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 XI



Arterial Aneurysms and Arteriovenous Fistulas

Sympathectomy as an Adjunct Measure in Operative Treatment

Harris B. Shumacker, Jr., M. D.

Part I. Preoperative Sympathectomy and Sympathectomy Coincidental With Vascular Surgery

  Interruption of the sympathetic pathways as a preliminary to surgical extirpation of an aneurysm was introduced in 1934 by Gage1 and in 1935 by Bird.2 They were unaware of each other's activities. Gage used alcohol injections and Bird ganglionectomy.

  Bird, although he was impressed with the apparent improvement of the collateral circulation achieved by this method, presented his contribution with definite reservations because he was hopeful that the pavex boot, which lead recently been introduced, might make unnecessary any other effort to foster development of the collateral circulation. Gage, on the other hand, was convinced that sympathetic interruption would continue to occupy a prominent place in the surgical management of aneurysms. He and Ochsner 3 subsequently reported additional

successful experiences with this procedure.

  After 1934 when Gage's original contribution appeared, a number of other observers reported the use of sympathectomy as an adjunct measure in the management of aneurysms and recommended that it be performed either before the vascular operation or as part of the vascular procedure.4

1 Gage, I. M.: Mycotic aneurysm of common iliac artery; sympathetic ganglion block as aid in development of collateral circulation in arterial aneurysm of peripheral arteries; report of a case. Am. J. Surg. 24: 667-710 Jun 1934.

2 Bird, C. E.: Sympathectomy as a preliminary to the obliteration of popliteal aneurisms; with a suggestion as to sympathetic block in cases of ligature, suture, or thrombosis of large arteries. Surg., Gynec. & Obst. 60: 926-929, May 1935.

3 (1) Gage, M., and Ochsner, A.: The prevention of ischemic gangrene following surgical operations upon the major peripheral arteries by chemical section of the cervicodorsal and lumbar sympathetics. Ann. Surg. 112: 938-959, Nov 1940.

(2) Gage, M.: The development of the collateral circulation in peripheral arterial aneurysms by sympathetic block. Surgery 7: 792-795, May 1940.

4 (1) Veal, J. R.: The value of sympathetic interruption following the surgical repair of peripheral aneurysms. M. Ann. District of Columbia 9: 227-230, 1940.

(2) Richards, R. L., and Learmonth, J. R.: Lumbar sympathectomy in treatment of popliteal aneurysm. Lancet 1: 383-384, 28 Mar 42.

(3) Pemberton, J. deJ., and Black, B. M.: Surgical treatment of acquired aneurysm and arteriovenous fistula of peripheral vessels. Surg., Gynec. & Obst. 77: 462-470, Nov 1943.

  (4) Kirtley, J. A., Jr.: Arterial injuries in a theater of operations. Ann. Surg. 122: 223-234, Aug 1945.

  (5) Freeman, N. E.: Secondary hemorrhage arising from gunshot wounds of the peripheral blood vessels. Ann. Surg. 122: 631-640, Oct 1945.

  (6) Harbison, S. P.: Experiences with aneurysms in an overseas general hospital. Surg., Gynec. & Obst. 81:128-137, Aug 1945.

  (7) Albright, H. L., and Van Hale, L. A.: Traumatic aneurysms; a study of 43 cases in an overseas general hospital. Surgery 20: 452-477, Oct 1946.

  (8) Warren, R.: War wounds of arteries. Arch. Surg. 53: 86-99, Jul 1946.


  All the reported cases were successful; the circulation apparently proved adequate after operations which necessitated ligation of the affected artery. The universally successful results reported in these contributions created the impression, perhaps contrary to the intent of the authors, that sympathectomy provides an almost impregnable defense against ischemic disaster in vascular lesions treated by ligation of main arteries.

  In the vascular centers during World War II the general usefulness of sympathectomy was confirmed, but at the same time certain limitations were revealed. At the vascular center of Ashford General Hospital sympathectomy was used only occasionally in the management of aneurysms; at the vascular centers of both DeWitt and Mayo the experience with sympathectomy was extensive. At the last-named hospital a study was made of sympathectomy as an adjuvant in the operative treatment of vascular lesions. The purpose of this study was twofold: to determine the merits and limitations of sympathectomy, and to formulate a logical plan for its use.

  There is ample experimental and clinical evidence to support the view that sympathectomy is of real help in maintaining an efficient circulation in various conditions in which the continuity of the blood flow through important arterial trunks has been interrupted. There are two concepts regarding the mechanism by which this efficient circulation is achieved:

  1. Sympathectomy ensures the fullest use of the existing collateral circulation because it eliminates vasoconstrictor impulses and thus contributes to the maintenance of near-maximum circulation. This concept is well established.

  2. Sympathectomy brings about more rapid growth of new collateral channels because it reduces peripheral resistance to the blood flow (or perhaps the new channels are brought about by some other, mechanism).This explanation is still hypothetical and will require considerable investigative support before it can be considered factual.

  The point of view might readily be defended that sympathectomy should be performed upon every patient with an aneurysm or fistula if cure might entail ligation of an important arterial stem. Clinical experience during World War II demonstrated, however, that large numbers of aneurysms and fistulas can be extirpated without serious ischemic difficulties provided (1) the data secured by careful testing of the collateral circulation are used as a guide in the selection of the proper time for operation, and (2) the operation is so carried out that no collateral blood vessels are needlessly sacrificed. Some investigators hold that sympathectomy should be used almost uniformly in the surgical management of aneurysms and fistulas, others take the position that the procedure should rarely, if ever, be employed. At the vascular center of Mayo General Hospital a practice between these two extremes was followed Sympathectomy was utilized whenever it might reasonably be expected to result in benefit to the patient. Undoubtedly it was carried out more often


than was actually necessary, as, for example, to prevent gangrene. In spite of this possibly excessive use, it may be said that if clear-cut benefit did not follow in every instance in which it was employed, at least no harmful results were noted.


  This analysis is based upon 75 sympathectomies (Table 35) performed on 288 patients either before or in the course of operations for aneurysm or arteriovenous fistula at the vascular center of Mayo General Hospital during World War II. 5 In an additional 13 patients 2 sympathectomies were carried out in anticipation of operation, but no operative treatment proved necessary because in all 13 spontaneous cure of the lesion by thrombosis occurred. One other sympathectomy might properly be added to this group. It was performed on a patient in whom physical findings pointed to a diagnosis of aneurysm, but surgical intervention established the fact that the signs were caused by costoclavicular compression of the subclavian artery (see Chapter IV).


  There were no deaths and no serious complications in these 302 patients.

  Indications. In the 77 patients upon whom sympathectomy was performed there were 7 chief indications for interruption of the sympathetic pathways (Tables 36-42):

  1. Evidence of poor collateral circulation in patients with lesions of sufficiently long duration to have produced, under ordinary circumstances, a fairly efficient collateral circulation.

5 Two hundred eighty-eight aneurysms and fistulas were treated surgically at this hospital, 63 lesions were treated in which surgery had been performed prior to transfer to this hospital, and 13 lesions were observed in which spontaneous cure occurred. These 364 lesions occurred in 351 patients (9 patients presenting a total of 22 lesions). Unfortunately In the distribution of patients and lesions, the differentiation became obscured and in this discussion the designation "patient" will be used for "patient/lesion" since so few patients presented multiple lesions.


  2. The presence of an inadequate collateral circulation in association with peripheral nerve injuries requiring surgical exploration. The management of this type of case was a problem. Early surgical treatment of peripheral nerve injuries was desirable because it resulted in better return of function than the late treatment of such lesions. On the other hand, it was hazardous to explore the nerve before the aneurysm or fistula had been treated. Every effort was made therefore to increase the efficiency of the collateral circulation as promptly as possible, particularly because it was usually feasible to carry out both the vascular operation and the neurosurgical procedure at the same time.

  3. The impossibility of testing the collateral circulation in certain patients in whom an important arterial trunk appeared to be involved. This indication was not often a valid reason for sympathectomy, but in-an occasional patient the affected artery could not be compressed at the site of the fistula or aneurysm and accurate preoperative testing of the collateral circulation was impossible.

  4. The association of the aneurysm or fistula with an ischemic lesion.

  5. Intense local vasospasm in the affected limb, or a rather severe generalized vasospastic disorder.

  6. The loss of one or more major arteries by previous injury or operation in a limb in which the cure of an aneurysm or fistula was likely to necessitate ligation of other important arterial channels.

  7. Causalgia in the affected limb which could be relieved temporarily, but not cured, by sympathetic blocks.

  Technique. At this center interruption of vasoconstrictor impulses was achieved by surgical resection rather than by paravertebral alcohol injection. Operative sympathectomy was preferred for two reasons (1) surgical interruption has a more lasting effect, and (2) neuritic pains sometimes constitute distressing sequelae of alcohol injection. Twenty-four of the 77 sympathectomies involved dorsal, and 53 lumbar, operations. Lumbar sympathetic ganglionectomy was performed through an anterior extraperitoneal muscle-splitting incision. These operations were carried out with the aid of spinal analgesia. Dorsal sympathectomy was accomplished by the preganglionic operation devised by Smithwick.6 It was carried out with the patient under intratracheal gas-oxygen-ether anesthesia.

