U.S. flag

An official website of the United States government

Skip to main content
Return to topReturn to top



Sympathectomy in Circulatory Disturbances of the Extremities

Norman E. Freeman, M. D.

  Sympathectomy was not a part of the therapeutic armamentarium in World War I because it had been used in man only occasionally and tentatively up to that time.1 Since then large series of cases have been reported from civilian practice giving the indications for the operation, the technical methods, and the results to be expected. Significant data concerning the operation were gathered from the military experience in World War II particularly because of the concentration of patients with vascular injuries and disturbances into centers for specialized treatment. The unequaled wealth of clinical material thus provided served to confirm and reemphasize certain concepts of the value of sympathetic denervation in the treatment of circulatory disturbances of the extremities.


  There are two chief reasons for the employment of sympathetic denervation of the extremities, namely, the prevention of vasoconstriction and the relief of pain. The importance of sympathectomy in the prevention of vasoconstriction is apparent from Ochsner's statement 2 that the prognosis in any patient with peripheral arterial disease depends upon the degree of vasospasm because the therapy largely concerns the relief of vasospasm. As to the relief of pain, although the exact mechanism by which pain in the extremities is relieved by interruption of sympathetic pathways is still not clear, the fact that many patients are thus greatly relieved has been accepted by most surgeons working in the field.

  A third, though much less frequent, indication for sympathectomy is the control of hyperhidrosis. Excessive sweating is due to abnormal activity on the part of the sympathetic nervous system and occasionally may become so incapacitating that surgical intervention is required. This is particularly true in instances of deficient circulation associated with vasospasm in the extremities. Hyperhidrosis may at times constitute a serious disability, chiefly because the resulting maceration of tissues predisposes them to infection.

1 Ochsner, A., and DeBakey, M.:Peripheral vascular disease; critical survey of its conservative andradicaltreatment. Surg. Gynec. & Obst. 70: 1058-1072, Jun 1940.

2 Ochsner, A.: Indications andtechnic for the interruption of impulses traversing the lumbarsympathetic ganglia. S.Clin. North America 23: 1318-1334, Oct 1943.



  Most of the 887 sympathectomies performed at the three vascular centers in the Zone of Interior during World War II were carried out on the basis of one or the other of the indications just listed. The 672 patients upon whom the operations were performed were all males and over 60 percent were in their twenties. The figures from Ashford General Hospital are typical: The average age of the patients subjected to operation there was about 22 years, the age range 19 to 38 years.

  Six hundred seventy operations were lumbar ganglionectomies and 217 were upper thoracic preganglionic sympathetic ramisectomies. Sympathectomy was performed for 1 extremity in 472 patients, for 2 extremities in 190 (bilateral), for 3 extremities in 5, and for all 4 extremities in 5 patients.

  Lumbar Sympathectomy. The indications for which lumbar sympathectomy was performed, and the proportionate distribution of cases, were as follows

  Trenchfoot and frostbite, with or without gangrene, 43 percent.

Thromboangiitis obliterans of the lower extremities, 19 percent.
  Vascular insufficiency, 19 percent. This group included patients with impaired circulation following ligation of major peripheral arteries; patients with arterial occlusion resulting from arteriosclerosis, thrombosis, and embolism; and patients on whom prophylactic sympathetic denervation was performed in order to facilitate the development of collateral circulation prior to excision of aneurysms of arteriovenous fistulas.

  Raynaud's disease and vasospastic states, 10 percent. This group included not only patients with true Raynaud's disease but also those who complained of sensitivity of cold; those suffering with post-traumatic vaso spastic states in which the factor of abnormal arterial vasoconstriction had been the primary reason for surgery; and those suffering from miscellaneous conditions such as acrocyanosis, hyperhidrosis, and scleroderma.

  Pain, 8 percent. The majority of the patients in this group were suffering from causalgia and from post-traumatic osteoporosis (Sudeck's atrophy).

Thrombophlebitis, 1 percent.

  Thoracic Sympathectomy. The indications for which thoracic sympathectomy was performed, and the proportionate distribution of cases, were as follows:

Raynaud's disease and vasospastic states, 36 percent.

Vascular insufficiency, 35 percent.

  Causalgia and other conditions associated with pain, 23 percent.

  Thromboangiitis obliterans of the upper extremity, 6 percent.


  The vascular centers were equipped with facilities for special diagnostic tests for the evaluation of the circulatory status of patients with disturbances


 of the peripheral circulation.In most instances, prior to sympathectomy, measurements of the skin temperature by means of thermocouples were made in constant temperature rooms, and pulsations of the peripheral arteries were determined by means of the oscillometer. The capacity of the blood vessels to dilate following release of vasoconstrictor tone was also observed.

  The chief indication for sympathectomy, as already noted, is abnormal vasoconstriction or vasospasm. Vasoconstriction is the normal physiologic response of the body to cooling. It occurs both in normal patients and in those with diseased or injured blood vessels.The diagnosis of an abnormal degree of vasoconstriction is simple in such conditions as Raynaud's phenomenon, for instance, in which so-called digital syncope can usually be produced by exposure of the patient to a cold shower, but in less severe cases it is far more difficult to reach a conclusion and diagnostic measures have developed only gradually.

