U.S. flag

An official website of the United States government

Skip to main content
Return to topReturn to top



Arterial Aneurysms and Arterial Fistulas

Anticoagulant therapy in Reparative Surgery

Harris B. Shumacker, Jr., M.D., David L. Abramson, M.D. *, and Herbert H. Lampert, M.D. **

    Prior to World War II occasional reports appeared in the literature concerning the clinical and experimental use of anticoagulants in arterial surgery. All, however, were limited to the use of heparin. 1 The anticoagulant dicumarol was not introduced until after the beginning of the war.2 While these two agents were not generally used in the vascular centers at the general hospitals in the Zone of Interior, they were employed occasionally on special indications at some of the centers. At Mayo General Hospital where they were used, a study was undertaken to test their efficacy in the prevention of thrombosis following arterial repair. The patients chosen for this study were those in whom some type of reparative or restorative procedure had been employed in an effort to maintain continuity of the artery following surgical obliteration of a peripheral aneurysm or arteriovenous fistula.


    Some type of reparative surgery (transfixion of fistula, lateral arteriorrhaphy, end-to-end suture, or vein transplantation) was carried out in 34 patients in a series of 288 aneurysms or arteriovenous fistulas treated at this installation. Twenty-two of these 34 patients received anticoagulant therapy for a more or less prolonged period of time. An additional patient received a single injection of heparin at the time of operation. Eighteen received combined heparin-dicumarol therapy after operation, 2 received dicumarol before operation and combined therapy afterward, 2 received only dicumarol which was given both before and after operation. In these 23 patients there were 3 instances of postoperative thrombosis.

*Clinical Associate Professor of Medicine, University of Illinois College of Medicine. Formerly Major, MC, AUS.
**Senior Clinical Physician in Internal Medicine and Cardiology, Mount Sinai Hospital. Formerly Major, MC, AUS.
1 (1) Murray, G. D. W.: Heparin in surgical treatment of blood vessels. Arch. Surg. 40: 307-325, Feb 1940.

  (2) Murray, G. D. W.: Heparin in thrombosis and blood vessel surgery. Surg., Gynec. & Obst. 72: 340-344, Feb 1941.

2  Link, K. P.: The anticoagulant from spoiled sweet clover hay. Harvey Lecture Series 39: 162-216, 1943-44.


    Of these 34 patients treated by reparative arterial surgery, 11 did not receive anticoagulant therapy. In 1 of these patients postoperative thrombosis occurred.

    In the majority of patients in this series, heparin was administered on the operating table as soon as the decision was made to attempt arterial repair and was continued for a period of hours until dicumarol, administration of which had been begun as soon after operation as the prothrombin time could be determined, had produced a satisfactory alteration of the prothrombin level.

    Crystalline heparin in aqueous solution was given intravenously in 50-mg. doses every 4 hours for a period of 48 hours. During this time its anticoagulant effect was measured occasionally, but not at regular intervals, by determining the clotting time according to the method of Lee and White.3

    The method of Quick 4 was used for the determination of the prothrombin level. In general, 300 mg. of dicumarol were given the first day, 200 mg. the second day, and 100 mg. the third day. Thereafter the dosage for each patient was determined daily on the basis of the level of his prothrombin time. The exact quantity required to maintain the prothrombin time at the desired level varied for the different patients and not infrequently for the same patient during the course of therapy. Generally a daily dose of 100 mg. was necessary, although in some instances smaller amounts were adequate. When a patient was being given heparin and dicumarol concurrently, blood for prothrombin determinations was drawn just before the administration of a dose of heparin in order to minimize any possible effect of the latter upon the prothrombin blood level determination. An effort was made to maintain the prothrombin level in the neighborhood of 20 to 30 percent of normal according to the Quick curve;5 this is roughly equivalent to a "clotting index" of 50. A control test using normal plasma was run with each daily set of determinations.

It was found to be a wise precaution todelegate dicumarol therapy to a single member of thestaff. Each morning a report of the prothrombin determinations of allpatients receiving thistherapy was submitted to the responsible staff member. On the basis ofthis information heissued orders regarding each patient's dosage of dicumarol for that day.

