Endothelial William Mark E.
Response
to Venous
Krupski, MD; Erwin R. Thal, MD; Bruce L. Gewertz, MD; L. Murphy, PhD; Herbert K. Hagler, PhD; William J. Fry, MD
\s=b\This investigation characterized venous endothelial healing after surgical manipulation. Procedures were performed on jugular and femoral veins in 21 mongrel dogs without systemic anticoagulation. Veins were harvested at varying intervals and vessel structure evaluated with light, transmission, and scanning electron microscopy. Veins that were mobilized or stripped of adventitia demonstrated 25% to 50% endothelial loss at one hour. Endothelial damage was rapidly repaired with complete healing observed in some veins at 48 hours. Tourniquets and clamps resulted in prominent medial and endothelial injury at occlusion sites. Eighteen of 24 transected veins remained patent for the study period. Endothelial healing was unaffected by tension at anastomoses. These observations confirm that venous endothelium receives nutrition by luminal diffusion. The healing process of venous anastomoses is characterized by an early fibrin sleeve sealing the anastomotic site; endothelial bridging of defects can be noticeably delayed by excessive fibrin
deposition. (Arch Surg 114:1240-1248, 1979)
Maximilian
Injury
Buja, MD;
recent reports have shown that the incidence of thromboembolism is actually higher in patients whose venous injuries are treated by ligation, thrombosis after peripheral vein repair remains a frequent complication.'-" Early patency rates in canine models range from 55% to 100%; late patency rates are generally higher due to recanalization.1!'-- Demonstration of altered blood coagulation in postoperative patients has led some investigators to advocate the use of anticoagulant and antiplatelet agents after venous anastomosis.21 Although it is generally accepted that venous anasto¬ moses are thrombogenic, the natural healing process of venous endothelium after injury and repair has not been closely studied. A better understanding of this process could allow more judicious use of anticoagulation and the further development of surgical techniques applicable to venous repair. This investigation was specifically designed to characterize the reaction of venous endothelium to
repair.11 Although
·--
modern era of vascular surgery began in 1877 with successful anastomosis of two blood vessels.1 Ironically, this historic achievement involved venous surgery, which has subsequently engendered far less inter¬ est than arterial reconstruction. Continuing experience in the management of arterial trauma has clearly demonstrated the desirability of primary repair. Treatment of peripheral venous injuries is more controversial. Early investigators recommended liga¬ tion for venous injury, even suggesting that limb salvage could be improved in patients with concomitant arterial insufficiency.-*4 Although DeBakey and Simeone' refuted this concept in a careful analysis of venous injuries in World War II, concern remained over the potential for venous thrombosis and pulmonary embolus after venous
TheEck's
for publication July 19, 1979. From the Department of Surgery (Drs Krupski, Thai, Gewertz, and Fry) and Pathology (Drs Buja, Murphy, and Hagler), University of Texas Health Sciences Center, Dallas. Read before the 27th scientific meeting of the International Cardiovascular Society, Nashville, Tenn, June 28, 1979. Reprint requests to Department of Surgery, 5323 Harry Hines Blvd, Dallas, TX 75235 (Dr Thal).
Accepted
Fig 1.—Light micrograph of histologie section of vein stripped 48 hours before harvesting. Periadventitia contains damaged and collapsed arteriole of vasa vasorum (large arrowhead). Perladventitial and adventitial (A) tissue is disrupted and is infiltrated with fibrin and neutrophils (small arrowheads). Media (M) and intima are intact ( 120).
Downloaded From: http://archsurg.jamanetwork.com/ by a University of Arizona Health Sciences Library User on 06/02/2015
electron micrographs of stripped veins. Left, One hour after stripping, intimai surface shows multifocal loss of endothelial cells (areas indicated by asterisk) (X650). Right, Two weeks after stripping, intimai surface is completely lined by intact endothelial cells with well-defined intercellular junctions ( 650).