  Brown,3 in 1926, was the first to approach the problem of placing vasomotor tone in man on a quantitative basis by comparing the rise in the skin temperature of the digits with the rise in the oral temperature after the ad ministration of typhoid vaccine; the change in the surface temperature of the foot provided the index of vasodilatation. White,4 in 1930, used paravertebral injection of procaine hydrochloride to block the sympathetic nerves to the extremity as a test to evaluate the benefit of sympathetic ganglionectomy. Morton and Scott,5 in 1931, used general, spinal, and local anesthesia to release vasomotor tone and designated the average maximum vasodilator response for healthy arteries as "the normal vasodilatation level."Gibbon and Landis 6 in 1932, differentiated vasospastic from occlusive arterial disease by the vasodilatation with associated rise in the skin temperature which occurs upon heating unaffected extremities or portions of the body.

  The tests described, which are based on the vasodilating capacity of the extremity, are useful in estimating the degree of organic vascular occlusion present in a patient, but they do not indicate the relative degree of vasoconstriction which intermittently or constantly affects the blood supply to the peripheral tissues. In order to make the diagnosis of abnormal vasoconstriction in this series, it was necessary, therefore, to utilize certain clinical signs.

  The combination of peripheral cyanosis, increased sweating, and constriction of the superficial veins of the extremities used by Freeman and Montgomery 7 as clinical evidence of high vasomotor tone proved useful at the vascular centers in the selection of patients with intermittent claudication

3 Brown, G. E.: Treatment ofperipheral vascular disturbances of the extremities. J.A.M.A.87:379-383, 7 Aug 26.

4 White, J. C.: Diagnosticblocking of sympathetic nerves to extremities with procaine; test toevaluate the benefit ofsympathetic ganglionectomy. J.A.M.A. 94: 1382-1383, 3 May 30.

5 Morton, J. J., and Scott, W. J.M.: Methods for estimating the degree of sympathetic vasoconstrictionin peripheralvascular diseases. New England J. Med. 204:955-962, 7 May 31.

6 Gibbon, J. H., Jr , and Landis,E. M.: Vasodilatation in lower extremities in response to immersingforearms inwarm water. J. Clin. Investigation 11:1019-1036, Sep 1932.

7 Freeman, N. E., and Montgomery,H.: Lumbar sympathectomy in treatment of intermittent claudication; selectionof cases by claudication test with lumbar paravertebral procaineinjection. Am. Heart J. 23:224-242, Feb 1942.


for sympathectomy. It was recognized, however, that a more objectivemethod was desirable.

  White and Smithwick, 8 on the basis of variations in surface temperature of the extremities in patients exposed to low room, temperatures (20?C.), differentiated "cold handed" from "hot handed" individuals. Naide,9 on the basis of skin temperature changes during the cooling period and subsequent vasodilatation stage of the test described by Gibbon and Landis,10 devised a quantitative test for basal vascular tone. Later he found it necessary to note only the response of the digital skin temperature of the hands to the preliminary cooling in order to separate their patients into those with a high degree of vascular tone and those with a normal or low tone. He advocated the use of this test in the selection of patients for sympathectomy because he found that while the collateral circulation after major arterial occlusion was excellent in 88 percent of his patients with low vascular tone, equally good circulation was present in only 34 percent of those with high tone.

  Forty-five patients with arterial injuries were studied by this test at DeWitt General Hospital prior to sympathectomy. The results were in agreement with the clinical evidence in 32 but did not conform with it in 13 patients. The results of operation indicated that the tests were valid in 20 patients, but in 11 others subsequent events showed that it had been misleading. In the remaining patients restoration of the continuity of the artery made postoperative evaluation by this method impossible. In this same group of 31 patients the end results in 26 were what might have been expected from a study of the clinical signs of abnormal vasoconstriction. Naide's test was frequently of value in confirming the clinical evidence of high vasomotor tone though it was occasionally misleading, and only rarely was it of diagnostic significance.

  A possible explanation of the discrepancy between the results of the cold test and the impression gained on clinical observation is that the physiologic stimulus present at the time the examinations were made in the hospitals, such as excitement or a mild emotional disturbance, differed from the stimulus present during exposure to cold. Another possible explanation is differences in the type of stimulus used. Finally, since the arterial occlusion in each case resulted from trauma, it is quite within reason that reflex vasoconstriction might have been set up in the region supplied by the injured vessels; this is a purely local phenomenon which might be quite unaffected by the general vasomotor tone of the upper extremities as measured by the test.

  The experience at the vascular center of DeWitt General Hospital with this and the other tests listed led to the conclusion that a valid qualitative

8 White, James C., and Smithwick, Reginald H.: The Autonomic Nervous System. 2nd ed. New York, The Macmillan Company, 1941.

9 Naide, M.: Test for vasculartone in humansand its application to the study of vascular diseases with specialreference to the etiology and prevention of thrombophlebitis. Am. J. M.Sc. 207:606-620, May 1944.

10 See footnote 6, p.424.


test for the diagnosis of abnormalvasoconstriction was still to be developed. Experience at theother centers were to the same effect.