  Anatomic Distribution of Operations. The majority of the 77 sympathectomies were inpatients with lesions of the larger peripheral arterial stems. If we exclude the 2 cases in which spontaneous cure without vascular surgery followed sympathectomy (Cases 27 and 47), it will be seen that the 75 sympathectomies represent 26 percent of the 288 patients with aneurysms or arteriovenous fistulas treated at this center by surgical methods.(Table 35.)One hundred thirty-six of these patients had lesions of the femoral, popliteal, or axillary arteries. It was upon this group that most of the sympathectomies were

6 Smithwick, R. H.: The rationale and technic of sympathectomy for the relief of vascular spasm of the extremities. New England J. Med. 222: 699-703, Apr 1940.


performed (55 sympathectomies).In the 44 patients with lesions of the sub clavian and brachial arteries, sympathectomy was performed only 9 times. In the remaining 108 patients with lesions of arteries not included in the foregoing breakdown, only 11 sympathectomies were performed. Moreover, as will be pointed out shortly, sympathectomy was performed upon a number of patients in the miscellaneous group because preoperative study had erroneously indicated that an important artery near the affected vessel was involved in the lesion.

  Proportionately, sympathectomy was performed more often upon patients with arterial aneurysms than upon those with arteriovenous fistulas (Table 35). The discrepancy is more apparent if only those patients with lesions of axillary, femoral, and popliteal lesions are considered, since it was upon these patients that sympathectomy was most frequently done. The relative percentages for axillary, femoral, and popliteal aneurysms were, respectively, 46.2, 66.7, and 78.6, in contrast to percentages for arteriovenous fistulas of 33.3, 32.0, and 34.1. Although these differences are primarily a reflection of the tendency to poorer collateral circulation in instances of arterial aneurysms, they can also be explained by the fact that associated nerve injuries requiring operative treatment were twice as common in arterial aneurysms as in arteriovenous fistulas. The difference cannot be explained by a higher incidence of involvement of major arterial stems in arterial aneurysm than in arteriovenous fistula; 46.3 percent of the arterial aneurysms operated on in this series involved the subclavian, axillary, femoral, and popliteal arteries, against 52.9 percent involvement of these arteries in arteriovenous fistulas. This is not a significant difference.


  The results of sympathectomy (Tables 36-42) can best be analyzed if they are considered in relation to the chief indications for the procedure. It should, of course, be borne in mind that in a number of patients sympathectomy was decided upon not for a single reason but for a combination of reasons.

  In all of the patients included in this analysis the status of the collateral circulation was investigated by the repeated use of a number of tests and observations (see Chapter II) . Chief reliance, however, was placed upon the reactive hyperemia test devised by Matas.

Poor Collateral Circulation

  Poor collateral circulation was the chief indication, or a major indication, for sympathectomy in 38 of the 77 patients upon whom it was performed (Tables 36 and 37). There were, however, other indications in this group. In 16 patients, for instance, associated peripheral nerve lesions required operative treatment. In a number of patients severe pain or pronounced vasospasm was present and in 1 patient superficial gangrene was the indication.


  In all 38 patients in this group, intermittent proximal occlusion of the involved artery had been practiced without apparent improvement in the collateral circulation, though in only a few instances had this method been given a prolonged trial.







  Of this group of 38 patients, sympathectomy was carried out when the lesion had been present 5 months or more in 18, 3 months but less than 5 months in 14, and less than 3 months in 5. In 1 patient (Case 30) the aneurysm was caused by medionecrosis and it was not possible to determine the time interval. In the 5 patients upon whom sympathectomy was performed even though the lesion had been present for less than 3 months, circumstances existed which made it seem advisable to undertake it without further delay.

  Two of these 5 patients (Cases 6 and 17) had severe pain. In 1 of them a popliteal aneurysm had ruptured subcutaneously and it was obvious that a further increase in its size might necessitate operation at any moment. The other had severe causalgia associated with a femoral arteriovenous fistula, a


saccular aneurysm and peroneal paralysis, in addition to a compound comminuted fracture of the femur, intense vasospasm of the foot, an ulcer of the heel, and superficial gangrene of the toes. His pain had been so severe that narcotics had been used to the point where there was real danger of addiction. In both of these patients pain was diminished or relieved shortly after sympathectomy had been performed, and tests thereafter showed the collateral circulation to be satisfactory.

  A third patient (Case 19), with a popliteal aneurysm of 2 months duration, had extremely poor collateral circulation and complete peroneal paralysis. The collateral circulation became adequate soon after sympathectomy and early treatment of the aneurysm and nerve lesion was possible. The fourth patient (Case 27) had a femoral aneurysm and the fifth (Case 28) a popliteal aneurysm. Both of these lesions had been present for 2 months and in both patients the collateral circulation was extremely poor. Although the tests did not show the collateral circulation to be entirely satisfactory in either patient after sympathectomy, they did show some improvement. Spontaneous cure occurred in 1 patient, and in the other success followed operative treatment 2 months after sympathectomy.

  Twenty-four of the 38 patients upon whom sympathectomy was performed because of inadequate circulation showed evidence of satisfactory collateral circulation subsequent to the operation, and in the majority, tests showed it to be adequate very shortly after sympathectomy. There was no great difference in this respect between those patients with arterial aneurysms and those with arteriovenous fistulas. In 8 of 14 patients with aneurysms and in 16 of 24 with fistulas, evidence of good collateral circulation became apparent after sympathectomy according; to the reactive hyperemia test and other tests used for this purpose.

  In these 24 patients (see Table 36) the vascular operation was not always performed as soon as tests showed the collateral circulation to be adequate; some patients were permitted to return home on sick leave between operations. In some instances the second operation was deferred because of an illness unrelated to the vascular lesion, e. g., malaria or gastroenteritis. In most instances, however, operation on the vascular lesion was performed shortly after sympathectomy. In 9 of the 24 patients it was done within 2 weeks, in 16 within 3 weeks or less, and in 18 within 4 weeks or less (the figures are cumulative). In only 6 instances was operation delayed 5 weeks or longer. In 1 of these patients (Case 17) tests showed the collateral circulation to be adequate shortly after sympathectomy, but operation had to be deferred because of a compound fracture of the femur and infected ulcers of the foot. In the other 4 patients tests showed that the collateral circulation improved slowly and steadily after sympathectomy.

  In all but 2 of the patients listed in Table 36, cure of the aneurysm or fistula required ligation of the affected artery. In the 2 exceptions (Cases


18 and 24) continuity of the artery was preserved or reestablished by ligation of the fistula or by bridging the arterial defect with a vein graft.

  In the statement of results the term excellent in this and in other tables should be understood to mean that the limb was of normal warmth and color under ordinary environmental conditions, that there was no significant or disturbing sensitivity to cold, and that there was satisfactory evidence of the return of nerve function. If the patient was not followed long enough for restoration of sensation and motor power to become evident, such evidence as progression distally of the point at which Tinel's sign might be elicited was interpreted to mean satisfactory improvement as compared with that recorded by patients with nerve injury but in whom no vascular disorder existed. Naturally, there was no regeneration of those nerves in which apparently irreparable damage had occurred and on which no reparative procedure could be carried out. The term excellent should not be understood to imply that no fatigue was present on exercise. This finding was invariably noted in patients who had had arteries ligated, especially the femoral or the popliteal.

  One patient (Case 1) had some cold sensitivity following an operation in which it was necessary to ligate the axillary artery, and in another (Case 22) gangrene of the distal third of the foot occurred after operation.

  Case 22. This patient, a 35-year-old soldier, had a popliteal arteriovenous fistula. No flush was present during the reactive hyperemia test. Sympathectomy was performed 4 months after the injury and immediately afterward tests showed definite improvement in the collateral circulation. The flush appeared promptly after deflation of the constricting cuff and steadily improved becoming full and complete in 2 minutes; there was no further improvement upon release of pressure from the popliteal artery. Similar results were obtained on several occasions.

  The fistula was explored 3 weeks after sympathectomy. A large communication was found between the popliteal artery and vein and there was present in addition a small saccular aneurysm of the popliteal artery from which emerged a large geniculate branch. This branch had to be ligated in the course of the excision. It seemed probable that during the preoperative tests the fistula had been occluded without interruption of the blood supply through this branch and that the information concerning the state of the collateral circulation had therefore been misleading. At the conclusion of the operation both the color and the warmth of the foot were poor and within a few days gangrene was apparent. Amputation of the distal third of the foot was necessary.7

  Comment. This case has been described in some detail because it illustrates a number of points: (1) that sympathectomy offers no guarantee against ischemic disaster; (2) that tests for collateral circulation may sometimes be fallible, and (3) that it is essential to take every possible precaution in vascular surgery. In this connection it should be mentioned that one essential precaution was omitted in the case described above. The drapes used at operation were applied in such a way that the foot was not in view during the procedure; a

7 The author of this chapter, in a total military and civilian experience covering 56 femoral and 55 popliteal aneurysms and arteriovenous fistulas treated by surgical means, observed only 1 other case in which gangrene occurred under these circumstances. In this case thrombosis developed on the sixth postoperative day, in a limb in which the circulation had previously been excellent.


precaution which should never be omitted in any instance in which doubt exists as to the adequacy of the collateral circulation. It was unfortunately omitted in this case because of the false sense of security which prevailed: Up to this time in the hospital experience, ischemic difficulty had never occurred in a limb following sympathectomy, and there had never been an episode to suggest that the reactive hyperemia test might give misleading information.