Fig 2.—Scanning
Fig 3.—Transmission electron micrograph of vein that was stripped four weeks before harvest Intimai surface is lined by endothelial cells (E) that are joined by intercellular junctions. Media contains smooth muscle cells (SMC) that are lined by basement membranes and have numerous cytoplasmic filaments, dense bodies, and pinocytotic vesicles. In other areas, intact venous media contains clusters of collagen fibrils as well as smooth muscle cells ( 10,000).
Downloaded From: http://archsurg.jamanetwork.com/ by a University of Arizona Health Sciences Library User on 06/02/2015
electron micrograph of vein one hour after clamping. Intimai surface is devoid of endothelium. Medial smooth muscle cells (SMC) show severe damage. Collagen (C) fibrils and elastic fiber (El) are also present; compare with Fig 3 and 6 ( 10,000).
Fig 4.—Transmission
surgical manipulation and the pattern ing after anastomosis.
of endothelial heal¬
measuring 1 to 1.5 cm was resected and anastomosis was performed under tension using interrupted (N 6) or continuous (N 6) technique. Other than the tension on the anastomosis, =
=
MATERIALS AND METHODS
Twenty-one mongrel dogs weighing from 15 to 30 kg were anesthetized with intravenous pentobarbital. Under sterile condi¬ tions, both external jugular and common femoral veins were carefully dissected. Systemic anticoagulation was not used. The dogs were divided into four groups. In group 1, left jugular and femoral veins (N 18) were =
mobilized over a distance of 5 to 6 cm. Care was taken to preserve the adventitial layer of the vein wall. Right jugular and femoral veins were mobilized over the same distance; all adventitia were meticulously stripped from these vessels. In group 2, veins were mobilized and occluded using a standard vascular clamp (N « 15) and Rommel-type tourniquet (N 9) on each specimen, isolating a segment of vein approximately 3 cm in length. Minimal occlusive force was used in applying clamps and tourniquets. Prior to occlusion, veins were flushed with a small amount of heparinized saline solution (10 units/mL of heparin sodium). After one hour, clamps and tourniquets were removed and the occlusion sites marked with fine vascular suture. In group 3, veins were transect¬ ed and underwent reanastomosis using interrupted (N 9) or continuous (N 9) technique. When using continuous technique, care was taken to place sutures loosely to avoid a "purse-string" effect. Number 6-0 polypropylene (Prolene) and No. 7-0 braided =
=
=
polyester (Ethibond) were randomly assigned to anastomoses. Vascular clamps were applied with minimal occlusive force on either side of the transected veins and dilute heparin solution was instilled proximally and distally. In group 4, a segment of vein
veins in group 4 were treated identically to those in group 3. All incisions were carefully closed and covered with collodion spray. At varying intervals (1 hour, 48 hours, 1-week, 2 weeks, and 4 weeks), the dogs were reanesthetized and vein segments were removed. Specimens were fixed in one of the following two ways: (1) ex vivo preparation in which removed vein segments were promptly opened, rinsed with tissue culture medium, pinned to cardboard under stretch, and immersed in aldehyde solution; or (2) in vivo preparation in which in situ vein segments were cannulated, rinsed with saline at approximately 15 mm Hg of pressure, instilled with aldehyde solution at the same pressure, and fixed in situ for two hours before removal. The ex vivo preparation was used for 15 dogs and the results compared with those obtained by in vivo preparation in six dogs. Fixation was performed using either a modified Karnovsky's solution (2.5% glutaraldehyde and 2% glutaraldehyde in 0.1M phosphate buffer, with a pH of 7.2 to 7.4), or 3% glutaraldehyde in 0.1M phosphate buffer (pH, 7.2 to
7.4). After gross examination, multiple specimens were obtained for microscopic study. Some specimens were submitted for paraffin embedding and preparation of sections stained with hematoxylineosin. Others were postfixed in 1% osmium tetroxide in 0.1M phosphate buffer (pH, 7.2 to 7.4), dehydrated through a graded series of alcohols and propylene oxide, and embedded in epoxy resin (Epon-Araldite). For light microscopy, l-µ thick epoxy sections were cut, stained with toluidine blue, and examined and photographed with a photomicroscope (Zeiss). For transmission
Downloaded From: http://archsurg.jamanetwork.com/ by a University of Arizona Health Sciences Library User on 06/02/2015
Fig 5.—Scanning electron micrograph of clamped veins. Left, One hour after clamping, endothelium has been lost from large area of intimai surface (indicated by *) ( 23). Center, Higher-magnification view shows areas devoid of endothelium (*) and patches of preserved endothelium (E) (x180). Right, Two weeks after clamping, intimai surface has intact endothelial lining ( 530).