  All lumbar sympathectomies at the vascular centers in the Zone of Interior were carried out through an extraperitoneal approach. At both Ashford General Hospital and Mayo General Hospital, the operation was per formed through a transverse, anterior, muscle-splitting incision. At DeWitt General Hospital a similar approach was used, with incision of the muscles of the abdominal wall." Spinal analgesia was used in all but a few cases, and in those endotracheal nitrous-oxide-oxygen anesthesia was employed.

  Atlas,12 who made a careful study of the anatomic variations of the lumbar sympathetic nerves, found that a ganglion is constantly present lying on the disc between the second and third lumbar vertebrae and noted recurrence of sweating along the medial side of the foot when it was not removed.This observation was borne in mind in sympathectomies performed at the vascular centers, and in all the second and third lumbar ganglia, with the intervening chain, were excised.Occasionally the first or the fourth lumbar ganglion was also removed.

  In the large series of operations reported from Mayo General Hospital the patients were given a regular diet and allowed to be up on the day after operation. The period of disability was considerably longer following the more extensive muscle-cutting procedure.

Thoracic Sympathectomy

  All upper thoracic or dorsal sympathectomies performed at the vascular centers were, in reality, preganglionic sympathetic ramisectomies, since the ganglia were not removed but were simply decentralized. At Ashford and Mayo the posterior muscle-splitting incision was used. The operation was done by the technique advocated by Smithwick,13 according to which the roots of the second and third intercostal nerves are severed either within, or just outside of, the dura, while the sympathetic chain is divided below the third dorsal ganglion. The upper portion of the sympathetic chain, together with the decentralized second and third thoracic ganglia, is then sutured to the muscles outside of the thoracic cage in order to forestall regeneration.

  In the patients operated on at Mayo General Hospital root section was occasionally omitted. It was carried out in all patients at the other centers. No differences in clinical results were noted.

11Harris, R. I.: Role of sympathectomy in treatment of peripheral vascular disease.Brit. J. Surg. 23:414-424, Oct 1935.
12Atlas, L. N.: Sympathetic denervation limited to the blood vessels of the leg and foot.Ann. Surg. 116:476-479, Sep 1942.

13 Smithwick, R. H.:Problem of producingcomplete and lasting sympathetic denervation of theupper extremity by preganglionic section.Ann. Surg. 112:1085-1100, Dec1940.


  At DeWitt General Hospital preference was given to the anterior approach first described by Royle and later adapted by Gask14 and by Telford 15 in England, and by De Takats16 in this country. The sympathetic chain is exposed behind the pleura through a supraclavicular incision and is divided below the third thoracic ganglion. The connections of the second and third ganglia are then cut and the decentralized chain sutured to the upper end of the scalenus anticus muscle.

  Endotracheal nitrous oxide-oxygen-ether was the anesthetic of choice for this operation. The postoperative discomfort and period of disability were shorter following thoracic sympathectomy through the anterior supraclavicular approach than after the dorsal operation.


Trenchfootand Frostbite

  At the time patients with trenchfoot and frostbite were observed at the vascular centers, which was always several weeks after the initial incident, the acute inflammatory reaction characteristic of these conditions had usually subsided and only disabilities resulting from damage to the soft tissues, peripheral nerves, and blood vessels remained.

  In most instances skin temperature and oscillometric determinations were made, while studies of the response to paravertebral injection of procaine were found to be especially useful in the selection of patients for sympathectomy. The criteria for operation used at Ashford General Hospital, which were much the same as those prevailing at the other centers, included diminution in arterial pulsations at the foot level as determined by oscillometer, skin temperature constantly below 79  ?F., and constant cyanosis and sweating of the feet.

  Generally speaking, results obtained from sympathectomy were both subjectively and objectively beneficial in patients thus selected. In a detailed analysis of one series of 55 patients with trenchfoot and frostbite it was found that in 35,with gangrene, lumbar sympathectomy had been performed in 10 instances bilaterally. Though healing ultimately occurred in all 35, most of the patients required minor amputations and some, various types of skin grafts. Eleven of the 20 patients without gangrene were also treated by sympathectomy of both lower extremities. In this group the complaints of coldness, cyanosis, and annoying sweating of the feet were relieved by the operation, but as a rule complaints of pain and tenderness, present in three-fourths of the patients, were not relieved.

  The results of sympathectomy at the vascular centers suggest that the residual hyperactivity of the sympathetic nervous system following trauma from cold (in addition to the actual destruction of tissue) and characterized

14 Gask, G. E.: Surgery of thesympatheticnervous system. Brit. J. Surg. 21:113-130, Jul 1933.

15 Telford, E. D.: Technique ofsympathectomy. Brit. J. Surg. 23:448-450, Oct 1935.

16 De Takats, G.: Effect ofsympathectomy onperipheral vascular disease. Surgery 2:46-60, Jul 1937.


by vasospasm, increased sweating, andsensitivity to cold can be relieved by this method. On theother hand, pain and tenderness, which are probably secondary to tissuetrauma, usually do notrespond to sympathectomy.