  In 14 patients (Table 37) the collateral circulation was adjudged inadequate following sympathectomy. In all, however, there was some improvement over that shown by preoperative tests. In a few the improvement was quite noticeable but in none did the collateral circulation become adequate according to criteria established. It will be noted that in this group of patients a longer interval elapsed between sympathectomy and operation than in the group in which sympathectomy produced better results. In only 1 of the 14 patients (Case 37) was the interval less than 5 weeks, and in 9 patients, almost two-thirds of the total number, it was 2 months or longer. In 1 patient in this group (Case 27) a satisfactory spontaneous cure occurred as the result of thrombosis. In 9 patients continuity of the affected artery was maintained or restored by ligation of the fistula, by end-to-end suture, or by vein transplant. In only 4 patients was it necessary to ligate the involved artery.

  In all 14 patients (Table 37) an excellent end result was obtained. The limbs were of good color and warmth, there was no sensitivity to cold, and in those patients with peripheral nerve paralysis, return of nerve function progressed satisfactorily. In those in whom the continuity of the affected artery was maintained or restored there was, in addition, no fatigue on exercise in any instance in which it, had not been present before operation.

  In all of the 13 patients upon whom operation was performed (in 1 patient [Case 27] spontaneous cure occurred by thrombosis and operation was unnecessary) color and warmth of the affected foot or hand were observed during operation over a prolonged period of precise occlusion of the involved artery with a rubber-shod clamp. In every patient but 1 the collateral circulation was observed to be adequate, though repeated preoperative tests had led to a different conclusion. The excellent postoperative results in the 4 patients in whom the involved artery was ligated proved that the collateral circulation was efficient. In some of the other patients anatomic conditions were found at operation which explained the reason for the misleading results of the tests carried out before operation. In some instances large collateral vessels were present which would necessarily have been occluded during digital compression, but which could be preserved at operation.

  That the collateral circulation was apparently satisfactory in all but 2 of the 14 patients in this group (Cases 27 and 31) does not necessarily indicate that it was made so by sympathectomy. All that can be said is that the same tests for the collateral circulation made before and after sympathectomy showed some improvement in all instances following sympathectomy and notable improvement in a few instances but in no instance did they show the


collateral circulation to be entirely adequate according to established criteria.

  Four types of response to sympathectomy were evident (Tables 36 and 37). (1) In some instances tests showed the collateral circulation to be adequate immediately after, or shortly after, sympathectomy. (2) In some instances the tests showed significant improvement soon after sympathectomy, and this was followed by slow, steady improvement until the collateral circulation became completely satisfactory some weeks later. (3) In some instances the tests showed some improvement, but never enough improvement so that the collateral circulation could be adjudged adequate before the vascular operation, though more precise tests performed with the lesion exposed at operation demonstrated that the collateral circulation actually was satisfactory. (4) In one instance tests showed the collateral circulation to be inadequate and according to all tests, including those made at operation, it remained so.

  It may be profitable to illustrate these four types of response with brief illustrative case reports:

  Case 5. A 32-year-old soldier who had been injured by shell fragments 15 May 1944 was admitted to Mayo General Hospital 1 September 1944 with a large aneurysm in the region of the right femoral artery. There was no essential difference in the color or temperature of the two feet. The reactive hyperemia test, carried out on several occasions between the date of admittance and the latter part of October, revealed no flush in the foot over a period of 3 minutes compression of the femoral artery although a brilliant flush appeared as soon as compression was released. Lumbar sympathectomy was performed 31 October. When the reactive hyperemia test was carried out on the third postoperative day, the flush was almost instantaneous, reaching the toes in 5 seconds and becoming complete and full in 60 seconds. There was no further improvement upon release of pressure from the femoral artery. Excision of the femoral aneurysm 6 November, with concomitant ligation of the femoral vein, produced an excellent result. There was good warmth and color in the foot at all times.

  Case 18. A 26-year-old officer who had sustained an injury from a shell fragment 16 April 1945 was admitted to Mayo General Hospital 29 May with a midfemoral arteriovenous fistula. There was no flush of the foot during the reactive hyperemia test. Sympathectomy was performed 14 July. When the test was carried out again a week later, a flush appeared in some of the toes in 15 seconds and became complete and of good quality in 2 minutes. A much more brilliant flush, however, appeared when pressure was released from the femoral artery. By 17 August the reactive hyperemia test showed an excellent flush which began at the toes in 10 seconds and was complete and full in 80 seconds. There was no improvement upon release of pressure from the femoral artery. The fistula and the involved area of artery and vein were excised 20 August; a vein graft was utilized to bridge the arterial defect. The graft was successful and the patient had excellent circulation in the limb.

  Case 29. A 20-year-old soldier injured 10 December 1944 was admitted to Mayo General Hospital 7 March 1945 with a fairly large popliteal aneurysm. After admittance and for the next few weeks, the flush during the reactive hyperemia test was poor and incomplete. Sympathectomy was performed 19 April. Thereafter the results of tests showed steady improvement. On 3 May a flush reached the toes in 30 seconds, improved slowly during the 2-minute period of observation, and became strikingly better upon release of pressure from the popliteal artery. By 30 July the flush reached the toes in 10 seconds. Though it improved considerably during the 2-minute period of observation, it improved still more upon release of pressure from the popliteal artery. When the aneurysm was


explored 3 August and the popliteal artery occluded with a rubber-shod clamp just proximal to the lesion, the foot was warm and of good color. The sac was opened and the artery transfixed above and below the defect. After operation the foot was always warm and of good color.

  Case 31. A 27-year-old soldier, injured 26 February 1945, was admitted to the Mayo General Hospital 4 May with signs indicative of a femoral arteriovenous fistula. After admittance and for several weeks thereafter, no flush occurred during the reactive hyperemia test. Sympathectomy was performed 5 June. It was followed by only slight improvement; the flush was of poor quality and incomplete. There was no noticeable change during the following weeks. In the belief that some local factor might be responsible for the poor results of the reactive hyperemia test and that the collateral circulation might actually be adequate, as it had proved to be in other cases, the lesion was cautiously explored 25 August. When a rubber-shod clamp was placed upon the artery above and below the fistula, the foot promptly became extremely pale and cool and did not change its appearance as long as the clamps were left in place. The wound was therefore closed without an attempt to extirpate the fistula. The limb was tested repeatedly during the next few months and finally, 11 January 1946, the vessels were again explored. As at the previous operation, precise occlusion of the artery produced persistent pallor and coldness in the foot. The fistula was carefully dissected out and it became evident that it could be transfixed and reinforced by a segment of the divided vein thereby preserving the continuity of the artery. This was done and the results were excellent.

Collateral Circulation not Determinable

  Sympathectomy was performed in 10 patients (Table 38) because it was impossible to compress the involved artery at the site of the aneurysm or fistula to test the status of the collateral circulation.In 3 patients the artery could




not be compressed near the affected portion because the aneurysm was extremely large and the attempted maneuver was accompanied by intolerable pain. In 4 more, compression of the artery was impossible because the lesion was located in the mediastinum. In the final 3, all with axillary or subclavian lesions, the overlying clavicle and heavy musculature made it impossible to occlude the lesion near the site of the defect. Since, with a single exception (Case 39), there was no way of determining what the results might have been had sympathectomy not been performed, it was difficult to evaluate the effect of the procedure. Certain data, however, suggest that it was beneficial. Thus in the exceptional case just mentioned (Case 39), the patient had a lesion within the mediastinum. During an exploratory operation, complete pallor of the hand occurred during temporary occlusion of the innominate artery proximal to the aneurysm. A partial proximal ligation was done, but since the resulting reduction in pulsation in the extremity was so transient (confirmed by oscillometric readings) it was obvious that the procedure had failed. Sympathectomy was performed, followed by complete proximal and distal ligation. Results were excellent. The aneurysm was cured and there was excellent circulation in the hand.

  Again, in two patients (Cases 41 and 48) some sensitivity to cold was present after operation. Since it is well established that this condition is


either improved or relieved by sympathectomy in instances involving ligation of the artery, there is every reason to believe that it would have been even more marked in these patients had sympathectomy not been performed.

  Two patients (Cases 43 and 48) who had had severe pain before sympathectomy were more comfortable after the operation.

  Four patients (Cases 40, 41, 42, and 48) who had sustained extensive damage to the brachial plexus experienced some return of nerve function following sympathectomy. In light of the state of the nerves at the time of exploration, any return of function must be considered remarkable.

  In one patient (Case 47) signs of a mediastinal subclavian fistula disappeared immediately after sympathectomy as the result of thrombosis. The possible role of sympathetic denervation in bringing about a cure by means of thrombosis will be discussed later.

  Obviously the sympathectomy performed upon one patient was needless (Case 45), for the vascular lesion was afterwards found to be inoperable.


  The only sure way to prevent vasoconstriction from jeopardizing the circulation after operation is to interrupt the sympathetic innervation. Since in operative cure requiring ligation of a main artery, maintenance of adequate circulation in a limb depends upon full utilization of existing collateral channels, persistent vasospasm should not be permitted to compromise circulation through these channels.

  Sympathectomy was performed upon 6 patients (Table 39) because of the primary indication of vasospasm. In all 6 there was evidence of intense persistent vasospasm in the affected limb and additionally, in 3, a longstanding history of a pronounced tendency to vasospasm in all the extremities. The results in these 6 patients, and in a number of patients with vasospasm in other categories, furnish evidence of the benefits to be secured by sympathectomy performed upon the indication of vasospasm. In all instances the hand or foot remained warm and the color good after surgical cure of the aneurysm or fistula. In several patients tests for collateral circulation showed it to be only fair before sympathectomy and good following the procedure.