    Eighteen patients in this series of 23 received combined heparin-dicumarol therapy after operation. In 1, ligation and transfixion of a fistula had been performed, in 3 lateral arteriorrhaphy, in 10 end-to-end suture, and in 4 vein transplantation. In these patients the administration of heparin was started during the operation procedure and continued for an average of 48 hours; dicumarol, which was administered as soon after operation as a prothrombin determination could be obtained, was usually continued for 21 days. In a few patients dicumarol was given for periods ranging from 10 to 16 days; in 4 patients it was given for even shorter periods of time.

3 Lee, R. I., and White, P. D.: A clinical study of coagulation time of blood. Am. J. M. Sc. 145: 495-503, 1913.

4 Quick, A. J.: Clinicalapplication of hippuric acid and prothrombin tests. Am. J. Clin. Path.10:222-233, Mar 1940.

5 Quick, A. J.: Nature ofbleeding in jaundice. J. A. M. A. 110: 1658-1662, 14 May 38.


    In each of these 4 patients there was a special reason for discontinuing the therapy. In the first patient the response was unusually prompt and the prothrombin level at the end of 6 days abnormally low. In the second patient lateral suture had been performed in the presence of gross infection. Since the prognosis for healing was poor and hemorrhage a possibility, dicumarol was discontinued after the third day. In the third patient dicumarol was discontinued on the fourth day because of persistent bleeding from the wound. In the fourth patient, in which a brachial aneurysm was excised and end-to-end suture accomplished, the usual heparin-dicumarol routine was instituted, but through error the patient received 300 mg. of dicumarol every 4 hours for 5 doses, in addition to 50 mg. of heparin every 4 hours for the same number of doses. When the mistake was discovered anticoagulant therapy was discontinued at once and 60 mg. of synthetic vitamin K were administered intravenously. The prothrombin level remained fairly low for several days, but no hemorrhagic difficulties ensued and the patient's condition was not alarming at any time.

    In 4 patients in this series of 23, dicumarol was given before operation and an adequate reduction in prothrombin level had already been attained by the time of operation. Two of these patients received, in addition, single doses of heparin during the operation. In all 4 patients dicumarol was continued for approximately 21 days.

    One patient in this series received no anticoagulant treatment other than a single dose of heparin given at the time of operation.


    Of the 23 patients receiving anticoagulant therapy 15 recovered with no complications, in 3 thrombosis developed, and in 5 some later difficulties with bleeding occurred. It should be emphasized that in spite of the thrombosis, there was no evidence of propagation of the clot in any of the 3 patients.No difficulty with hemostasis was experienced at operation in any case in this series regardless of whether heparin was administered immediately before the anastomosis was accomplished, or whether the prothrombin time had been altered before the operation through the administration of dicumarol. Fibrin foam was frequently placed in the wound as a safeguard and this precaution appeared to give some protection against bleeding from the operative site without adding any risk of intravascular clotting.

    Details of the cases in which complication followed the operative procedure are as follows: In 3 patients receiving anticoagulant therapy, thrombosis developed.In 2 of these patients, end-to-end suture had been performed for a traumatic aneurysm of the distal end of the brachial artery. In both instances the proximal end of the artery was perfectly normal, but the distal segment was somewhat scarred. Further resection of that end, in order to have available for suture a longer strip of normal artery, was impossible because of the


proximity of the bifurcation of the artery into its radial and ulnar branches. Both patients received heparin immediately after operation. The first also received dicumarol for a period of 14 days. The second was the patient mentioned previously who through error was given a large amount of dicumarol during the first 20 hours following operation and following this discovery no anticoagulants at all. The thrombosis which developed in both patients can very logically be attributed to the local arterial damage.

    The third patient in whom thrombosis developed had had a defect in the brachial artery repaired by vein graft. An adequate prothrombin level had already been achieved with dicumarol at the time of operation. A single dose of heparin was given during the operative procedure and dicumarol was continued for about 14 days thereafter. While no explanation for the thrombosis in this patient is apparent, the artery was unusually small and it is not unlikely that an imperfect suture may have been the cause even though the completed anastomosis appeared to be satisfactory at the time of operation.

    In 5 patients difficulty with bleeding was encountered. In 1 patient a hematoma of moderate size developed in the wound. Lateral suture of the subclavian artery had been performed and the patient was receiving dicumarol. The operative site was explored on the 10th day and a large clot evacuated. No bleeding was encountered and the wound was closed; convalescence was thereafter uneventful. A second patient, in whom a vein graft to the popliteal artery had been performed, had a small hematoma which was evacuated without difficulty. There was no further bleeding during the 6-week period in which dicumarol was administered.

