned to duty.

  Comment.  For a number of years it has been well known that certain vascular and neurologic symptoms in the upper extremity may result from structural abnormalities or be associated with positional changes. Particularly well known are the scalenus anticus syndrome and the clinical pictures associated with cervical ribs and with tendinous or cartilaginous bands extending from rudimentary cervical ribs or transverse processes. Also well known are the symptoms resulting from compression of the subclavian vessels and the brachial plexus between the first rib and the clavicle, and neurovascular complaints resulting from the relatively normal obliteration of arterial pulsation in the upper extremity in various positions of hyperabduction when they are maintained for any length of time.

The case just described differs from previously reported cases of costo-clavicular compression. Although the patient had no subjective complaints, he presented physical findings which pointed to the diagnosis of aneurysm. Prior to operation, in fact, no doubt existed concerning the diagnosis, partly because the patient had sustained a wound which might have injured the subclavian or axillary artery and partly because of the signs present. Furthermore, aneurysms of considerable size in this area may exist in the absence of a pulsating mass, without thrill, and with only moderately strong pulsations and moderately intense bruits. In this case there was no mass and no thrill could be felt, but pulsation was prominent and the bruit was intense.

It is unfortunate that because of a diagnostic error this patient was subjected to two unnecessary surgical procedures, but under the circumstances it is difficult to see how they could have been avoided since, so far as can be determined, this particular syndrome had never before been reported and therefore was not suspected. The excessive development of the arm and shoulder muscles prevented testing of the collateral circulation, and preoperative sympathectomy was regarded as desirable for two reasons:  First, arterial aneurysms, in contrast to arteriovenous fistulas, provide notoriously poor stimuli for the development of a collateral circulation. Second, the possibility that surgical cure of an aneurysm will entail ligation of the involved artery always exists when an operation of this type is undertaken. It is true that the subclavian and axillary arteries can often be severed with relative impunity, but it is equally true that gangrene may follow such a procedure.

  Since the question might be raised as to whether the clinical picture was the same before and after the operation, it can be said that nothing was done during the exploratory operation which could have altered the regional anatomy and physiology, and that the postoperative findings offer adequate explanation of the picture presented before operation.



  In a large number of the arterial aneurysms and arteriovenous fistulas observed at the vascular centers during World War II operative cure (Tables 8-9) was possible, partly because of the very considerable number of these lesions from which cases for this procedure could be selected, partly because of a deliberate effort to increase the number of indications for operation, and partly because competent vascular surgeons had been chosen to head these centers and carry out the delicate and difficult procedures involved. Otherwise, the therapy of these vascular lesions did not differ from that which would be employed under similar circumstances in civilian practice.





  Most of the 221 arterial aneurysms were treated by endoaneurysmorrhaphy (99 cases) or excision of the aneurysm (98 cases).  Most of the 593 arteriovenous fistulas were treated by quadruple ligation and excision (526 cases). The techniques used were almost without exception those employed for these lesions when encountered in civilian practice.

  Anesthesia.  Fractional spinal analgesia was preferred for operations on the lower extremities since all procedures in this region were likely to be long and tedious. Thiopental sodium anesthesia administered intravenously, and supplemented as necessary by nitrous oxide and oxygen, was preferred for operations on other parts of the body. In operations about the neck and head an intratracheal tube was usually introduced to ensure smooth anesthesia and maintain an open airway.

  Almost without exception there were no complications of any consequence attributable to anesthesia at any of the vascular centers. This is a notable achievement since over half of the operations required 3 hours or longer and some lasted 8 hours. The advantages of the types of anesthesia selected are apparent for operations of such length.


Preoperative Care

  The status of patients with arterial aneurysms who are encountered in civilian practice may or may not be good. Often it is not good because of advancing age, the presence of organic cardiac or cardiorenal disease, hypertension, or similar complications. The great majority of aneurysms encountered in civilian practice have been brought about by an underlying disease and the surgical risk must be estimated with this in mind. Arteriovenous fistulas, on the other hand, are almost without exception of traumatic origin and the patient is quite likely to be a good surgical risk.

  The great majority of patients with arterial and arteriovenous fistulas treated at the vascular centers in the Zone of Interior during World War II were young men in excellent physical condition. They had no background of disease, cardiac or otherwise. Some patients it is true had other injuries, but for the most part the vascular lesion was the only disability, or the major disability, and the surgical risk had to be estimated principally on the basis of this lesion.  
  Preoperative preparation, therefore, was generally limited to tests of the circulatory status of the affected part and the institution of methods to improve the circulation. The usual routine was to keep patients with arterial aneurysms under close observation in the hospital until the proper time for operation arrived. If no contraindications existed they were permitted to walk, but their activities were restricted to those which did not require strenuous effort. Patients with arteriovenous fistulas, if they showed no evidence of cardiac enlargement, and many of them did not, were usually allowed to


leave the hospital on furlough but were required to return at regular intervals for checkups until the appropriate time for operation arrived.

Since surgery of vascular lesions of these types had to be delayed to permit the development of an adequate collateral circulation, any danger of infection from the original wound had long since passed before the operation was under taken. Preoperatively, therefore, chemotherapeutic and antibiotic agents were not administered.