Miscellaneous Indications

  In 2 patients (Table 40) sympathectomy was carried out chiefly because of causalgia. In 1 patient (Case 56) prompt and complete relief was experienced. In the other (Case 57), the patient reported "about 90 percent" relief. As a result, it was possible to withdraw rapidly the narcotics to which he was almost addicted. This patient also had a number of other complications, including nerve paralysis, a compound comminuted fracture of the femur, and ulcers and superficial gangrene of the toe and heel. All of these lesions healed and there was adequate circulation in the extremity.



  Another patient (Case 58) had gangrene and infection of all the toes and a compound comminuted fracture of the femur. The gangrenous toes were amputated at the time that the sympathectomy was performed. The end results in this patient were excellent. The fistula was cured through surgical intervention, the fracture healed, the infection cleared, and the patient had a useful limb with excellent circulation.

  Three patients (Cases 55, 59, and 60) had peripheral nerve lesions in limbs in which the circulation was obviously impaired or the collateral circulation regarded as only questionably adequate. One of these had in addition an ulcer of the foot. Following sympathectomy, results were good in all three.

  In the last patient in the category of miscellaneous indications (Case 61), three previous attempts by other surgeons had failed to effect a cure of an arteriovenous fistula. Both the ulnar and the radial artery had previously been ligated and divided. It seemed likely that important collateral vessels might have to be sacrificed in a radical excision of the cirsoid aneurysm which



involved the ulnar aspect of the forearm. Sympathectomy was performed in the hope that it would provide protection against ischemic difficulties. Operative cure necessitated excision of the flexor sublimis muscle and of the extensive vascular channels in and about this muscle. Results were good; circulation in the hand remained excellent.

Errors of Localization

  In 7 patients (Table 41) sympathectomy was performed in error because an aneurysm or a fistula was thought, before operation, to involve arteries other than those actually affected. Such tests as could be performed indicated poor collateral circulation in all. In 3 patients (Cases 62, 63, and 64) digital pressure sufficient to still the aneurysm or fistula invariably occluded the over-



lying femoral as well as the involved profunda femoris artery. In all of these the lesion was, therefore, thought to be in the femoral or common femoral artery. In each case, however, it was found at operation to be in the profunda femoris artery.

  In two other patients (Cases 65, 66) the fistula was in such close proximity to the popliteal artery that the bruit and thrill could be eliminated only by digital pressure which compressed the adjacent popliteal artery. In the first patient the lesion was in the geniculate artery and in the second in the posterior tibial artery. In a sixth patient (Case 67) a fistula between the transverse cervical artery and the internal jugular vein could be occluded only by pressure which obliterated the brachial pulse. This fistula was thought therefore to


involve the subclavian vessels. In the remaining patient (Case 68), actually with an arteriovenous fistula of the posterior tibial artery, it was thought that the anterior tibial vessels might also be involved since the fistula could be closed only by compression which occluded the anterior as well as the posterior tibial artery. This patient had an extensive fracture of the fibula and it is likely that it contributed to the difficulty of accurate digital compression.

  In all 7 patients tests showed the collateral circulation to be poor before sympathectomy and good afterward. It is apparent, in retrospect, that had correct localization been possible in each of 3 patients, sympathectomy could safely have been omitted. It is also apparent in retrospect that with 1 possible exception (Case 65), arteriograms might have established the correct site of the lesion in all 7.


  In 9 patients (Table 42) sympathectomy was performed at the time that the operation upon the aneurysm or fistula was done. In 1 patient (Case 77) the sympathectomy was carried out primarily because the sympathetic chain was easily exposed in the operative incision through which the iliac artery had




been isolated as a preliminary precaution. The fistula was so high that it was between the medial circumflex femoral artery and the common femoral vein. In another instance (Case 73) sympathectomy might also have been omitted had the fistula been correctly located before the exploration. This patient had a very large aneurysm in the anterior aspect of the thigh which could be stilled only by pressure which occluded the femoral artery. The profunda had been ligated in this patient shortly after injury. Tests showed the collateral circulation to be extremely poor before sympathectomy and excellent afterward. In this case the aneurysm involved only the lateral femoral circumflex artery.

  Results were good in 7 of the 9 patients in this group. In 1 patient (Case 75), though the foot at all times after operation was warm and of good color, some peroneal sensory loss without motor involvement was added to the preexisting tibial and saphenous anesthesia. At no time, however, was there evidence of circulatory impairment and it seems likely that this complication was the result of pressure from the tourniquet rather than of postoperative ischemia. In the other patient (Case 71), ischemic paralysis developed after operation. The circulation was obviously impaired for a few hours after operation, but thereafter the limb rapidly regained normal warmth and color. In this patient, sympathectomy and the operation upon the vascular


lesion were performed concomitantly without retesting the collateral circulation following sympathectomy. This precaution should never be omitted.

  That it is sometimes advantageous to perform sympathectomy at the same time as the vascular operation is obvious from a consideration of the patients in this group in whom severe pain was present. To illustrate, in 1 patient (Case 70) , an extremely large popliteal aneurysm ruptured subcutaneously and was so painful that the patient writhed in agony. Large doses of morphine afforded no relief. The reactive hyperemia test was characterized by a complete absence of flush. Sympathectomy was performed and a complete and intense flush was present within 30 seconds after release of the constricting cuff. Aneurysmorrhaphy was therefore done at once. The patient was entirely comfortable after operation and there was excellent circulation in the foot.


Exercise Tolerance and Cold Sensitivity

  Two functional disorders which commonly follow the ligation of an artery for the cure of an aneurysm or a fistula are decrease in exercise tolerance, and sensitivity of the limb to cold. Since reduction in exercise tolerance is particularly noticeable after ligation of the popliteal, femoral, and common femoral arteries, a consideration of the effect of sympathectomy upon intermittent claudication or its equivalent is best limited to patients with lesions of these vessels. In addition, certain other patients should be excluded from any study of exercise tolerance: (1) those patients who had difficulty walking prior-to sympathectomy because of fractures, amputated digits or parts of limbs, or motor loss from peripheral nerve injury, (2) those patients who were unable to walk any distance before operation, and this at a time when the continuity of the blood flow through the affected artery was uninterrupted, and (3) those patients whose arterial lesions were successfully repaired and continuity of the blood flow preserved.

  Exercise tolerance in the lower extremity was determined by having the patient walk at a normal pace with a pedometer, or over a measured course, until he was forced to stop because of fatigue or, occasionally, cramps in the limb. The distance walked was approximately the same in the two groups. Patients whose operations had required ligation of the popliteal artery and upon whom sympathectomy had not been done were forced to stop after they had walked an average of 0.68 mile, while those who had had sympathectomy performed could walk 0.73 mile. When ligation of the femoral or the common femoral artery had been necessary, the distance covered was the same (an average of 0.73 mile) regardless of whether sympathectomy had or had not been done.

  These data seem to demonstrate that exercise tolerance was almost precisely the same regardless of whether or not sympathectomy was performed.


  The two groups, unfortunately, were not carefully controlled from several important standpoints. The time interval between operation and final walking test varied from patient to patient. This is an important factor since some patients reached their maximum tolerance rapidly, others more slowly. It proved impossible to analyze the intensity of the exercise pain in some of the patients. Exaggeration was always a possibility in those who desired to avoid duty and be separated from the service. Finally, the two groups were not entirely comparable because from the various tests and observations it was evident that the collateral circulation was more efficient in the group in which sympathectomy was considered unnecessary.

  More clear-cut information is available concerning the effect of sympathectomy in the prevention of sensitivity of the limb to cold after operation. Sensitivity to cold of varying degrees was a complaint of 5 of 18 patients upon whom sympathectomy had been done. In these patients the continuity of the subclavian, axillary, or brachial arteries had been interrupted either through ligation or reparative procedures which failed. In 31 patients with popliteal or femoral lesions treated similarly, only 1 complained of sensitivity to cold. Two of the 6 patients who complained had this sensitivity before operation.

  In contrast, cold sensitivity was a complaint of 13 of 40 patients upon whom sympathectomy had not been done. In these patients the continuity of the subclavian, axillary, or brachial arteries had been interrupted. In 65 patients with popliteal or femoral lesions treated similarly, 5 complained of sensitivity to cold. Eight of the 18 patients who complained had this sensitivity prior to operation.

  Sensitivity to cold was a complaint of 14 patients with aneurysms or arteriovenous fistulas who were received at the Mayo General Hospital after they had been operated on at other installations (see Part II this chapter). There were 5 such complaints in 11 patients with lesions of the main arteries of the upper extremity, and 9 in 30 patients with lesions of the main arteries of the lower extremities. None of the patients who complained of cold sensitivity after operation had been treated by sympathectomy. A single patient in this group had had sympathectomy performed at the time of the vascular operation. He did not complain of sensitivity to cold.

  Altogether, of the 50 patients upon whom sympathectomy had been performed prior to an operation which resulted in interruption of the main arterial stem to the extremity, 6 complained of sensitivity to cold (12 percent). In contrast, of 146 patients who had similar operations without sympathectomy, 32 complained of cold sensitivity (21.9 percent).

Intrasaccular Thrombosis

  Not uncommonly sympathectomy was followed by a noticeable increase in the mural thrombus within the aneurysmal sac. Sometimes this process was very extensive. In one patient, for example (Case 62), a large pulsating aneurysm


in the thigh became progressively smaller and firmer after sympathectomy and eventually lost its pulsation entirely. Clinically it appeared that the lesion had been completely obliterated, but arteriograms revealed the persistence of a sac several centimeters in diameter. At operation the laminated mural thrombus was found to be many times larger than the remaining aneurysmal cavity.