    A third patient developed a large hematoma after excision of a femoral arteriovenous fistula repaired by vein transplant. Two hours after the operation brisk bleeding occurred. The wound was explored and a small branch to the muscle found to be the source of the bleeding.  This artery was ligated, the wound closed, and no further difficulty occurred although dicumarol was continued for several weeks.

    In a fourth patient a slow, steady oozing of bright red blood from the wound was noted shortly after a femoral arteriovenous fistula had been excised and a vein graft performed. Heparin was continued intermittently in spite of the bleeding until the fourth day after operation when an adequate prothrombin level was obtained with dicumarol. By this time the blood loss had been sufficient to result in a significantly lowered erythrocyte count. The patient was given a transfusion of whole blood and 50 mg. of synthetic vitamin K. The wound was thereafter explored, but was closed promptly when only diffuse capillary bleeding was found. No further bleeding occurred.

    It is noteworthy that in none of these 4 patients did thrombosis of the repaired arterial segment take place.

    In a fifth patient the brachial artery remained patent after a lateral suture in the presence of gross infection. Hemorrhage, however, occurred through


the sutured defect on the 13th postoperative day. Exploration of thewound revealed that nohealing of the suture line had taken place. The segment was excised andthe artery ligated. Theresults in this patient were good. Since dicumarol had beendiscontinued for 10 days before thebleeding took place, it seems unlikely that this agent was implicatedin the bleeding.Hemorrhage was probably caused by nonhealing consequent to infection.

Analysis of Results in Patients Not Receiving Anticoagulant Therapy

    No control group, as such, existed for the patients who received anticoagulant therapy following arterial repair. Anticoagulants were reserved, as a rule, for those in whom the type of repair was such that the hazard of thrombosis was greater. Since the 11 patients in this series denied this therapy were those with the simpler types of repair, they cannot be considered proper controls. Notwithstanding, the results in these patients are considered of value for comparative purposes.

    In 9 of the 11 patients not receiving anticoagulants following arterial repair, recovery was uncomplicated. In 1 patient thrombosis developed and in 1 the fistula recurred. In the former, a femoral arteriovenous fistula was resected and the rent in the artery repaired by lateral suture. The success of the operation seemed questionable when it was performed because there was obvious injury to the arterial wall in the neighborhood of the fistula and because the vessel was constricted to about one-half its normal diameter by the suture. Blood flowed freely through the segment, however, and it was decided to do nothing further. Two hours after operation it was apparent from the absence of pulsations in the popliteal, the dorsalis pedis, and the posterior tibial arteries that thrombosis had occurred. When the wound was reexplored a thrombus was found which was sharply limited to the repaired segment of the artery and was without proximal or distal propagation. The segment was excised and the artery ligated. In retrospect it would seem that in this instance the segment should have been excised originally and vein transplantation performed. The use of anticoagulants also would have been wise.

    In a second patient a fistula was ligated and transfixed, but division of the vein was not performed nor a cuff of vein used to buttress the ligated fistula. This was the only instance in this series in which this technique was used. Recurrence was prompt and subsequent excision and quadruple ligation of the vessels were required.

    In the remaining 9 patients not treated by anticoagulants there were no untoward results.


    The efficacy of heparin in the prevention of thrombosis in aneurysms treated by arterial repair had been suggested by experimental and clinical evidence for some time before World War II. Later it became evident that


dicumarol was of benefit in reducing the incidence of postoperativevenous thrombosis andpulmonary embolism and, on the basis of experience in other types ofsurgery, it seemed likelythat it would be equally effective in reducing the incidence ofpostoperative thrombosis afterarterial repair. The experience at Mayo General Hospital does notfurnish complete proof of itseffectiveness, however, since no true control group existed and sincemost patients in this seriesreceived heparin as well as dicumarol. Untreated cases recorded in theliterature cannot be usedfor controls because so many other factors are important in evaluatingthe success of arterialsurgery, as, for example, the type of repair, the proficiency of thesurgeon, the presence orabsence of infection, and the degree and extent of the local arterialinjury or disease.

    It is believed that dicumarol was of benefit in preventing arterial thrombosis in the present series, but that it is not a complete safeguard is suggested by the observation that adequate and prolonged anticoagulation therapy did not prevent thrombosis in all patients. The conclusion therefore seems warranted that, helpful as anticoagulants may be, their use will not be attended with success unless (1) local damage to the sutured artery is not great and (2) the surgical repair is properly executed.