  Patients whose arterial lesions had produced cardiac dilatation of any degree or who presented evidence of actual or impending cardiac failure were prepared for operation as they would have been under similar circumstances in civilian practice.

  When indicated, an internist was called in consultation. Physical therapy was employed as indicated before as well as after operation. It was of the greatest usefulness in patients in whom contractures had developed as a result of disuse, or in whom concomitant lesions of nerve or bone, or nerve and bone, complicated the vascular lesion.

Postoperative Care

  Dressings covering the operative wound were always small and simple and served only to protect the incision. However, when pressure for the obliteration of an aneurysmal sac was indicated an elastic bandage was employed. In order to prevent swelling it was so applied as to cover the whole extremity from toe to thigh or from fingers to axilla. Pressure was maintained until the patient was ambulatory or until there was no longer any evidence of an accumulation of fluid in the wound.

  Patients operated on for lesions of the smaller vessels, particularly vessels of the upper extremity, were allowed to be ambulatory within 24 hours of operation provided there were no contraindications. When their lesions involved larger blood vessels they were kept in bed for a period of 10 days to 2 weeks until the wound was well healed.

  Patients with a considerable degree of cardiac dilatation or with evidence of cardiac failure required special treatment. A longer period of bed rest was usually necessary and the return to normal activities was always graduated and cautious. Medication depended upon the indications of the special case. Treatment was usually carried out in cooperation with the internist.

Chemotherapy and Antibiotic Therapy

  The sulfonamide drugs (antibiotics were not available in the first months of operation of the vascular centers) were used in relatively few cases. When such treatment was regarded as necessary, the agents were given in standard dosages. They were never administered for purely prophylactic reasons and sulfanilamide and similar preparations were never placed in wounds.

When penicillin became available, however, it was administered routinely as a prophylactic measure. It was given (intramuscularly, as a rule) immediately after operation and continued until it was thought that all danger of


infection had passed. The usual dose was 30,000 units every 3 hours for a period of 5 days. Reactions were minimal, and limited to urticaria and fever. They were observed only occasionally and in every instance cleared up as soon as therapy was discontinued.

  Streptomycin did not become available until the vascular centers were in process of deactivation. It was not regarded as indicated in any instance of aneurysm or arteriovenous fistula then under observation.


  More than half (447) of the 814 vascular injuries in this series were associated with injuries of the nerves, bones, or soft tissues (Tables 10 and 11).  Frequently these injuries were of an extremely serious character and introduced problems of repair which required patience and ingenuity for their solution. These considerations are discussed in detail in the volumes of the history devoted to neurosurgery and orthopedic surgery and will not be repeated here.

  Other preoperative complications, with the exception of 2 cases of major causalgia in connection with arterial aneurysms, were confined to arteriovenous fistulas. They consisted of major causalgia in 7 cases; gangrene, which was seldom extensive, in 13 cases; gas gangrene in 2 cases, in both of which it was readily controlled; and Streptococci viridans bacteremia in 1 case.

  The single instance of Streptococcus viridans septicemia in this series is worth recording in detail for two reasons: (1) it originated in vegetations in a femoral arteriovenous fistula, and (2) it was cured by surgical excision of the fistula. A number of instances of septicemia of this origin have been recorded but, so far as can be determined from a survey of the literature, there are on record to date only three instances of cure of the bloodstream disease by surgical removal of the arteriovenous fistula.  The first of these cases, reported by Hamman and Rienhoff 3 in 1935, was an arteriovenous fistula of the external iliac vessels which had resulted from a rifle injury 17 years earlier. The second, reported by Touroff, Lande, and Kroop4 in 1942, was an arteriovenous fistula of the profunda femoris vessels combined with a saccular aneurysm which had resulted from a gunshot wound 9 years earlier. The third, reported by Lipton and Miller 5 in 1944, was a femoral arteriovenous fistula of 15 years duration which also had resulted from an accidental gunshot injury.

Case 2.  This 26-year-old soldier was wounded (after 18 months in the Southwest Pacific theater) on Leyte 21 December 1944.  He received penetrating grenade-fragment wounds in the right thigh, left arm, left chest, and left thigh. There was considerable bleeding from the right thigh wound, but it was controlled by a compression bandage. His wounds were debrided at a mobile surgical hospital and he was evacuated to a numbered general hospital. There, on about 18 January 1945, he called the attention of the attending medical

3 Hamman, L., and Rienhoff, W. F., Jr.:  Subacute Streptococcus viridans septicemia cured by excision of an arteriovenous aneurysm of the external iliac artery and vein. Bull. Johns Hopkins Hosp. 57: 219-234, Oct 1935.
Touroff, A. S. W.; Lande, H., and Kroop, L: Subacute Streptococcus viridans septicemia cured by excision of an infected traumatic arteriovenous aneurysm. Surg., Gynec. & Obst. 74: 974-982, May 1942.

5 Lipton, S., and Miller, H.: Streptococcus viridans septicemia-subacute bacterial endarteritis of an arteriovenous aneurysm. J.A.M.A. 126: 766-769, Nov 1944.


officer to a "purring" sensation in the right thigh which he had first noted on the third day after injury. Examination disclosed the classical signs of an arteriovenous fistula and he was evacuated by air to the Zone of Interior.