  Before any patient with thromboangiitis obliterans was selected for sympathectomy, his vascular status was determined by means of the oscillometer with skin temperature measurements. Special attention was also paid to the capacity of the collateral vessels to dilate after release of vasomotor tone following spinal analgesia and paravertebral block. In a few patients with marked circulatory impairment, although increase in skin temperatures was slight, considerable improvement in the condition was noted after sympathectomy. As White17 commented, wide clinical experience is essential in selecting the patients for operation from the small group in which sympathectomy has been shown by experience to be sometimes surprisingly worthwhile even though preliminary tests do not indicate a striking relief of vasospasm. Harris,18 in harmony with this concept, wrote that it was his practice to use lumbar sympathectomy in all severe cases of thromboangiitis obliterans regardless of the results of the preoperative tests.

  In the vascular centers it became more and more the tendency in the selection of patients for sympathectomy in cases of thromboangiitis obliterans, to rely on clinical evidence of abnormal vasoconstriction, as shown by cyanosis, sweating, and the presence of constricted veins. Demonstrations of improvement in the ability of the patient to walk after paravertebral block was particularly useful in the selection of individuals for early operation.

  At Ashford General Hospital the indications for sympathectomy were eventually crystallized about as follows: (1) Evidence of a large vasospastic element characterized by cyanosis, coldness, and sweating. (2) Pain, constantly present. (3) Quantitative elevation of temperature after spinal analgesia or sympathetic block. (4) The presence of ulcers or gangrene. At the other centers the indications for sympathectomy were essentially the same. In numerous instances at all centers the indications were multiple.

  Resection of the sympathetic ganglion was carried out for 1 extremity in 33 of the 152 patients with thromboangiitis obliterans whose records were available for this analysis, and for 2 extremities (bilateral) in 18 patients. The operation was also performed in 2 of the patients in whom 3 extremities were affected. Good results were observed in all. Gangrenous areas demarcated and separated promptly and healing was rapid. Ulcers also healed. Resting pain was relieved or partially relieved. Functional capacity, particularly the ability to walk, was greatly increased. Oscillometric readings and measurements of skin temperatures furnished objective evidence of subjective improvement.

17 White, J. C.: Progress insurgery of theautonomic nervous system, 1940-1942.Surgery 15: 491-517, Mar 1944.

18 Harris, R. I.: Obliterativevasculardisease; treatment by sympathectomy. Canad. M.A J. 45:529-533, Dec 1941.


  Minor amputations were necessary in 12 of the patients treated by sympathectomy on the indication of thromboangiitis obliterans, but only 3 of the 53 patients were not improved at all. Major amputation of the leg was required in 2 of these unimproved patients, but in 1 instance it was necessary because of contractures about the knee and ankle and not because of the vascular disease per se. It is of interest that the 3 patients who were not at all improved by sympathectomy had all continued to smoke in spite of advice and prohibitions to the contrary.

  There was general agreement in all the vascular centers that sympathectomy is the treatment of choice in thromboangiitis obliterans, especially in the more severe cases. The disease is frequently associated with abnormal vasoconstriction, and the circulation will be improved to the extent that circulatory deficiency can be relieved by removal of this added factor. On the other hand, De Takats19 warning should be heeded: "sympathectomy deprives the extremity of its vasoconstrictor tone. It does not influence the course of Buerger's disease." In other words, as long as obliteration of blood vessels continues to recur, the disease may be improved or arrested, but it cannot be said to be cured.

Raynaud'sDisease and Vasospastic States

  In the group of patients with Raynaud's disease and other vasospastic conditions, study of the circulation after removal of vasomotor tone was of great significance as a preoperative test. In 1 group of 57 patients the response to paravertebral block was used for this purpose in 37, the effects of spinal anesthesia were observed in 17, and the results of the vasodilatation test (reflex heat) were used in 3 patients. Differentiation was made between patients with true Raynaud's disease and other vasomotor disorders on the basis of the production of actual spasm of digital arteries characterized by digital syncope on immersion of the entire body for 2 minutes in a cold shower. It was sometimes observed that digital syncope was not produced by this procedure even through characteristic attacks were occasioned in cold weather. The occurrence of typical attacks was therefore considered more significant than the results of the test. The urgency of the indications for sympathectomy depended upon the severity and frequency of arterial spasm, especially if destruction of the digital pulp had occurred or was impending. When cold sensitivity and frequency of attacks were incapaciting, sympathectomy was regarded as mandatory.

  At Ashford General Hospital considerable improvement was noted in all 30 patients upon whom sympathectomy was performed for Raynaud's disease. This operation was carried out for involvement of 1 extremity in 2 patients, of 2 extremities (bilateral) in 23, of 3 extremities in 2, and of 4 extremities in 3 patients. Although after operation digital syncope could still be elicited by

19 De Takats, G.: Value of sympathectomy in the treatment of Buerger's disease. Surg., Gynec. & Obst. 79:359-367, Oct 1944.


the cold shower test, it was considerablydiminished in degree, extent, and duration. Within 3 to5 weeks of operation definite evidence of partial resumption ofvasomotor tone in the upperextremity, which could be prevented by blocking the peripheral nerveswith procaine, wasnoticed in 2 patients subjected to multiple procedures. The experienceat Ashford GeneralHospital is in harmony with that reported by the other centers.