    It is of interest, as mentioned previously, that in none of the 3 patients in whom thrombosis occurred following anticoagulant therapy was there any evidence of extension of the clot from the repaired segment. However, in the 1 patient who did not receive anticoagulants following arterial repair and in whom thrombosis occurred, there was also no evidence of extension of the clot. An analysis of these 4 cases suggests the extreme importance of the maintenance of a good blood flow in the distal segment through adequate collateral channels and it seems reasonable to assume that anticoagulant therapy will prove of aid in the important problem of prevention of distal extension of the clot.

    It would seem pertinent to this discussion to mention the results, from the standpoint of postoperative thrombosis, in the other 254 lesions treated at this center. In this group the affected artery was ligated and divided and anticoagulants were not used prophylactically.In two patients postoperative thrombosis occurred.

    In the first patient a partial hemiplegia, presumably caused by distal propagation of a thrombus, developed some hours after ligation of the internal carotid artery for an intracranial carotid aneurysm. This patient recovered.

    In the second patient disaster of great moment occurred. The patient had a femoral arteriovenous fistula, just distal to the point at which the profunda is given off, which necessitated excision and quadruple ligation of the vessels. Good circulation was maintained in the foot until the sixth postoperative day when there was sudden intense pain in the calf and the foot.This was followed by swelling of the leg, coldness and pallor of the foot.The patient stated that this foot was numb. On examination it was evident that extensive venous


and arterial thrombosis had occurred. In spite of poor response tospinal analgesia in respect toboth color and temperature, sympathectomy was performed as a lastresort. The warmth andsensation of the foot were significantly improved by this procedure,but the increase incirculation was insufficient to prevent gangrene of the sole of thefoot and amputation wassubsequently necessary. Anticoagulants were begun shortly after thedifficulty was first noted,but no apparent therapeutic effect resulted.

    Although this was the only case in the large experience at the Mayo General Hospital in which postoperative thrombosis occurred in an extremity after nonreparative surgery for aneurysm or arteriovenous fistula, it suggests the thought that the routine use of anticoagulants in the surgery of peripheral aneurysms and fistulas, whether or not reparative surgery is done, may be valuable in avoiding the uncommon but disastrous occurrence of postoperative thrombosis. Such a program would undoubtedly result in a higher instance of hematomas and oozing, but these difficulties, experience indicates, would probably not be serious.

    With regard to the choice of anticoagulant agents dicumarol has apparent advantages over heparin both in cost and in ease of administration. It must not, however, be used in any institution in which it is not possible and economically feasible to provide facilities for accurate daily determinations of the blood prothrombin level. Its uncontrolled administration is fraught with great danger, and under these circumstances heparin can be employed more safely.

    The experience at the vascular center of Mayo General Hospital suggests that it is more desirable to obtain an adequate prothrombin level before operation by means of dicumarol than to begin anticoagulant therapy with heparin at the time of the surgical procedure. Under this plan, there is no necessity to use heparin intravenously during the first few days after operation.6


    Although the clinical applicability of anticoagulant therapy cannot be precisely defined from the Mayo General Hospital experience, it is evident that this method must be used on very strict indications and contraindications. The experience at this installation suggests that surgery of the peripheral arteries can be undertaken safely at a time when a full anticoagulant effect has already been obtained either from dicumarol or heparin. In such cases, however, it is imperative that adequate hemostasis be achieved before the wound is closed. Fibrin foam proved helpful in control of capillary bleeding, and other coagulant sponges would probably be equally useful. In spite of such precautions it must be assumed that the use of anticoagulant agents will result in

6 Additional postwar studies on the efficacy of these agents in the prevention of thrombosis were carried out by one of the authors of this chapter. See Kiesewetter, W. B., and Shumacker, H. B., Jr.: Experimental study of comparative efficacy of heparin and dicumarol in prevention of arterial and venous thrombosis. Surg., Gynec. & Obst. 86: 687-702, Jun 1948. Ed.


certain increases in the incidence ofhematoma and in persistent oozing from wounds, though thepossible dangers of such complications are minimal if they arerecognized promptly and treatedproperly by withdrawal of anticoagulant therapy, by surgery, if that isindicated, and by the useof reversing agents such as synthetic vitamin K.

    Although no proof of theeffectiveness of anticoagulants was furnished by this study, theimpression was gained. that this therapy renders less likely thrombosisof the repaired segmentof an artery.