  The patient was admitted to the vascular center of Mayo General Hospital, 11 February 1945. His past history was noncontributory (cardiovascular history entirely normal) except for recurring attacks of tonsillitis. Within the last 14 months he had had 3 attacks of malaria. His wounds had healed uneventfully and he had no symptoms other than the "purring" sensation in the right thigh and a little numbness along the medial aspect of the knee.





He denied having had edema, cyanosis, coldness, pain, or weakness in the right lower extremity.

  A general physical examination revealed nothing remarkable except for the generalized yellowing of the skin commonly seen in patients who have had prolonged atabrine therapy.  The sclerae were normal in color. There was no tortuosity or thickening of the peripheral arteries and no evidence of cardiac enlargement.  The heart sounds were normal and there were no murmurs.

  Examination of the right thigh revealed an abnormal pulsation over the course of the femoral vessels in the proximal third, associated with a continuous thrill and a loud, continuous bruit which could be heard down to the knee and up into the right lower quadrant of the abdomen.  Both thrill and bruit could be obliterated by direct pressure over the femoral vessels about 15 cm. below the inguinal ligament. When this maneuver was carried out the pulse rate slowed from 76 to 52 beats per minute and blood pressure rose from 124 mm.of mercury systolic and 64 diastolic to 138 mm.of mercury systolic and 86 diastolic. When atropine was administered before compression of the femoral vessels, blood pressure changes were of the same magnitude as before but the pulse decreased by only 8 beats per minute. The dorsalis pedis and posterior tibial pulses were normal in both feet and the venous filling time was the same on both sides. In a room at a temperature of 22oC. the skin temperature of the toes on the right foot varied from 19o to 20 o  C. and on the left foot from 21o to 21.5o entigrade.  After 5 minutes of total arterial occlusion, and with the fistula completely compressed by digital pressure, reactive hyperemia began in the toes on the right foot in 20 seconds and was complete and full in the foot in 90 seconds.

  The electrocardiogram was normal except for a QRS complex of 0.12 seconds duration.  Roentgenograms showed no cardiac enlargement. The frontal cardiac area was 146 sq. cm. (Fig. 11A) as compared with a predicted frontal area of 143 sq. cm. for a man of the patient's weight and height.

  A diagnosis of femoral arteriovenous fistula, with collateral circulation adequate for surgical extirpation was made.  On 1 March 1945, the 17th day of hospitalization, and before a date for operation had been set, the patient had chills and fever which recurred in 48 hours. The clinical diagnosis of malaria was confirmed by a blood smear positive for Plasmodium vivax.  The recurrence was promptly controlled by atabrine and after the blood smears were reported as negative the patient was permitted to leave the hospital on furlough. When he returned 5 April 1945, he stated that shortly after he left the hospital he had had a recurrence of chills and fever, this time associated with a sore throat, and that a week later he had experienced abdominal soreness and a mild but persistent diarrhea. He had also suffered a moderate loss of appetite. He regarded the illnesses as a recurrence of malaria and had taken atabrine, but without results. He had been confined to bed for most of his 3-week furlough and had lost 16 pounds in weight.

  Examination on his readmittance to hospital revealed some tenderness in the region of the transverse colon and a barely palpable spleen. Stool examinations revealed infection by both hookworm and amebae. Appropriate treatment for these parasites resulted in prompt disappearance of the intestinal symptoms.

  The patient's temperature had been 102o F. when he was readmitted to the hospital, and though it did not rise above 99.2o F. for the next 3 days, it was higher the following day (Chart 18) and remained abnormal until after administration of penicillin. Atabrine was discontinued 14 April because blood smears had been reported negative for malaria parasites since 3 March. On 16 April the soldier had a severe chill followed by a temperature rise to 102.4 o F., and 3 days later a similar episode occurred though at this time the temperature rose only to 101.3o Fahrenheit.

  Following the chill and temperature elevation the patient became dyspneic and orthopneic and complained of pain and tenderness over the sternum. Within a few hours the chest pain shifted laterally and became localized in the left fourth interspace in the anterior axillary line. There was marked tenderness in this area. Examination of the chest re-


Figure 11. Roentgenograms showing change in heart size and evidence of pulmonary infarction. 

  A. 18 February 1945, heart, size is normal. Frontal cardiac area 146 square centimeters.
  B. 5 April, heart size has increased. Frontal cardiac are 168 square centimeters.
  C. 23 April, heart size difficult to estimate because pathologic changes in the lungs and elevation of diaphragm. There is increased density in the left lower lung field from pulmonary infarction.
  D. 4 June, 6 weeks after operation the heart size is normal, the lung field is clear.  (The method of Ungerleider and Gubner was used in computing the size of the heart.)


Chart 18. Temperature chart of patient with Streptococcus viridans septicemia preceding and following operation. 

vealed (for the first time) a soft, blowing, nontransmitted systolic murmur in the third interspace to the left of the sternum. The first heart sound was roughened and the apical impulse was displaced 3 cm. to the left of the midclavicular line. Over the left lower lobe posteriorly, breath sounds were absent, crepitant rales were heard, and there was marked dullness to percussion. The temperature was now 101.2 0 F., the pulse was 120 beats per minute, and the respirations were 20 per minute. On the basis of the clinical findings a tentative diagnosis of pulmonary infarction was made.