  In 2 other patients cures of this type occurred. The first patient (Case 27) had a pulsating femoral aneurysm about 7 cm. in diameter. Tests of the collateral circulation gave evidence of its complete inadequacy. Sympathectomy was performed 2 months after injury. After this procedure the aneurysm became by degrees somewhat firmer and smaller and pulsated less vigorously. The process was slow during the first few months but finally, 6 months after sympathectomy, only a small, firm, nonexpansile mass remained. Arteriograms revealed almost complete obliteration of the sac. During the next few weeks this mass practically disappeared and it was evident that a satisfactory cure had been obtained.

  In the second patient (Case 47) there were signs suggestive of an arteriovenous fistula of the proximal portion of the left subclavian vessels. After sympathectomy the thrill and continuous bruit disappeared and only a short systolic bruit remained. The extra-mediastinal portions of the vessels were explored in the belief that a saccular aneurysm remained. No aneurysm was found, and there was such dense scarring about the vessels as the mediastinum was approached that exploration was discontinued on the assumption that the remaining systolic bruit was probably the result of partial compression of the artery by scar tissue.

  In evaluating the results of sympathectomy in this group it must be borne in mind that spontaneous cure occurs occasionally without operation in both arterial aneurysm and arteriovenous fistula (see Chapter XII). It took place in 11 of the 224 patients in this particular series upon whom sympathectomy was not done (4.9 percent). On the other hand, a cure by thrombosis occurred in but 2 of the 77 patients upon whom sympathectomy was performed (2.6 percent). It is difficult, therefore, to be certain that thrombosis in these patients was actually the result of sympathectomy. The decision is particularly difficult when complete thrombosis occurs gradually and over a long period of time, as happened in 1 patient (Case 27). When it occurs promptly after sympathectomy as in some of the cases recorded in the literature, the march of events seems to suggest that it has resulted from sympathectomy. The cause-and effect relationship seems all the more likely because of the incomplete thrombosis of the sac which is commonly observed after sympathetic denervation.

  The question arises whether an increase in mural thrombosis or complete thrombosis of the sac can be associated with extension of the clot distally. The course of a patient in the author's experience in civil practice suggests that, uncommon as the occurrence may be, it can take place.



  The patient was a 49-year-old man with an arteriosclerotic aneurysm of the right popliteal artery. When he was admitted to the hospital the aneurysm pulsated vigorously and there was a loud systolic bruit. There was also evidence of peripheral arteriosclerosis. On the affected side the dorsalis pedis pulse was present but the posterior tibial, absent. Both pulses were present in the left foot. The collateral circulation was very poor. Although these findings were checked the day before sympathectomy was performed, the pulses, unfortunately, were not palpated nor were oscillometric studies made immediately before operation. The day after sympathectomy the aneurysm was observed to be somewhat firmer and there was only a faint, shock-like sound in place of the systolic bruit previously heard. The dorsalis pedis pulse had disappeared and oscillometric studies confirmed the impression that the popliteal artery was occluded distal to the aneurysm. This impression, as well as the fact of a recent increase in the extent of mural thrombus, was verified at operation 8 days later. Although no proof exists in this case, the evidence suggests that the extension of the mural thrombus following sympathectomy brought about occlusion of the artery distal to the aneurysm. An excellent result followed aneurysmorrhaphy.


Effects of Sympathectomy

  In considering the role of sympathectomy as an adjuvant measure in the surgical treatment of aneurysms and arteriovenous fistulas, it is important to keep in mind that in the hands of those familiar with it, this procedure is associated with minimal risk and very little discomfort. Patients can be ambulatory the day after either dorsal or lumbar sympathectomy. In the experience at the Mayo General Hospital there were no deaths and complications were extremely uncommon.

  Data derived from the tests given the patients in this series supply convincing evidence that sympathectomy is useful in rendering the collateral circulation more efficient by eliminating vasoconstriction in the collateral vessels. This is exemplified by the fact that following sympathectomy the hand or foot was almost always of excellent warmth and color in patients in whom cure of the vascular lesion entailed arterial ligation. It is even more strikingly exemplified by the remarkable improvement in the collateral circulation which so often occurred immediately after sympathectomy was carried out. These data do not offer convincing proof that sympathectomy actually increases the collateral circulation by fostering the growth of new collateral channels. However, data derived from tests made on these patients in whom the collateral circulation showed only slight improvement immediately after sympathectomy, but steady significant improvement during the ensuing weeks and months might suggest that sympathectomy had this effect. Perhaps a more plausible explanation, however, would be simply progressive dilatation of existing collaterals. In contrast it is noteworthy that in a number of patients upon whom sympathectomy was not performed, tests showed that the circulation had improved during short periods of observation.


  The immediate effect of sympathectomy upon the status of the collateral circulation can be predicted with reasonable accuracy by comparing the results of the reactive hyperemia test in any given patient under ordinary environmental conditions with those obtained during reflex vasodilatation with the patient under anesthesia or, better still, after sympathetic block with procaine. Such studies are unnecessary as routine measures, however, and one can proceed with sympathectomy without additional testing in any case in which the indications are plain. Should the preoperative tests for collateral circulation show it to be poor and should no distinct improvement follow sympathectomy, the situation should be considered sufficiently precarious to warrant the use of all additional measures which might add some safeguard against ischemic difficulties. These include intermittent occlusion of the artery, delay in operative treatment, and efforts to preserve the continuity of the artery at the time of operation.

The Rationale of Sympathectomy

  The Mayo General Hospital experience indicates that a rational plan for the use of sympathectomy in aneurysm or arteriovenous fistula can best be formulated on the basis of the various indications for its use.

  Inadequate Collateral Circulation. Sympathectomy seems indicated when there is evidence of poor collateral circulation provided the lesions are of sufficient duration to have produced under ordinary circumstances good collateral circulation, and provided the simpler means of improving the collateral circulation have shown no results during a short period of trial. The literature is full of statements concerning results of intermittent proximal occlusion of the affected artery on the collateral circulation and in some instances there is clear evidence that such an effect has been achieved. In many patients, however, it is difficult to see any beneficial results. It would be a significant contribution to the problem of the treatment of aneurysms and arteriovenous fistulas if controlled experimental and, preferably, clinical studies should establish beyond question the value and limitations as well as the criteria for continuing or abandoning such procedures. From the experience at this center it would seem that sympathectomy is indicated if there is no definite improvement in the collateral circulation following such efforts as intermittent proximal occlusion of the affected artery carried out over a period of some weeks. It has been established that permanent partial occlusion of the proximal artery is sometimes helpful in increasing the collateral circulation in instances of arterial aneurysm, but this procedure is not applicable in instances of arteriovenous fistula.

  Since sympathectomy is as simple as partial proximal ligation and probably safer, and since partial proximal ligation may jeopardize the chances of successfully maintaining or reestablishing the continuity of the involved artery when the major operation is attempted, sympathectomy seems to be the wiser procedure in arterial aneurysms in which the collateral circulation does


not appear to be adequate. In the majority of instances in this series in which sympathectomy was carried out because the results of tests showed the collateral circulation to be unsatisfactory, improvement was noted fairly promptly. Even in the instances in which the tests did not show the circulation to be entirely adequate following sympathectomy, the procedure at least assured the surgeon that vasoconstriction had been eliminated.

  Associated Peripheral Nerve Lesions. When the aneurysm or arteriovenous fistula is associated with a peripheral nerve lesion which requires operative treatment, a precarious collateral circulation seems a clear-cut indication for sympathectomy. It is imperative that the nerve lesion be treated as promptly as possible and it is entirely too hazardous to attempt nerve repair before operative cure of the aneurysm or fistula has been achieved. Any effort to improve the collateral circulation and permit early operative treatment both of the arterial and the nerve lesion is therefore worthwhile.

  Sympathectomy would seem to be indicated particularly when the combined vascular-nerve lesion is associated with obvious evidence of impaired circulation in the extremity regardless of whether or not the collateral circulation appears to be adequate. In such cases it must be assumed that operative cure of the vascular lesion may require ligation of the artery and further reduction in the circulation of the limb. Furthermore, an adequate vascular supply fosters nerve regeneration exactly as impairment of the circulation may produce ischemic nerve injury. The data in this series are insufficient for a quantitative comparison of nerve regeneration in patients with unimpaired circulation and those with impaired circulation upon whom sympathectomy was done. The impression seems warranted, however, that nerve regeneration was comparable in the two groups. The data do offer conclusive proof that sympathectomy may be of great benefit in instances of ischemic nerve injury with impaired circulation resulting from arterial ligation, division, or thrombosis.

  Collateral Circulation Undeterminable. When the anatomic location of the aneurysm or fistula is such that digital occlusion of the artery is not possible, sympathectomy seems advisable if the artery affected is one in which ligation is sometimes followed by ischemic difficulties. When the artery cannot be compressed precisely at the site of the defect, proximal or distal compression may reveal some information concerning the state of the collateral circulation. It must be borne in mind, however, that tests of this kind may supply erroneous information either because collateral channels which can be preserved during the operative procedure are occluded, or because the reverse is true and channels which may have to be sacrificed have not been occluded.

  Ischemic Lesions. When ischemic lesions are present distal to an aneurysm or a fistula, sympathectomy is indicated since cure of the vascular lesion may entail ligation of the artery thereby still further reducing the circulation. Even if the blood flow through the artery can be preserved, increase in the circulation would be highly desirable since the ischemic lesion supplies unequivocal evidence of insufficiency.