Fig 6.—Transmission electron micrograph of vein one week after clamping. Intimai surface is lined by endothelium (E) and media has intact smooth muscle cells (SMC). Collagen fibrils are also present ( 10,000).
Downloaded From: http://archsurg.jamanetwork.com/ by a University of Arizona Health Sciences Library User on 06/02/2015
Fig 7.—Scanning electron micrograph of 48-hour-old anastomosis performed with interrupted suture. Left, Low-magnification view shows large gaps between sutures (S) that are focally connected with
strands of thrombus
(arrow). Sutures
are
also covered by thin layer of thrombus (x13). Right, material composed of fibrin and leukocytes partially is devoid of endothelium ( 300).
Higher-magnification view shows thrombotic bridging gap in anastomosis. Adjacent intima
Fig 8.—Scanning electron micrographs of anastomoses performed with interrupted suture technique. Left, One week after
surgery, surface of anastomosis has been extensively covered by endothelium, al¬ though gaps in anastomosis are still pres¬ ent (X150). Right, Two weeks after surgery, this particular anastomosis has become completely lined with endothe¬ lium (x54).
Downloaded From: http://archsurg.jamanetwork.com/ by a University of Arizona Health Sciences Library User on 06/02/2015
Fig 9.—Scanning electron micrographs of vein harvested four weeks after anastomosis was performed using interrupted suture technique. Left, anastomotic site is covered by moderate amount of thrombus (T) ( 41). Right, Higher magnification shows that surface of anastomosis is devoid of endothelium and consists of fibrin with trapped WBCs and RBCs ( 3,167).
Fig 10.—Scanning electron micrographs of anastomoses produced by continuous suture technique. Left, Forty-eight hours after surgery, vein has puckered appear¬ ance, and some thrombus (T) is present along anastomotic site ( 30). Right, Four weeks after surgery, vein shows massive thrombus ( ) ( 26).
Downloaded From: http://archsurg.jamanetwork.com/ by a University of Arizona Health Sciences Library User on 06/02/2015
electron microscopy, ultrathin sections were cut, mounted on copper grids, stained with uranyl acetate and lead citrate, and examined and photographed with a transmission electron micro¬ scope (JEOL 100 C). For scanning electron microscopy, specimens were dehydrated through a graded series of alcohols, critical point dried, mounted on stubs using a colloidal carbon paste, coated with 30 mm of gold palladium alloy using a sputter coater (Dentón), and examined and photomicrographed with a scanning electron
microscope.