Roentgenograms  made on 5 April revealed increased density which obliterated the left lower lung field, a finding compatible with a diagnosis of pulmonary infarction. The frontal cardiac area, which had been 146 sq. cm. on 13 February,  had increased to 168 sq. cm. by 5 April-17 percent above normal (Fig. 11B). By 3 April, 4 days after the development of chest symptoms and signs, it was approximately 174 square centimeters. A more exact measurement was not possible because of the pulmonary changes (Fig.11C). On the same day an electrocardiogram showed sinus tachycardia with slight slurring of QRS in the third lead and a QRS complex of from 0.10 to 0.11 seconds duration.

  The persistence of intermittent chills and fever after a diagnosis of malaria had been eliminated, together with the clinical and roentgenographic signs suggestive of pulmonary embolism, pointed to a diagnosis of septicemia, the most likely source of which was thought to be vegetations within the femoral arteriovenous fistula. Blood taken for culture on 20 April was reported 24 hours later as positive for Streptococcus viridans. At this time leukocytes numbered 12,400 per cubic millimeter of blood, with a differential count showing 28 percent lymphocytes. The concentration of hemoglobin was 11.9 gm. per 100 cc. of blood, the prothrombin time 30 seconds, and the hematocrit reading 36 percent. There was no increase in tenderness to pressure over the arteriovenous fistula. No petechiae were observed.


Figure 12. Kodachrome of specimen removed at operation viewed from venous side of arteriovenous fistula. Note numerous vegetations along margins of fistula.


  The patient, whose expression was one of grave apprehension, was obviously acutely ill. Dilaudid (dihydromorphinone hydrochloride) was given for the relief of chest pain. Penicillin, begun on 19 April, was administered intramuscularly in doses of 25,000 units every 3 hours for 3 days and then in doses of 12,500 units every 3 hours for the next 3 days. Blood taken for culture on 21 April was again reported positive for Streptococcus viridans but repeated cultures for the next 6 days showed no growth of organisms. The patient became afebrile 48 hours after penicillin was begun and there were no further chills (Chart 18). On 24 April, immediately after an intravenous injection of 100,000 units of penicillin, the patient was operated upon.

  With the patient under spinal analgesia a longitudinal incision about 12 cm. long was made over the course of the femoral vessels beginning about 6 cm. below the inguinal fold. The fascia was divided and the sartorious muscle retracted. A great deal of scarring was observed in the region of the vessels. The femoral artery, which was about normal in size, and the femoral vein, which was dilated to a diameter of about 2 cm., were isolated above and below the fistula. A continuous thrill was present. The saphenous and the femoral nerves which were adherent to the vessels were dissected free without injury. The artery and vein were divided above and below the fistula and the stumps transfixed with silk. A large fistula almost 1.5 cm. in diameter communicating between the artery and vein was excised. The wound was closed in layers with interrupted silk sutures. The operative procedure was carried out without difficulty and the patient's condition was good throughout.

  Examination of the excised specimen (Fig. 12) revealed a large arteriuovenous fistula almost 1.5 cm in diameter.  The margins of the fistula were studded with minute small vegetations. Cultures made directly from these vegetations revealed a heavy growth of Streptococcus viridans, although at the time of operation blood cultures had been reported as negative for these organisms. Histologic examination revealed that these vegetations were literally studded with cocci (Fig. 13).

  The patient's postoperative course was uneventful. He remained afebrile, and his pulse and blood pressure were within normal range. The bruit and thrill were no longer present in the area of the fistula and the temperature and color of the right leg were normal. The patient was allowed out of bed on the fourth day. For a few days thereafter he complained of a pulling sensation in the left chest on deep inspiration but he had no dyspnea and orthopnea and this minor discomfort rapidly disappeared. The signs of infarction in the left lower lobe gradually disappeared and 6 weeks after operation a clinical and roentgenologic examination revealed no abnormalities in this area. The size of the heart returned to normal (Fig. 11D).

When the patient was first discharged from hospital he experienced  slight edema of the right leg after it had been in a dependent position for a long  period.  When examined 5 July 1945, no edema was present and he stated that there had been no evidence of it for a week; at this time he was wearing an elastic stocking. His only complaint was a sense of fatigue in the calf after he had walked for half a mile. The feet showed no abnormal color changes. The posterior tibial and the dorsalis pedis pulses were present on both sides but were reduced in volume on the right. In a room in which the temperature was 25.5o C. the temperature of the toes on both feet varied between 29 and 30 Centrigrade. The oscillometric reading in the distal thigh was 2.5 at 70 mm. of mercury on the right and 6 at 120 on the left. In the calf the oscillometric value was 3 at 60 mm. of mercury on the right side and 7.5 at 90 on the left. In the ankle it was 2 at 70 mm. of mercury on the right and 5 at 80 on the left. The pulse was 56 beats per minute and the blood pressure 128 mm. of mercury systolic and 70 diastolic.

  The patient was reexamined about 3 weeks later. The elastic support had not been worn in the interim and he had experienced no edema. After a walk of about 1 miles he stated he had a sense of tightness in the right calf, but the discomfort was not great enough to make him cease walking. There had been no recurrence of fever, chills, or pulmonary and cardiac symptoms. He looked well and was apparently in excellent general condition.