  Vasospasm. If the main artery to a limb must be ligated for the cure of a vascular lesion, the surgeon cannot afford to take the chance of having the remaining circulation jeopardized by vasospasm in the collateral vessels. Vasospasm, of course, is not present in every limb in which a main arterial stem is ligated. Indeed, the vascular tone after ligation may be high, low, or normal. In some patients the limb actually exhibits evidence of vasodilatation and increased stability of skin temperature under varying environmental circumstances; these findings suggest the effect of periarterial sympathetic interruption consequent to ligation and division of the artery. Whenever there is evidence of pronounced vasospasm it appears unwise to run the risk of its occurrence after operation. Such a problem might be handled in two ways: (1) Proceed with the operation and then if alarming vasospasm ensues attempt to control it by sympathetic blocks or, if necessary, by sympathectomy. (2) Perform sympathectomy either before the vascular operation or at the same time, ever though tests for collateral circulation show it to be good. This is the safer method and one to be recommended in instances in which intense vasospasm is already present.

  Miscellaneous Indications. Sympathectomy may be required in the occasional patient in whom one or more of the main arteries to a limb have already been occluded by previous injury or operation and in whom cure of an aneurysm or fistula may require ligation of other arteries essential to nutrition and proper function. Similarly, an occasional patient will be encountered in whom an aneurysm or fistula is associated with severe causalgia which can be temporarily relieved, though not cured, by sympathetic blocks.

  Unnecessary Sympathectomy. Sympathectomy should not be performed unless. there is some specific indication for its use, and even if the indications listed are strictly adhered to as a basis for the employment of the operation it will undoubtedly be performed in certain instances in which it might safely have been omitted. These instances occur chiefly when, during operation, it is found possible to maintain the continuity of blood flow through the affected artery by some reparative procedure.Since, however, the feasibility of such procedures cannot be foreseen, and since one cannot always be sure that the repair will be successful, it is wise to make certain that the collateral circulation is adequate before an attempt is made to extirpate an aneurysm or fistula. In practice, adherence to this policy will mean the performance of an occasional unnecessary sympathectomy.

  Sympathectomy is sometimes done needlessly because of improper localization of the aneurysm or fistula. Fortunately, this seldom happens. Difficulties are most often encountered in lesions located high in the profunda femoris artery in which digital pressure sufficient to still an aneurysm or abolish the bruit and thrill of a fistula is likely to compress the overlying femoral artery. In such cases arteriograms will ordinarily establish the correct location of the lesion.


  Even though preliminary sympathectomy is not performed, numerous patients will recover from operative cure of an aneurysm or fistula with no signs of gangrene or ischemic paralysis. When, however, such serious complications appear to offer a definite threat, it seems wise to take every precaution to avoid them, even at the risk of performing an unnecessary operation. One cannot afford to gamble with the viability of an arm or a leg. Indeed, as the reported series of cases illustrates, such disasters may occur even if sympathectomy is performed. It should be emphasized that sympathectomy offers no guarantee that some ischemic difficulty may not occur after ligation of an important artery: Sympathectomy must therefore be looked upon as an aid in, but not as a sure preventive of, ischemic troubles. Furthermore, the utmost care in testing the collateral circulation and in avoiding injury to collateral vessels at the time of operation must be exercised or sympathectomy will be associated at times with disastrous results.

  Exercise Fatigue and Cold Sensitivity. Fatigue on exercise and sensitivity to cold are two symptoms which commonly occur in limbs following the operative cure of aneurysm or arteriovenous fistula even though circulation to the limb is otherwise adequate. As already pointed out, although no definite conclusions can be formulated in regard to exercise fatigue, the results in the series of patients observed at the Mayo General Hospital give the impression that sympathectomy seldom has any marked beneficial effect upon exercise tolerance. The results of sympathectomy are more clear-cut in respect to cold sensitivity. Though this complication sometimes occurs following arterial ligation in patients upon whom sympathectomy has been performed it occurs far more frequently in those upon whom sympathectomy has not been carried out.

  Intrasaccular Thrombosis. As Gage 8 noted in his original publication, intrasaccular thrombosis frequently occurs following sympathetic denervation in instances of arteriovenous fistulas. DeBakey 9 recorded a case of carotid-cavernous sinus fistula in which cure by thrombosis occurred after sympathectomy, and Colsen and Giddy10 noted progressive thrombosis of the aneurysmal sac within a few days after they had performed lumbar sympathectomy in preparation for exploration of a popliteal aneurysm. In the latter case an apparently satisfactory cure was obtained without operative treatment. In this series of patients extension of the mural thrombus within the sac was often noted following sympathectomy and in two instances apparent cure by thrombosis followed interruption of the sympathetic pathways. Since spontaneous cure by thrombosis occurs occasionally without sympathectomy it is impossible to state that a cause-and-effect relationship exists when a cure by thrombosis

8 See footnote 1, p. 318.

9 Martin, J. D., Jr., and Mabon, R. F.: Pulsating exophthalmos; review of all reported cases. J.A.M.A. 121: 330-335, 30 Jan 43.(Discussion by Ochsner.)

10 Colsen, K., and Giddy, P.: Popliteal aneurysm cured by lumbar ganglionectomy. South African M. J. 18:242, 22 Jul 44.


follows the operation, although there is evidence which suggests that this is true.

  Sympathectomy Combined with Operation on the Vascular Lesion. Sympathectomy and the vascular operation need not be performed as separate procedures. They can undoubtedly be combined more often than was done in the present series of cases. Unless there is evidence of an unquestionably adequate collateral circulation before operation, the procedure at operation should be (1) to perform sympathectomy, (2) to test the collateral circulation, and (3) to proceed with, or abandon, the contemplated operative attack upon the vascular lesion according to the results of these tests.

Part II. Postoperative Sympathectomy

  The usefulness of sympathectomy in increasing the efficiency of the collateral circulation, in providing maximal or near-maximal circulation in an injured limb, in favorably influencing existing ischemic difficulties, and in alleviating such associated conditions as causalgia, made it reasonable to suppose that it might also be of benefit in correcting certain circulatory conditions which sometimes follow the surgical treatment of aneurysms and fistulas. Similarly, experience with sympathectomy in other disorders in which blood flow through an important artery had been interrupted by disease, injury, or operation suggested that this procedure would be of value in correcting these conditions.

  At the vascular center of Mayo General Hospital certain generalizations concerning circulatory disorders following surgical treatment of aneurysms and fistulas were confirmed: Circulatory disorders can be lessened if tests for the efficiency of the collateral circulation are used as a guide in the selection of the proper time for operation and if operative techniques are employed which in no way injure the collateral blood supply. In addition, these disorders can practically be eliminated if the continuity of the blood flow through the involved artery can be maintained or restored by some reparative procedure. Nonetheless, circulatory disorders occur occasionally in spite of all efforts to avoid them, and this chapter is a report of a study made at this center on the value of postoperative sympathectomy in correcting such disorders.


  In addition to the 75 sympathectomies performed upon 288 patients either before or during the course of operation for aneurysms and arteriovenous fistulas, and the 2 performed upon 13 patients in whom cure by thrombosis occurred and surgical intervention proved unnecessary, 19 sympathectomies were performed after operation in the 213 patients in whom preoperative sympathectomy had not been performed.(Table 43.)

11 See footnote 5, p. 320.



  Sixty-two patients were admitted to the Mayo General Hospital after operation elsewhere for aneurysms and arteriovenous fistulas (Table 43), 43 of which had been performed overseas. Upon 19 of the 62 patients sympathectomy had been performed, 1 before and 18 after vascular surgery. In addition there was one other admittance of a patient, in whom spontaneous cure of the vascular lesion had occurred (Case 82, Table 44), and upon whom sympathectomy was performed after neurosurgery.

  The clinical material thus amounted to 116 patients; in 76 sympathectomy was carried out before or in the course of vascular surgery, in 37 after vascular surgery, and in 3 the vascular operation was not necessary because spontaneous cure by thrombosis occurred (Table 43).

  Although the proportion of sympathectomies was essentially the same in the patients operated on at Mayo General Hospital (32.6 percent) and in those operated on elsewhere (31.7 percent), a breakdown of the figures discloses significant differences (Table 43). Sympathectomy was carried out before or at the time of the vascular operation in 26.0 percent of the patients of


Mayo General Hospital, but in only 1.6 percent of the patients operated on elsewhere. The incidence of postoperative sympathectomy at Mayo General Hospital in the 213 patients upon whom sympathectomy had not earlier been performed was 8.9 percent, against 30.6 percent in the patients operated on elsewhere. In both groups the indication for sympathectomy was proportionately more often aneurysm than arteriovenous fistula (Table 43).

  The 38 postoperative sympathectomies were performed upon 19 patients with aneurysms and 19 with arteriovenous fistulas. The upper extremities were involved in 17 patients and the lower in 21. In 3 patients (Cases 105, 106, and 107, Table 47) sympathetic interruption was achieved by means of alcohol injection. In the remainder it was brought about by surgical excision of the sympathetic pathways. Ganglionectomy was used when the lower extremities were involved, the approach being through an anterior extraperitoneal incision. The Smithwick type of preganglionic operation was done when the upper extremities were affected.

  There were no deaths in the 38 patients and no serious complications resulted.