  Although patients with Raynaud's disease of the upper extremities were considerably improved by sympathectomy, they were not cured and none could be returned to duty. Results observed after sympathectomy of the lower extremities were, as expected, better than those obtained by operation on the upper extremities. This difference is probably to be explained by the anatomic arrangement of the vasoconstrictor nerves to the upper, in comparison with the lower, limbs.

  Post-traumatic Vasospastic States. Patients with post-traumatic vasospastic states had a variety of complaints of which vasospasm, edema, and pain were the most common. This group was differentiated from patients with causalgia, post-traumatic painful osteoporosis (Sudeck's atrophy), and various reflex dystrophies chiefly on the basis of clinical evidence of abnormal vasoconstriction.They are discussed in Chapter XV.


  Paravertebral injection of procaine into the region of the thoracic or lumbar sympathetic ganglia was the one really significant preoperative diagnostic test in selection for sympathectomy of patients who were suffering from causalgia and other painful conditions of the extremities. The prompt relief of pain after the injection was usually striking and conclusive. Skin temperature and oscillometric determinations were significant only in demonstrating that the sympathetic block had been successful. Subsequent sympathectomy was generally effective in relieving the condition. Equally good results were obtained in patients with pain in either the upper or the lower extremity. Of the 57 patients treated for causalgia by sympathectomy at Mayo General Hospital, excellent results were reported in 46, and good eff ects in 9.In 2 patients the end result was described as fair or poor. Sympathectomy was performed chiefly because of pain in the extremities in 60 patients at Ashford General Hospital. The operation was almost uniformly effective in relieving this symptom.

  The exact mechanism through which sympathectomy relieves pain in the extremities is not fully understood. The dramatic relief afforded by sympathetic block furnishes, however, strong clinical evidence to support the concept that nerves mediating painful sensations from the extremities actually traverse the sympathetic ganglia. Attempts to demonstrate their presence, whether by anatomic studies or by physiologic investigations, have generally been unsuccessful, although some observations in this field are promising. In order to


explain the clinical observations otherconcepts have been advanced. Lewis20 suggested thatthe relief of pain after sympathectomy is the result of thepersistently increased blood flowthrough the cutaneous vessels consequent upon loss of vasomotor tone.The essentialdisturbance, according to Livingston,21 is a "centralperturbation offunction involving thespinal cord centers"; he also suggested that the "sympathetic nervesmay contribute to thedevelopment of peripheral tissue changes, which may lead to additionalafferent impulses."

  Both of these explanations are based on the concept that relief of pain in causalgia depends on paralysis of efferent sympathetic nerves with consequent loss of vasomotor tone. Clinical evidence, however, suggest that the relief observed depends on the suppression of afferent sensory stimuli. Only the failure of experimental investigation to reveal the transmission of pain stimuli from the extremities across sympathetic ganglia prevents the acceptance of the simpler explanation which accords so readily with clinical observations.

  Probably the most reasonable explanation of causalgic pain is that advanced by Doupe 22 and coworkers who ascribe its peculiar qualities to direct cross stimulation of sensory fibers by efferent sympathetic impulses at the, point where the nerve trunk is injured, rather than to the indirect action of the vasoconstrictor response which they also produce. Final explanation of the mechanism of the relief achieved in causalgia and other painful conditions of the extremities by sympathectomy or by injections of procaine around the sympathetic ganglia must await further experimental evidence. There was general agreement in the vascular centers, however, that sympathectomy, whatever the mechanism, was quite effective in the treatment of these painful conditions.

Vascular Insufficiency

  Records show that in the category of vascular insufficiency were included (1) those patients with impaired circulation following ligation of major vessels of the extremity, (2) those in whom arterial thrombosis developed at the site of trauma, and (3) those in whom sympathectomy was performed as a prophylactic measure prior to operation for arterial lesions. Sympathectomy was performed in well over half of these patients for symptoms arising from obstruction of the main arterial supply to the limb. Preoperative tests included skin temperature and oscillometric determinations, also observation of the vascular supply after spinal analgesia or paravertebral block.

  The chief indications for sympathectomy in patients in the first two Categories were impaired circulation, nerve paralysis, gangrene, and ulceration. The results of sympathectomy in this group were extremely satisfactory. In spite of the severe impairment of circulation, improvement was noted in all patients. Ulceration and gangrenous areas healed although minor amputations

20 Lewis, Thomas: Pain. New York, The Macmillan Company, 1942.

21 Livingston, William K.: PainMechanisms. APhysiologic Interpretation of Causalgia and Its Related States. NewYork, The Macmillan Company, 1943.

22 Doupe, J.; Cullen, C. H., andChance, G.Q.: Post-traumatic pain and the causalgic syndrome. J. Neurol.,Neurosurg. & Psychiat. 7:33-48, Jan-Apr 1944.


and skin grafts were necessary in some.Satisfactory progress in nerve recovery took place.Amputation of the extremity was ultimately necessary in only 2patients, in 1 of whomcirculatory impairment was not the cause; the operation was done forcontracture and recurrentosteomyelitis.