Figure 13. Photomicrograph of portion of excised arteriovenous fistula showing two bacteria studded vegetations (H. and E. stain, high power magnification). The dark-staining areas are masses of cocci. Cultures of these vegetations yielded a pure growth of Streptococcus viridans.

  Comment.  Transient Streptococcus viridans bacteremia is not infrequent in the course of upper respiratory or sinus infections, or following the extraction of infected teeth. Presumably such transient bacteremia is the original source of infection in cases of subacute bacterial endocarditis. It is therefore significant that the patient had had recurrent attacks of tonsillitis since childhood and had had a severe attack associated with pharyngitis and accompanied by chills and fever about a month before the diagnosis of septicemia was made. It is quite conceivable that this attack resulted in infected vegetations within the arteriovenous fistula.

  Although this patient had a multiplicity of other diseases (tonsillitis and pharyngitis, malaria, hookworm, and amebic infection) the recognition of Streptococcus viridans infection was not long delayed. The diagnosis was relatively easy despite the coincident malaria because of the severe chills and fever, the signs and symptoms of pulmonary infarction, and the presence of an arteriovenous fistula. The findings on blood culture merely confirmed the clinical suspicion.

  The rapid increase in cardiac size which occurred in this case (Fig. 11A-C) was probably the direct result of the altered circulatory dynamics produced by the presence of the large femoral arteriovenous fistula. Similar increases


were observed in other patients with similar lesions during the period of observation prior to excision of the fistula.

This case illustrates the necessity of bearing in mind the presence of arteriovenous fistula when looking for a source of infection in instances of Streptococcus viridans septicemia. It also exemplifies the dramatic cure which can be achieved in such instances by surgical removal of the defect. The patient was treated with penicillin over a 6-day period and the blood stream was thus rendered sterile prior to operation, but the heavy growth of Streptococcus viridans obtained from within the fistula after it had been removed is evidence that the apparent cure would not have been permanent.


  The criteria for a successful operation were considered fulfilled if the objective evidences of the lesion which had been present before operation had disappeared after operation and there was no evidence of recurrence during the postoperative period of hospitalization. These standards were deemed adequate since the period of hospitalization is much longer in military than in civilian practice and since any recurrence of aneurysmal lesions becomes evident rather promptly.

Of the 801 operations performed at the vascular centers during World War II (in 13 instances in this series of 814 cases, surgery was unnecessary because spontaneous cures occurred) there were only 4 fatalities, 10 failures, and 36 instances of complications following operation. Much of the success can be attributed to the youth and generally good circulatory status of the patients in whom these lesions occurred, but other factors contributed and are worthy of note. The thorough preoperative examination to which each patient was submitted and the detailed evaluation of his vascular status established the fact that the operation was being performed at the optimum time on a patient whose vascular status was as good as it could be made. Furthermore, the surgeons who performed the operations in these centers were men preeminent in the field of vascular surgery and had at their disposal the appropriate precision instruments and other necessary equipment.  

It is possible, too, that the adjunct use of certain measures, particularly anticoagulant therapy (see Chapter X) and sympathectomy (see Chapter XI) played some part in the results. On the other hand, neither of these methods was employed at one center (Ashford) where the results were quite as good as those obtained at the other centers. The conclusion seems warranted, therefore, that while these methods may be useful in selected cases, they are certainly not indispensable.


  The number of postoperative complications in this series (36) is remarkably small when weighed against the number of operations (801) performed for aneurysms and arteriovenous fistulas. These complications included:


  1. Nineteen instances of wound infection: 12 in arteriovenous fistulas and 7 in aneurysms. In 1 arterial aneurysm the infection progressed to necrosis.

  2. Four instances of hemorrhage: 3 in arteriovenous fistulas and 1 in an aneurysm.

  3. Five instances of hematomas in the wound: 4 in arteriovenous fistulas and 1 in an aneurysm.

  4. One instance of mediastinal abscess in an arteriovenous fistula.

  5. Three instances of major causalgia: 2 in aneurysms and 1 in an arteriovenous fistula.

  6. Two instances of hemiplegia: 1 in an aneurysm and 1 in a fistula.

  7. Two instances of gangrene, both in arteriovenous fistulas.


  There was immediate evidence of failure following 10 operations performed at the vascular centers. The failures occurred in

  1. Two arterial aneurysms, both in the subclavian artery.

  2. Eight arteriovenous fistulas: 1 in the aorta and vena cava, 4 in the carotid, 1 in the superficial temporal artery, and 2 cirsoid aneurysms.


  Four deaths occurred in the 801 cases treated surgically at the vascular centers. The distribution of the lesions was as follows:

  1. Aneurysm of the deep cervical artery.

  2. Aneurysm of the subclavian artery.

  3. Arteriovenous fistula between subclavian artery and vein.

  4. Arteriovenous communication between the internal carotid artery and the cavernous sinus.

  The details of these cases are as follows:

  Case 3.   A 22-year-old soldier was operated on 4 months after injury for a large aneurysm on the left side of the neck which was causing a good deal of pain.  Intratracheal anesthesia was used. After the carotid artery had been isolated and a tape had been placed about it, temporary occlusion of the vessel did not end the pulsation in the area.  Compression of the vertebral and subclavian vessels was equally ineffective.  It was apparent from these phenomena that some one of the small branches of the carotid artery was the site of the aneurysm, but the sac was so large that the vessel could not be isolated.