  The chief indications for postoperative sympathectomy in these 38 patients were (1) distressing sensitivity of the limb to cold, (2) associated severe peripheral nerve damage in limbs in which the circulation was definitely impaired, (3) persistent edema, (4) ischemic nerve paralysis, (5) causalgia relieved temporarily but not cured by sympathetic blocks, (6) obviously impending gangrene, and (7) evidence of sympathetic overactivity, with or without poor collateral circulation, in a limb in which another arteriovenous fistula required excision. The data concerning the 38 patients are summarized in Tables 44 through 47, grouped according to the primary indication for sympathectomy. In many instances, however, there were several circumstances which influenced the decision to perform sympathectomy.

  Cold Sensitivity. The largest number of postoperative sympathectomies, 17, was performed because of sensitivity of the affected limb to cold (Table 44). Nine patients in this group had arterial aneurysms and 8, arteriovenous fistulas. In 9 the upper extremity was involved and in 8 the lower extremity. In 1 patient (Case 82, Table 44) cure of the aneurysm by spontaneous thrombosis had occurred; in the others excision of the lesion had been necessary. The duration of the lesion at the time of the vascular operation ranged from 1 to 4 months and averaged about 1 months in the group operated upon overseas; the range was from 2.5 to 11 months and averaged about 5 months in those operated upon at Mayo.

  All of the patients in this group had annoying coldness of the affected hand or foot upon exposure to cold and most of them had cyanosis, mild in some instances and severe in others. Some patients described the sensation as


aching, others as burning or tingling.Most of them had some degree of numbness of the fingers or toes and nearly all of them complained of stiffness of the digits.When paresis was present it was aggravated during exposure to cold, and pain, parestbesia, or hypesthesia also tended to become more intense. The affected hand or foot was generally warm and of good color in a warm environment. One patient (Case 91, Table 44), for example, always had a warm, well-colored foot under ordinary circumstances and showed no evidence of circulatory insufficiency except for the usual fatigue on exercise which was




a problem with all such patients. He lived in Minnesota, however, and on exposure to cold his foot became icy cold and numb, and ached severely. Since there was no improvement in this condition over a period of months, he decided to have sympathectomy performed. Complete relief followed sympathetic ganglionectomy.

  It was the practice to test each patient's ability to withstand exposure to cold. If only mild manifestations of sensitivity were present, sympathectomy was withheld. It was occasionally withheld even when more pronounced, but this was because the patient planned to live in a warm climate and it was felt that the condition would cause little discomfort or disability. However, if the patient planned to live in a cold climate, symptoms of the same degree, were considered an indication for sympathectomy. The situation was always carefully explained to the patient and he was allowed to compare the reaction of the limb to a cold environment under ordinary circumstances and following sympathetic block produced by procaine. After this experience, he was offered the opportunity to elect operation. An occasional patient felt that his work, avocations, and general interests would make prolonged exposure to outside cold unnecessary. The great majority, however, chose to have sympathectomy performed. The advice of fellow patients who had obtained relief from similar complaints undoubtedly influenced the choice of many.


  Eight of the 17 patients in this group had, in addition to cold sensitivity, associated peripheral nerve lesions. A few showed evidence of sympathetic overactivity under ordinary environmental conditions, and in 1 instance the vasospasm was rather pronounced. In several patients there was obvious evidence of impaired circulation. One patient had mild causalgia. One had fairly severe frostbite of both feet which resulted from exposure at the time of injury. The injury produced an aneurysm of one external iliac which was treated by excision of the lesion and division of the affected artery. The frostbite continued to cause cold sensitivity in this patient, but the discomfort was much more severe in the limb in which the artery had been divided.

  Results of sympathectomy in this group of patients with associated peripheral nerve lesions were excellent. All except 2 obtained full relief, and these 2 experienced definite if not complete relief. The patient who suffered from causalgia gained considerable relief of pain; pain was also relieved in all other instances in which it had been present except in the patient with frostbite who continued to have mild burning in the foot. Nerve regeneration in these patients was as satisfactory as could be expected in the light of the trauma which had been sustained.

  Peripheral Nerve Damage. Sympathectomy was performed after operation in 5 patients (Table 45) who, in addition to severe nerve damage, showed evidence of reduced blood flow. One patient had markedly impaired circulation, 1 had extreme hyperesthesia of the hand, and 2 had some sensitivity to cold in the affected part.

  In almost every instance sympathectomy had been considered before operation but had been deferred because of evidence of excellent collateral circulation and in the hope that the damage to the nerve might prove to be caused in large part by pressure of the aneurysm or otherwise to be less severe than had been anticipated. In all five patients the neurologic condition observed at operation was worse than had been expected.

  The results of sympathectomy in all of these patients were extremely gratifying. Signs of nerve regeneration indicated progress as satisfactory as might have been hoped for, considering the extent of the initial damage. The circulation in each of the limbs was visibly improved and sensitivity to cold relieved or reduced.

  Persistent Edema. In 5 patients (Table 46) the indication for postoperative sympathectomy was persistent edema. The edema in these patients ranged from moderate to massive. All had failed to improve under such conservative measures as rest, elevation of the limb, elastic support, and gradually increasing periods of exercise. Sympathetic blocks had resulted in only transient improvement or had produced no effect at all other than temporary warmth and dryness of the foot. All the patients complained of coolness of the affected foot or hyperhidrosis, all had cyanosis, 2 had nerve paralysis, and 1 had indolent ulcers. Two had been operated on overseas when the vascular lesions were, respectively, of 2 weeks and 2 months duration. In the 3 upon whom



the operation was carried out at Mayo General Hospital the lesions were of 2, 3, and 6 months duration.

  Results were excellent in 3 patients, moderately good in 1, and fair in another. In addition to the good effect which sympathectomy had upon the edema, warmth and normal color prevailed in all the limbs treated, and satisfactory return of nerve function took place.

  Miscellaneous Indications. Sympathectomy was performed upon 11 patients for miscellaneous indications (Table 47). Six of these patients had arterial aneurysms and 5, arteriovenous fistulas.

  Ischemic paralysis was the primary indication for sympathectomy in 3 patients (Cases 105, 111, and 112). In 2 of these (Cases 111 and 112) neurologic difficulty followed operations for the cure of femoral arteriovenous fistulas.


The clinical records accompanying the patients contained no data concerning the state of the collateral circulation before operation. One patient (Case 111) had an extensor paralysis of two toes, anesthesia of the foot, and stocking hypesthesia almost up to the knee. Sympathectomy was performed and improvement observed almost immediately. He continued to improve during the ensuing weeks. When last observed motor power had returned to his toes, hypesthesia had disappeared from his leg, and his foot, although still hypesthetic, was no longer anesthetic. The other patient (Case 112) had a stocking anesthesia with complete sciatic paralysis after operation. Return of function in the tibial nerve had been prompt after repair of the vascular lesion, but when he was admitted to the Mayo General Hospital he had complete peroneal motor loss and almost complete sensory loss. The neurologic condition seemed stationary. Sympathectomy was performed and within a few days return of peroneal function was noted. He improved steadily thereafter and complete recovery ensued.






  The third patient with ischemic paralysis (Case 105) had paralysis of the upper extremity after injury. When examined at the Mayo General Hospital, 3 months after excision of a brachial aneurysm and 4.5 months after injury, the gauntlet type of anesthesia and other findings made it evident that his residual paralysis was largely ischemic in character. The hand was cold and cyanotic and he complained of aching upon exposure to cold. Improvement was noted shortly after sympathectomy. The anesthesia receded steadily and 9 months later sensation in the extremity was normal except for hypesthesia of the dorsum of the hand and anesthesia of the palmar surface of the fingers. The hand was warmer and of good color and withstood exposure to cold fairly well. Return of motor function, although considerable, was still incomplete.

  Two patients in the miscellaneous group (Cases 113 and 114) had multiple arteriovenous fistulas. In one (Case 113) there was some sensitivity to cold and evidence of extremely poor collateral circulation following excision of an external iliac fistula. The operation had been performed overseas. When the patient was admitted to Mayo General Hospital a second fistula remained to be excised. At operation this was found to involve the neighboring hypogastric vessels. Sympathectomy was performed and immediately afterward tests of the collateral circulation showed it to be adequate. This fistula was therefore excised. Thereafter, circulation in this limb was excellent. The second patient (Case 114) had a congenital cirsoid aneurysm of the foot with


distinct individual fistulas of the anterior and posterior tibial and peroneal vessels. Excision of the posterior tibial fistula was followed by extreme hyperhidrosis and cyanosis of the foot which persisted for several months. Sympathectomy was performed and thereafter the foot remained warm and dry. The remaining fistulas were resected without difficulty.

  Sympathetic interruption was carried out in 2 patients (Cases 106 and 115) because of obviously impending gangrene. One of these patients (Case 106) had been operated upon overseas nearly 3 months after an injury which had produced partial paralysis of the brachial plexus. An aneurysm of the axillary artery was encountered unexpectedly at operation and excised. After operation the hand became cold and cyanotic, and the thumb, in particular, seemed devoid of circulation. Alcohol injection of the dorsal sympathetics was accomplished successfully, but gangrene of the thumb occurred and required amputation. Fairly good circulation was restored in the remainder of the hand. In this patient definite ischemic paralysis had been superimposed on the preexisting damage to the brachial plexus. Steady improvement in nerve function occurred, but recovery was incomplete. The second patient (Case 115), who at first had had good warmth and color in the foot after excision of a femoral arteriovenous fistula, on the sixth postoperative day suddenly complained of pain and swelling in the foot and leg with coldness, numbness, and paresis of the foot. It was apparent that extensive arterial and venous thrombosis had occurred. Although spinal analgesia produced little improvement in the circulation, sympathectomy was performed because the situation was desperate. The cold level became lower and circulation improved but the improvement was insufficient to prevent gangrene of the toes and sole of the foot, and amputation was required.