  The chief indications for sympathectomy after excision of aneurysms or arteriovenous fistulas were sensitivity to cold, nerve damage, impaired circulation, and pain. In all instances cold sensitivity and pain were relieved by operation. Improvement in the function of peripheral nerves also resulted. Amputation was necessary in only one patient; thrombosis of the collateral vessels developed after excision of a femoral arteriovenous aneurysm. The circulation improved following sympathectomy, but gangrene of the foot developed later and amputation became imperative.

  Ninety-three of the 155 sympathectomies available for analysis and categorized as operations for vascular insufficiency were performed for prophylactic reasons in order to facilitate the development of collateral circulation before operation. In spite of the arterial lesions circulation was adequately maintained in the majority of the patients before operation, but there were 2 instances of gangrene and 5 others in which circulation was impaired. Associated nerve damage was present in 25 patients but in all improvement followed sympathectomy.

  The chief indication for sympathectomy in the group of prophylactic operations was poor collateral circulation as determined, prior to excision of the arterial aneurysm or arteriovenous fistula, by the Matas-Moschcowitz test: The main artery was temporarily occluded at the site of the arterial lesion, the distal tissues were rendered ischemic for 5 minutes by an Esmarch bandage, and the time necessary for flushing of the extremity after release of the bandage, noted. The collateral circulation was regarded as good if, while the main artery was still occluded, color returned to the distal portion of the extremity in 60 seconds or less. If flushing was delayed more than 3 minutes, the collateral circulation was considered unsatisfactory.

  In the patients in this group observed at DeWitt General Hospital the average time required for color to return to the extremity before sympathectomy was 3 minutes and 15 seconds. After the operation the time was reduced to 45 seconds. Definite improvement in the collateral circulation, as determined by the fluorescein wheal test, was also noted in these patients.

  The indications for sympathectomy following traumatic or therapeutic occlusion of major arteries gave rise to no differences of opinion when actual gangrene or ulceration, ischemic paralysis, pain, and other results of impaired circulation were present. All of these are conditions known to be greatly improved by sympathectomy. They may also be prevented if sympathectomy is performed as soon as vascular impairment becomes evident, as advocated by Learmonth 23 in the treatment of acute vascular injuries. In two patients

23Learmonth, J. R.: Surgery of blood vessels. Edinburgh M. J. 47: 225-240, Apr 1940.


included in this series the foot became extremely ischemic after ligation of the popliteal artery for an arterial lesion. Sympathectomy was performed immediately, in each instance while the patient was still under the original anesthetic, and results in both patients were excellent.

  There was not complete agreement at the vascular centers, however, over the advisability of sympathectomy as a prophylactic procedure either before or after excision of aneurysms and arteriovenous fistulas. The surgeons at Mayo General Hospital and DeWitt General Hospital were in favor of the procedure.The surgeons at the vascular center of Ashford General Hospital were not. In contrast to 149 operations performed for vascular insufficiency at Mayo and DeWitt General Hospitals, only 6 of the 469 sympathectomies performed at the Ashford center were for this condition. Elkin, at Ashford, has repeatedly stated that it is his personal belief, based on experience, that sympathectomy is not necessary in such cases if sufficient time has been allowed to elapse for a collateral circulation to develop. In his opinion, 24 'It is an assumption, and nothing more or less than that, that gangrene is prevented by sympathectomy." His extensive experience in vascular surgery makes it quite clear that sympathectomy is not required to prevent gangrene, and there were no instances of it following excision of aneurysms or arteriovenous fistulas at the Ashford General Hospital where interruption of the sympathetic pathways has been omitted.

  On the other hand, in spite of the Ashford experience, the question arises as to the degree of vascular insufficiency which persists after operation. The collateral circulation in the patients managed at that hospital without sympathectomy may have been quite adequate to maintain tissue nutrition while they were at rest so that gangrene did not develop after operation, but whether it was sufficient to permit normal activity after they left the hospital is another matter. The improvement noted at the other centers after sympathectomy in patients who had undergone ligation of major arteries, and in others who complained of symptoms of vascular insufficiency after excision of arterial lesions, indicated the value of this procedure in these circumstances. Bigger's 25 experience may also be cited in this connection. He wrote:

  In the literature dealing with the treatment of traumatic vascular lesions of important blood vessels, much is said, and properly, about the cure of the lesion and the avoidance of acute ischemic gangrene, but remarkably little consideration is given to the permanent reduction of blood supply to the tissues, especially the muscles, distal to the lesion.

  Bigger noted excellent immediate results following excision of arteriovenous fistulas when sympathectomy was omitted but found late results less satisfactory. Of 8 patients examined between 9 months and 8 years after operation, 7 had marked symptoms of chronic circulatory deficiency distal to the obstruction. Bigger therefore concluded that permanent interruption of

24 Elkin, D. C.: Personalcommunication tothe author.

25 Bigger, I. A.: Treatment oftraumaticaneurysms and arteriovenous fistulas. Arch. Surg. 49:170-179, Sep 1944.


the sympathetic nerves to the extremity mayhelp prevent chronic circulatory deficiency distal tothe obstruction. Others share his opinion. Learmonth 26advocatedsympathectomy in allpatients in whom permanent ligation of the femoral or poplitealarteries is expected and Gageand Ochsner 27 recommended chemical section of thesympathetic nervesbefore operations onmajor arteries.