  A cautious attempt was therefore made to free the sac with the idea of isolating the vessel when it was elevated.  During this attempt, however, the sac was accidentally entered.  The opening into the artery could easily be occluded by introduction of the finger, but each time that suture transfixion of the opening was attempted a considerable amount of blood was lost.  Although the patient had been receiving plasma and blood throughout the procedure, he was now discovered to be in shock and in a matter of moments, pulseless.  His blood pressure could not be determined and for several minutes breathing had to be maintained by artificial respiration.

  It was finally possible to suture the artery within the aneurysm and there was some improvement in his condition following the introduction of more blood, but he never regained


consciousness. Convulsions ensued and death occurred 24 hours after operation. Necropsy revealed diffuse acute and anoxic edema and necrosis of the middle layer of the right cerebral cortex. The vessel involved proved to be the deep cervical artery.

Comment.  This death should not have occurred. An inexperienced anesthetist gave the anesthetic and the surgeon had no warning of the patient's condition until it had become irreversible. Had the proper warning been given, the operation could probably have been terminated quickly and without further loss of blood by packing the sac and suturing the aneurysm, or bleeding could have been controlled by pressure until the patient's status permitted the loss of more blood while another attempt was made to suture the vessel.

  Case 4. A 24-year-old soldier was operated on 11 months after injury for a large subclavian arterial aneurysm which was partly in the neck and partly in the mediastinum.  Intratracheal anesthesia was used. Excellent exposure was obtained by resetting the inner portion of the clavicle and splitting the sternum.  The subclavian artery was ligated proximal and distal to the opening in the sac.

  Although there was no further bleeding, the sac continued to pulsate feebly and the introduction of a needle revealed arterial blood under pressure.  This blood was obviously coming from the vertebral artery which was the only vessel trapped between the two ligatures.  The aneurysm was so large that access to the vertebral artery in its free portion was completely prevented.  The sac was therefore gradually freed without special difficulty.  When the posterior aspect had been dissected free almost to the point of the opening, the sac was quickly opened and back bleeding from the vertebral artery readily controlled.  At this time the wound was dry and the major portion of the sac could be excised.

  The patient had received blood throughout the operation and his systolic blood pressure had never fallen below 90 mm.of mercury.  Except that he was notably drowsy after operation, his condition seemed good.  The blood pressure was constantly in the neighbor hood of 130 mm. of mercury systolic and 80 diastolic and the pulse, which was of good volume, in the neighborhood of 100 beats per minute.

  The night after the operation the patient woke suddenly and complained of a severe occipital headache.  Shortly afterward he became stuporous and then comatose, and death occurred about noon the following day.  There was a rapid rise of both temperature and pulse rate during the night and just prior to death they were greatly elevated.

  Necropsy revealed encephalomalacia of the left cerebellum with some diffuse old subarachnoid hemorrhage and scarring.  No thrombus could be demonstrated in the vertebral artery or in the circle of Willis.  The right vertebral artery and both carotid arteries were intact.

  Comment. The course of events in this case was unexpected and cannot be explained. In retrospect, it seems that the management of the case was correct and that the operative procedure was properly performed.

  Case 5.   A 20-year-old paratrooper was stabbed with a knife in the left infraclavicular region.  Bleeding, which was profuse, was controlled by pressure but neither formal debridement nor exploration was carried out; the wound was merely closed with sutures.  Slight drainage from the wound had continued for approximately 3 weeks after the injury. Then complete healing occurred.  Within a short time after the injury the patient became conscious of a buzzing sensation in the left upper chest and in the region of the shoulder.

  Two and one-half months after injury he was admitted to a vascular center in the Zone of Interior.  Examination given at the time of admittance revealed the typical symptoms and signs of an arteriovenous communication in the first part of the subclavian vessels. There were no trophic changes in the left upper extremity, also no evidence of cardiac decompensation.


  At operation, which lasted 8 hours, the left clavicle was resected, and the median half of the bone removed, the first portion of the subclavian artery was then identified and the fistula located between the subclavian artery and vein.  It lay so near the aortic arch that its isolation was extremely difficult and on several occasions brisk bleeding was encountered, though it could always be satisfactorily controlled by means of sutures or ligatures.

  Convalescence was satisfactory until the fourth postoperative day when it became apparent that the wound was infected. Six days later a secondary hemorrhage occurred from the wound; it was finally controlled in the operating room by suture and ligature.  For the next 5 days recovery was uneventful.  Then bleeding from the wound recurred.  This time the attempt to control the hemorrhage by surgery was ineffective for all the structures were so friable that isolation of the bleeding vessel was not possible and the patient died on the operating table.  A continuous transfusion of whole blood was given from the time he was placed on the operating table until death occurred.

  Comment.  The wound infection which caused the hemorrhage must be considered responsible for this death. Penicillin had been administered immediately after the first operation but despite this precaution infection developed.

  Case 6. A 35-year-old soldier sustained a wound of the head and neck. The injury was followed by typical symptoms and signs of an arteriovenous communication between the internal carotid artery and the cavernous sinus.  A bruit was audible to the patient and was particularly prominent when he was in bed at night.  The left eye became prominent and its conjunctiva edematous and engorged.  Six months after injury he was admitted to a vascular center in the Zone of Interior.