  In another patient (Case 107) sympathectomy was performed the day after aneurysmorrhaphy because the foot became alarmingly cold and pale. In this patient the foot regained good warmth and color promptly.

  Sympathectomy was carried out in 3 patients (Cases 108, 109, and 110) because of causalgia. All 3 had suffered peripheral nerve damage and 1 had severe sensitivity to cold. In each, temporary relief followed a series of sympathetic blocks, but lasting effect was not achieved. After sympathectomy causalgia was promptly and completely relieved in 1 patient and greatly relieved in another. The third patient continued to complain of pain after operation. He had, in addition, other complaints which suggested hypochondriasis. Under reassurance and exercise the pain of which he complained finally disappeared.


  If patients who have circulatory difficulties following the operative cure of peripheral aneurysms or arteriovenous fistulas are to be provided with the best possible limbs, and if the desire is not merely to avoid such gross ischemic disasters as gangrene, repeated investigation of the circulatory status


of the affected limbs is necessary. Just the casual observation of patients on a ward will not reveal the true circulatory status of their injured extremities. It is important to test these affected limbs in order to ascertain whether exercise can be tolerated without crippling pain, to make certain that no distressing symptoms ensue when they are exposed to cold, to see that persistent edema or ischemic paralysis is not present, and to make sure that return of nerve function in instances of associated peripheral nerve injury is not being compromised by inadequate circulation. The experience at this center during World War II shows that in certain instances the performance of sympathectomy will correct these circulatory difficulties.

Special Effects of Sympathectomy

  Exercise Tolerance. As was noted in the discussion of preoperative sympathectomy, fatigue on exercise occurs in a striking fashion in the lower extremity after ligation of the iliac, femoral, or popliteal artery and in a less notable degree in the upper extremity after ligation of the subclavian, axillary, or brachial artery. Exercise tolerance of the lower extremities was determined in the patients upon whom sympathectomy had been performed postoperatively, by measuring the number of yards they could walk at a normal pace before they were compelled to stop because of fatigue, or, less commonly, because of cramps in the calf. Exercise tolerance of the upper extremity was measured by having the patient squeeze once every second a rubber bulb which was connected to a 5-gallon bottle with a mercury manometer. The number of squeezing movements which the patients could perform before fatigue prevented further effort was recorded, as was the height of the manometer at the end of the test.

  Statistically significant comparisons were not possible because these studies were carried out both before and after sympathectomy on only a small number of patients. Many patients were not tested because of associated fractures or motor paralysis. Results of the test showed, however, that patients upon whom sympathectomy had been performed after ligation of the femoral or popliteal artery experienced fatigue or cramps after walking an average of 0.82 mile; approximately the same distance walked by patients with similar lesions treated by ligation but without sympathectomy (see Part I).Many patients showed no or only slight improvement in exercise tolerance after sympathectomy. On the other hand, a striking increase in exercise tolerance was occasionally observed. One patient (Case 84, Table 44), who was admitted 4.5 months after excision of a femoral aneurysm, could walk no more than half a mile because of aching and fatigue in the calf. Within 2 weeks after sympathectomy had been performed on the indication of sensitivity of the foot to cold, he could walk 2 miles before experiencing fatigue in the calf. Improvement of such a degree was, however, exceptional.

  Cold Sensitivity. Next to diminished ability to exercise the limb without fatigue, the commonest postoperative circulatory difficulty is cold sensitivity.


If, as has already been pointed out, it is of mild degree, or if it occurs in patients who live in warm climates, it is not likely to cause real discomfort or disability. When it exists in rather severe degree or is present in patients whose place of residence and type of work require exposure to low temperatures, the symptoms are distressing and often disabling.

  As the patients in this series show, sympathectomy which was performed on this indication yielded excellent results. Sympathetic blocks with procaine followed by exposure to cold served as excellent guides as to what might be expected of sympathectomy.

  Persistent Edema. Persistent edema of significant degree is unusual after operation for aneurysms or arteriovenous fistulas. If it occurs it should be treated by rest, elevation of the limb, elastic support, and graduated activity with the limb dependent for a sufficiently long period of time. If these measures prove ineffectual after a fair trial, sympathetic blocks or sympathectomy may be of real value.

  It was noted that if they caused transient though not permanent diminution in swelling, sympathectomy was likely to be very helpful, but even if no effect upon the edema was noted after the blocks, sympathectomy was occasionally effective. Certainly sympathectomy was of real value in every instance when it was performed on this indication in this series---even if the prospects indicated by sympathetic block had been poor.

  Ischemic Difficulties. When ischemic difficulties are present, the efficacy of sympathectomy depends upon the capacity of the collateral circulation to improve through the elimination of persistent or intermittent vasoconstriction and the maintenance of a state of vasodilatation. As with other disorders in which the arterial blood flow has been interrupted, this capacity varies in different individuals.

  In the Mayo General Hospital experience, sympathectomy was strikingly successful in ischemic nerve paralysis. The results in the two patients in whom gangrene was imminent were less striking. Nonetheless, the procedure should be given consideration in such patients even if the limb shows only slight improvement with sympathetic block or spinal analgesia. In occasional instances of impending or spreading gangrene caused by occlusion of arteries from other causes, sympathectomy has produced striking benefits even after a poor response to preliminary tests.

  The experience at this center also suggested that sympathectomy was indicated in patients in whom severe nerve damage was associated with obviously impaired circulation. The evidence of good return of nerve function in these patients would seem to justify the procedure. Unfortunately, there were available no carefully controlled clinical studies comparing nerve regeneration in patients in whom sympathetic interruption had been carried out with those in whom it had not been carried out. Furthermore, sympathetic procaine block did not invariably furnish reliable information concerning the results to be expected following sympathectomy.


  It was noted at this center that if the nerve paralysis had not produced sensory loss in the entire hand or foot, improvement in the circulation produced by sympathetic procaine block could be readily demonstrated by skin temperature changes in the normally innervated digits. If, on the other hand, normal sensation and intact sympathetic innervation had been lost in all parts of the hand or foot, ordinarily no rise in skin temperature could be demonstrated during the block. That sympathectomy may actually increase the circulation in these patients was suggested by a number of phenomena, namely, the decreased tendency to dependent cyanosis which was sometimes noted after operation, diminution in sensitivity to cold, an occasional increase in circulation as demonstrated by the oscillometer and, in particular, the striking improvement in nerve function which so often followed if part of the neurologic damage was ischemic in character. Sympathectomy naturally did not alter the tonus of blood vessels if there was anesthesia of portions of the skin and consequently local sympathetic denervation. It was equally apparent, however, that blood flow in the limb was increased by the elimination of vasoconstriction from the vessels of the limb as a whole.

Factors Influencing the Development of Postoperative Circulatory Difficulties

  A comparison of the patients operated on overseas and those treated at the Mayo General Hospital reveals certain profitable data concerning factors influencing the occurrence of postoperative difficulties for which sympathectomy is required. If only those patients are considered in whom the main arteries to the extremities were involved (innominate, subclavian, axillary, brachial, external iliac, common femoral, and popliteal arteries), 7 of 42 patients in the overseas group were treated by sympathectomy (40.5 percent) against 17 of 182 patients in the Mayo General Hospital group (9.3 percent). If those patients upon whom sympathectomy was carried out before or at the time of operation are eliminated, likewise those in whom the continuity of blood flow through the affected artery was successfully maintained, the figures for the overseas group become 17 of 41 cases (41.5 percent) and for the Mayo General Hospital group 17 of 108 cases (15.7 percent).Sympathectomy was performed upon 1 patient, in each group for the additional reason that a second arteriovenous fistula required excision.

  In general, the time interval between injury and operation for the vascular lesion was shorter in the overseas group than in the Mayo General Hospital group. In the former it ranged from a few days to 4.5 months and averaged about 1 months. In only about 17 percent of the patients was the interval 3 months or longer. In the Mayo General Hospital group, exclusive of 1 patient with a lesion of 13 years duration, the interval ranged from 6 weeks to 15 months and averaged about 5 months. In about 85 percent of those in this group the interval was 3 months or longer.

  Unfortunately, the overseas group and the Mayo General Hospital group are not entirely comparable. Early operation was apparently forced by com-


plicating circumstances in a somewhat higher percentage of the patients operated on overseas than in those treated at the Mayo General Hospital. There were also other differences. Often patients in the overseas group were operated upon by a number of surgeons whose experience in vascular surgery was limited; whereas those in the Mayo General Hospital group were treated by a few surgeons who were specialists in the field of vascular disorders. Careful testing of the collateral circulation before operation was routine at the Mayo General Hospital. If the clinical case records are taken at their face value and if the replies of patients to questioning can be accepted, these tests were carried out only rarely in patients operated upon overseas. The two groups are therefore not comparable because of these extremely important factors. Nevertheless, it may be concluded from the data available that postoperative circulatory disorders follow more frequently operations performed upon aneurysms and fistulas of relatively short duration. This hypothesis can probably be expressed more accurately by stating that the incidence of postoperative difficulties is less when an adequate collateral circulation is established before surgery is undertaken. It must be emphasized that although the collateral circulation tends to be better when the lesion has been in existence a long time, especially in patients with arteriovenous fistulas, there are frequent exceptions to the rule, and careful testing is mandatory.