  Since the development of a collateral circulation determines the ultimate outcome in vascular operations involving ligations of major arteries to the extremities, and since abnormal vasoconstriction apparently interferes with the development of collaterals, sympathectomy should perhaps be reserved for those patients who show evidence of increased vasomotor tone. Prophylactic sympathectomy, although it increases the assurance of an adequate postoperative blood supply, might well be reserved for those patients in whom the location of the arterial lesions and the results of the preoperative tests indicate that severe disturbances of circulation may follow ligation of the major artery to the extremity. Even in those patients it might be justifiable to hold the operation in reserve and to perform it only in the event that arterial ligation leaves an incapacitating circulatory disturbance.

  Unnecessary Prophylactic Sympathectomy. Though preoperative sympathectomy improves the collateral circulation in patients with aneurysms and arteriovenous fistulas upon whom vascular surgery is to be performed, pro phylactic operation is not always necessary. Spontaneous closure of the fistula and spontaneous cure of an aneurysm may occur (see Chapter XII). Moreover, the main artery to the extremity may prove not to be the one involved, or subsequent arterial repair may be found to be possible.

  Clinically satisfactory spontaneous closure of a subclavian arteriovenous fistula and spontaneous cure of a femoral arterial aneurysm were observed following sympathectomy at one vascular center and a case similar to the latter was observed at another center. Sympathectomy, however, cannot be given full responsibility for inducing the closure since spontaneous cures were recorded in similar cases in which sympathectomy was not carried out. It is self-evident that in these instances prophylactic sympathectomy was unnecessary. On the other hand, spontaneous cure of an aneurysm or spontaneous closure of an arteriovenous fistula may occasionally be followed by sufficient impairment of circulation to necessitate subsequent sympathectomy.

  Inaccurate localization of an arterial lesion before operation may also lead to unnecessary sympathectomy. A branch rather than the main artery may be the vessel involved. This happened in several instances in this combined series. In 4 instances the lateral circumflex femoral or the profunda femoris artery was involved instead of the femoral artery. In 2 other instances a geniculate artery was involved instead of the popliteal as had been supposed. In still another instance the transverse cervical branch, instead of the sub-

26  See footnote 23, p. 432.
27 Gage, M., and Ochsner, A.: 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.


clavian artery, was the site of the arteriallesion. In 1 case reported from a vascular center anarteriovenous fistula, which involved the arch of the aorta and thesuperior vena cava, could notbe excised and sympathectomy of the left upper extremity was thereforeperformedunnecessarily.

  Restoration of the continuity of the artery instead of excision of the lesion may also make preliminary sympathectomy unnecessary. In several instances of arteriovenous fistulas in this series, and in a small number of arterial aneu rysms, successful repair of the arteries and restoration of the circulation rendered the prophylactic sympathectomy which had been performed quite superfluous. It may be argued that preliminary sympathectomy possibly contributed to the success of the arterial repair. In 8 patients, however, suture of the artery was successfully employed without preliminary sympathectomy, while in 2 instances, in both of which sympathectomy had been done, arterial repair was not successful.


  Many patients with longstanding thrombophlebitis were referred to the vascular centers for treatment of the resulting disabilities which included cyanosis, sweating, diminished peripheral circulation, and pain. Temporary relief following paravertebral block suggested that lumbar sympathectomy might be of benefit in these patients. The value of this method in the early treatment of thrombophlebitis of the lower extremities, first advocated by Leriche and Kunlin 28 had been well established before World War II by the subsequent studies of Ochsner and DeBakey. 29

  Seven patients referred to the vascular centers for the treatment of late thrombophlebitis were treated by this method. Although the sweating and peripheral vasoconstriction present were abolished and pain was moderately relieved, congestion and edema caused by venous obstruction were not benefited. In 3 or 4 patients operated on at Ashford General Hospital on this indication, complaints of cyanosis, edema, congestion of the limb, and pain on walking were increased after the operation. The other patient was slightly improved. On the face of the evidence, therefore, sympathectomy does not seem of value in the treatment of the late residua of thrombophlebitis.


  The 887 sympathectomies performed in the vascular centers in the Zone of Interior were carried out without a single fatality attributable to the operation. This record may in part be ascribed to the fact that the patients were mostly young and healthy, and in part to the fact that sympathectomy has become a standardized procedure which, in the hands of well-qualified surgeons, should always carry small risk.

28 Leriche, R., and Kunlin, J.: Traitement imm?diat des phl?bites post-op?ratoires par l'infiltration novocainique du sympathique lombaire. Presse m?d. 42:1481-1482, 22 Sep 34.

29 Ochsner, A., and DeBakey, M.:Treatment ofthrombophlebitis by novocain block of sympathetics; technique ofinjection. Surgery 5:491-497, Apr 1939.