  The left common carotid artery was ligated and as a result there was diminution of the bruit and decreased prominence of the left eye. Vascularity of the conjunctiva also became less.  Since this operation did not eliminate the bruit completely, a second operation was performed at which the internal carotid artery was isolated and divided.  The bruit was absent for about 2 weeks after this operation, then recurred at occasional intervals.

  When the patient was on a furlough he had an episode of unconsciousness which was believed to be the result of a cerebral accident.  When he returned to the hospital, the left eye had again become prominent, though not as prominent as when he was first seen, and the bruit was again audible.  It was then decided that intracranial ligation of the affected artery must be attempted.

  The internal carotid artery was excellently exposed by retraction of the left frontal lobe and was isolated at its point of exit from the cavernous sinus.  Two tantalum clips were applied to it.  Operation was carried out without incident and without undue loss of blood but the patient did not respond and died on the third postoperative day.  At necropsy one of the tantalum clips was found on the carotid artery and the other on the posterior communicating branch.

  Comment. Death in this case can be attributed to an insufficient arterial blood supply to the cerebral tissues.


  Since multiple vascular injuries were so frequent in World War II it was to be expected that certain casualties in whom sequelae from these injuries developed would sometimes present multiple lesions. Of the 814 aneurysms and arteriovenous fistulas observed at the vascular centers in the Zone of Interior, 46 occurred in 20 patients (Table 12).



  Fifteen patients had 2 lesions each, 4 had 3 lesions, and 1 patient had 4 lesions. The 20 patients thus presented a total of 46 separate vascular sequelae of their combat-incurred injuries.

  In 15 patients the multiple lesions were arteriovenous fistulas; in the 5 remaining the distribution was as follows: 2 had 2 aneurysms, 2 had an aneurysm and an arteriovenous fistula, and 1 patient had an aneurysm and 2 arteriovenous fistulas.

  Because of its unusual character the case report of the patient with 4 arteriovenous fistulas is reported in detail.

  Case 7.  A 22-year-old sergeant was wounded on 16 August 1944 in Southern France by an exploding land mine. He received a compound comminuted fracture of the right fifth metatarsal bone in addition to multiple soft-tissue injuries of the back, thighs, legs, and feet.  Bleeding was minimal.  The wounds were dressed and casts applied to both legs.

  The patient was transferred by ship to a station hospital in Italy. The cast was removed from the left leg on 1 September and from the right leg on 1 November. The patient was found to have some weakness of extension of the right foot, paralysis of the right toe extensors, hyperesthesia of the dorsal surface of the right toes, and hypesthesia in the distribution of the left saphenous nerve. He was evacuated to the Zone of Interior and on 31 October admitted to a general hospital.

  Examination at the general hospital confirmed the neurologic findings in the right foot. The fracture was found to be healed. This examination also disclosed signs of the presence


of an arteriovenous fistula located anteriorly in the distal portion of the right leg. The patient was therefore transferred to a vascular center.

  On 7 November 1944 when admitted to the vascular center at Mayo General Hospital he had few complaints.  He stated that he had some weakness of extension of the right foot and toes, and occasionally a slight burning pain on the dorsal surface of this foot, but the hyperesthesia had disappeared.  Along the left internal malleolus he still had some residual hypesthesia.  He said the right foot was somewhat warmer than the left and that sweating was excessive in both feet.  He denied having had color changes, venous difficulties, trophic changes, precordial distress, dyspnea, or cyanosis.

  Physical examination revealed 150 wounds.  All of them were well healed and most were very small.  The majority were scattered over the legs, feet, and thighs. Roentgenologic examination of the extremities revealed numerous small metallic fragments scattered throughout the soft parts in these areas.

  Neurologic findings were limited to a small area of hypesthesia along the left internal malleolus, loss of extensor power in the right toes, and some weakness of extension and eversion of the right foot.

  All the toes of both feet became slightly cyanotic when in a dependent position. The peripheral arteries were neither tortuous nor thickened, the veins neither dilated nor thrombosed. Those in the right foot filled in 5 seconds, in the left after 15 seconds. The veins on the right side were somewhat better filled.  Sweating of both feet was moderate.  Slight pitting edema was present in the right foot and leg. There were no trophic changes.

  About 10 cm. above the right internal malleolus, between the tibia and fibula anteriorly, an abnormal pulsation was present associated with a continuous thrill and a bruit audible down into the foot and up to the knee.  Direct pressure over the anterior tibial vessels in this area obliterated the thrill, pulsation, and bruit, and produced a drop in the pulse rate from 72 to 67 beats per minute.  When the arteriovenous fistula in the region of the right anterior tibial vessels was obliterated by direct pressure, a loud, continuous bruit was audible posteriorly in the calf about 20 cm. above the malleolar area.  There was no thrill in this area and the bruit could be obliterated by rather broad pressure against the calf muscles proximal to the point of maximum bruit.  About 20 cm. above the malleoli on the posterior medial aspect of the left leg a continuous thrill could be felt and a continuous bruit was heard.  The bruit could not be obliterated by pressure in this area but ceased when the popliteal artery was compressed.