  The possibility of sterility as the result of interference with ejaculation after bilateral lumbar sympathectomy might constitute a definite contraindication to the performance of this operation, especially in young men. Ochsner30 has maintained that bilateral resection of the first, lumbar interferes with ejaculation, and White and Smithwick31 stated "After removal of the [superior hypogastric] plexus or injury to the first lumbar ganglia the power of ejaculation is lost," but added in a footnote "We are obtaining increasing clinical evidence that males may not be sterile or have any alteration of sexual function after bilateral excision of the first or even the upper three lumbar ganglia." In the majority of sympathectomies performed at the vascular centers, the first lumbar ganglion was not removed. Lake 32 found in 2 patients normal numbers of spermatozoa after bilateral resection of the second, third, and fourth lumbar ganglia, and De Takats and Helfrich 33 reported similar findings after extensive bilateral lumbodorsal sympathectomy.

  Particular attention was paid to this problem in studying the patients upon whom sympathectomy was performed. No patients complained of difficulty with ejaculation in the group reported from Mayo General Hospital. Three patients at Ashford General Hospital claimed temporary impotence but the bilateral operation had not been performed in any of them and it seems probable that the condition was on an emotional basis.

  Complications and Sequelae. While it was inevitable that a certain number of incidental complications should have occurred, such as infections and hematomas in operative wounds, these were infrequent. Pneumothorax, once unilateral and once bilateral, followed thoracic sympathectomy in 2 instances. It was readily treated by aspiration of the pleural cavity. Inadvertent resection of the fourth rib instead of the third, during sympathectomy by the dorsal approach, led to incomplete denervation in 1 patient. Transient spinal radiculitis developed in 1 patient after dorsal sympathectomy.

  Following lumbar sympathectomy a number of patients complained of pain in the thigh and of paresthesias in the distribution of the genitofemoral nerve. The incidence of this complication was reported as 25 percent from Ashford General Hospital and was about the same at the DeWitt General Hospital. On the other hand, it was notably lower at Mayo General Hospital-not more than 5 percent--and in only 2 patients did the pain persist for more than 10 days. The infrequent occurrence of this troublesome symptom at this center may possibly be explained by the more limited operative exposure which was used. The pain in the thigh, although annoying, disappeared spontaneously. Sweating of the body may be greatly increased after

30 See footnote 2, p. 422.

31 See footnote 8, p. 425.

32 Lake, N. C.: Sympathectomy andsterility.Brit. M. J. 1:843, 24 Jun 44.

33 De Takats, G., and Helfrich, L.S.:Sterility of the male after sympathectomies. J.A.M.A. 117:20-21, 5 Jul41.


quadrilateral sympathectomy, with resultantparalysis of the sweat glands of all 4 extremities,and 2 patients complained of excessive perspiration after theiroperations.

  After thoracic sympathectomy partial recurrence of vasospasm characterized by digital syncope was frequently observed, especially in patients with Raynaud's disease, when the cold shower test was given. This was not true of the patients upon whom the lumbar operation had been performed. This observation, which is in keeping with that of the majority of surgeons working in this field, has been the subject of considerable controversy. Lewis 34 noted that vasoconstriction still took place in the fingers after excision of the stellate and upper thoracic ganglia. He therefore ascribed the vasospasm in Raynaud's disease to "local fault," the assumption being that the blood vessels of the digits are abnormally sensitive to cold while the vasomotor reactions are normal. The possibility that the recurrent vasoconstriction is caused by sensitization of the blood vessels to circulating adrenalin as the result of sympathetic ganglionectomy was advanced by Freeman, Smithwick, and White.35 It was to avoid this sensitization phenomenon that Smithwick 36 as suggested the preganglionic type of sympathectomy which has now come to be so extensively practiced. Even after this operation, however, residual vasospasm is frequently noted in patients with Raynaud's disease. It was reported from all of the vascular centers.

  The fact that blocking the peripheral nerves with procaine, as previously noted, was followed in two patients by noticeable elevation in temperature of the digits (an observation which confirms those of other investigators) indicates that there is some persistent innervation by vasoconstrictor fibers. Kuntz and Dillon 37 attributed residual vasoconstriction to efferent fibers arising from the first dorsal root, since they had found evidence of such neural pathways in the experimental animal. Ray, Hinsey, and Geohegan 38 showed by stimulation of the anterior roots in man that preganglionic impulses may be transmitted through the first dorsal root. Regeneration of preganglionic nerves has also been suggested as an explanation.Obviously, the problem of how to produce a lasting preganglionic denervation of the upper extremity is not yet fully solved.

34 Lewis, T.: Experiments relatingto theperipheral mechanism involved in spasmodic arrest of the circulation inthefingers, variety of Raynaud's disease. Heart 15:7-101, Aug 1929.

35 Freeman, N. E.; Smithwick, R.H., andWhite, J. C.: Adrenal secretion in man; reactions of blood vessels ofhuman extremity, sensitized by sympathectomy to adrenalin and toadrenal secretion resulting from insulinhypoglycemia.Am. J. Physiol. 107:529-534, Mar 1934.

36 See footnote 13, p. 426.

37 Kuntz, A., and Dillon, J. B.:Preganglionic components of the first thoracic nerve; their role insympatheticinnervation of the upper extremity.Arch. Surg. 44:772-778, Apr 1942.

38 Ray, B. S.; Hinsey, J. C., andGeohegan,W. A.: Observations on the distribution of the sympathetic nerves tothepupil and upper extremity as determined by stimulation of the anteriorroots in man. Ann. Surg. 118:647-655, Oct1943.