  Blood pressure varied from 126 mm. of mercury systolic and 72 diastolic to 116 mm. of mercury systolic and 76 diastolic. It did not change when all the pulses in both legs were occluded, but the pulse rate dropped at this time from 84 to 68 beats per minute.

  Skin temperatures of the toes on both feet were about equal. On one occasion, in a room at 22oC., the temperature of the toes on each foot varied from 22o to 23o C. and on another occasion from 32.5o to 35.5o Centigrade. When the reactive hyperemia test was carried out with the anterior tibial artery occluded, a good flush began in 25 seconds and was complete in 50 seconds.

  It was concluded that the patient had 3 arteriovenous fistulas, 1 involving the right anterior tibial vessels and 2 probably involving the right and left posterior tibial vessels.

The patient was given a brief furlough at the conclusion of the examinations described. On his return he complained of some discomfort in both calves on walking and some swelling of both ankles. Examination at this time revealed moderate edema of both ankles. The circumference at this level was 2.5 cm. greater on the right than on the left side.

On 29 December 1944 the anterior tibial vessels of the right were explored and a double fistula found between the anterior tibial artery and two anterior tibial veins. The veins lay on either side of the artery.  Each of these communications was about 4 mm. in diameter. The involved vessels were ligated proximally and distally, and the fistula excised. The postoperative course was smooth and the patient showed no evidence of vascular insufficiency at any time.


  A second operation, this time on the left leg, was done 8 January 1945 through a posterolateral incision. The posterior tibial and the peroneal veins were found to be considerably dilated and to communicate through two channels. The posterior tibial vein itself communicated through a fistula about 5 mm. in diameter with the posterior tibial artery which also exhibited a saccular aneurysm on its wall opposite to the communication. This aneurysm was well organized and approximately 1 cm. in diameter. The lesions were excised and recovery was uneventful.

  A third operation was carried out 12 February 1945, this time to locate the second aneurysm in the right leg. The posterior tibial artery and vein were exposed through a posteromedial incision in the calf.Both were found to be of normal size and appearance and no thrill could be palpated in this area. Palpation beneath the gastrocnemius, however, revealed a continuous thrill on the posterior aspect of the soleus muscle somewhat medial to the operative wound. The patient's muscles were so large and well developed that access to the involved area was not possible through the original incision and a second longitudinal incision was made posteriorly and directly over it. The gastrocnemius was divided in the direction of its fibers and the soleus exposed. Within the body of the latter muscle was found a small artery, about the diameter of a matchstick, which communicated with a vein about twice its size through a small saccular aneurysm 1 cm. in diameter. These vessels appeared to be muscular branches from the posterior tibial vessels.  Artery and vein were ligated proximally and distally and the aneurysm and communicating vessels exised.

  When these procedures had been completed the thrill and bruit originally present in this area were no longer present but a fairly loud and continuous bruit could be heard somewhat lateral and cephalad to the incision.  Again the bulk of the muscles prevented access to the affected area through the original incision and a third longitudinal incision was made somewhat higher along the posterolateral aspect of the calf.  When the fibers of the gastrocnemius muscle were separated and the soleus exposed, a small artery was found in the body of that muscle which communicated through a fistula with a vein dilated to the diameter of almost I centimeter. These vessels were thought to be muscular branches of the peroneal vessel. The vessels communicating with it were quadruply ligated and the fistula excised.  Recovery was uneventful.  A diagrammatic sketch of the fistulas found is shown in figure 14.

  When his convalescence from the third operation was completed, the patient was able to walk at a reasonably brisk pace for more than a mile with no resultant discomfort in the calves of his legs.  He regained fairly complete power of extension of the right foot; likewise, function of all the extensors of the toes except that of the great toe.

  The extremities no longer presented any edema and there was no substantial difference in their circumferences.  Both feet were warm.  Clinically the temperature appeared to be the same on both sides, but when a thermocouple was used it was found that the toes on the right foot tended to be somewhat warmer than those on the left.In a room in which the environmental temperature was 23o C., the temperature of the toes on the right foot varied from 32 to 33 C., and that of the toes on the left foot from 28o  to 30 Centigrade. The pulsations of the posterior tibial artery were normal on the right side but absent on the left side. Both dorsalis pedis pulses were normal.  Oscillometric readings at the ankles were 4.5 units on the right side and 3.5 units on the left at 90 mm. of mercury.

After operation the pulse was 72 beats per minute and the blood pressure 108 mm.of mercury systolic and 72 diastolic. The cardiac size was unchanged. The predicted frontal cardiac area for a man of this patient's height and weight was 124 square centimeters. The actual measurements on admittance and after each of the 3 operations were 112 square centimeters.  Electrocardiograms taken before each of the operations showed no diagnostic deformity.

  The patient was returned to duty after a neurosurgical consultant had agreed that the peroneal paralysis had so nearly disappeared that no operative procedure was indicated.


Figure 14. Diagrammatic showing of location and general appearance of the four arteriovenous fistulas present following injury by land mine. No effort has been made, to draw the lesions to scale.

Comment. Careful palpation for thrill and auscultation for bruit should be an essential step of the final stages of operative procedures for the cure of arteriovenous fistulas. This case illustrates the wisdom of this precaution particularly in the battle casualty whose wounds are multiple and relatively small.