RESULTS
Vein specimens fixed ex vivo showed changes identical with those prepared by the more complicated in vivo technique. In group 1, all veins in this group remained patent without gross or microscopic thrombus formation. Both mobilized and stripped veins showed acute disruption of periadventitial and adventitial connective tissue with damage of the vasa vasorum (Fig 1). The acute injury was followed by chronic inflammation and fibrosis in the periadventitia. Both mobilized and stripped veins demon¬ strated focal endothelial loss at one hour (Fig 2, left). Endothelial repair occurred in some veins as early as 48 hours, with all damage repaired by two weeks (Fig 2, right and 3). There was no evidence of subendothelial fibrodys¬ plasia or medial damage in either mobilized or stripped veins at any observation time. Although no intraluminal thrombus was noted in group 2, gross and microscopic examination of the occlusion sites for both vascular clamps and tourniquets showed periad¬ ventitial and adventitial hemorrhage and necrosis. Endo¬ thelial damage was uniformly observed at each occlusion site. The endothelial injury extended over a variable distance and was associated with disruption of medial smooth muscle cells (Fig 4 and 5). Endothelial regeneration at clamp sites began at 48 hours and was complete at two weeks (Fig 6). Aside from the initial disruption of medial elements, no persistent changes were present in the media or adventitia. Complete occlusion was found in only four of 18 speci¬ mens in group 3, resulting in an overall patency rate of 78%. Microscopic evidence of thrombus was uniformly present, characterized by fibrin deposition and trapped blood cells at anastomotic sites (Fig 7). Veins with minimal fibrin deposition demonstrated extensive reendothelialization by two to four weeks (Fig 8). Healing was associated with focal fibrosis and thickening of the wall adjacent to the anastomotic site. In veins with more extensive fibrin deposition or gross thrombus, the healing process, includ¬ ing the amount of repair and endothelial bridging of the defect, was noticeably delayed (Fig 9). Endothelial regen¬ eration remained incomplete for as long as four weeks in these specimens. Excessive fibrin deposition was not asso¬ ciated with a specific anastomotic technique or suture material (Fig 10). Although the occlusion rate was higher for continuous technique (4/9) than interrupted technique (0/9), the sample was too small for valid statistical analysis. Two of the four occluded anastomoses were performed with polypropylene suture and two were performed with braided polyester suture. There was no evidence of recanal¬ ization after occlusion in this series.
In group 4, anastomosis under tension resulted in complete occlusion in three of 12 specimens, resulting in an overall patency rate of 75%. An additional four specimens (33%) were narrowed at anastomotic sites to less than 50% of the original circumference. Histologie findings paral¬ leled those in group 3 with microscopic thrombi at anasto¬ moses, fibrin deposition, and slow endothelial bridging. The incidence of thrombosis in interrupted anastomoses (1/6) and continuous anastomoses (2/6) was similar. Type of suture material played no apparent role in thrombus formation. COMMENT
The increased use of peripheral veins as bypass conduits for aortocoronary, aortorenal, and other arterial bypasses has led to a wealth of data concerning the behavior of venous autografts in arterial systems.-'*-" After initial desquamation in the first week of graft placement, endo¬ thelium rapidly regenerates with half of the total graft surface covered by one month.2"' Despite this restoration of a smooth luminal surface, arterialized veins are plagued by subendothelial alterations, including cellular proliferation and collagen deposition in the intima, media, and adventitia.2T-" Mural edema, inflammatory changes, necrosis of myocytes, and medial and adventitial hemorrhagic foci are frequently noted. In addition to such primary derange¬ ments in the vascular wall, endothelial disruption and rupture of intracellular junctions in venous autografts results in a more favorable environment for deposition of fibrin and platelets. The cause of subendothelial changes remains controver¬ sial; forceful distention prior to placement, wall ischemia, and persistently increased intraluminal pressure have been investigated. The importance of mechanical factors is supported by the demonstration of necrotic smooth muscle cells in the tunica media shortly after graft placement. With time, these medial defects are filled with collagen. Several investigators have suggested that damage to the vasa vasorum during distention of venous autografts is an important mechanism. Subsequent impairment of nutrient blood flow through the media could contribute to subendo¬ thelial fibrodysplasia. Using a somewhat different model for interruption of the vasa vasora, Brody et al-1 found that "ischemie" vein segments demonstrated medial fibrosis whether inter¬ posed in a high-pressure (arterial) or low-pressure (venous) system. In addition, he described noticeable intimai prolif¬ eration and fibrosis, medial fibrosis, and loss of myocytes when "ischemie" vein segments were subjected to high intraluminal pressure. Although our investigation did not address the changes observed in arterialized vein segments, we observed no such alterations in intimai histology nor did we find any medial fibrosis in either stripped or mobilized veins. Since loss of vasa vasora was well documented histologically in both stripped and mobil¬ ized veins, it is suggested that at least in low-pressure systems, venous endothelium, subendothelial intima and media receive sufficient oxygénation and nourishment from luminal diffusion. Furthermore, the identical histo-
Downloaded From: http://archsurg.jamanetwork.com/ by a University of Arizona Health Sciences Library User on 06/02/2015
logic appearance of stripped and mobilized veins indicates that vasa vasora are damaged even when an attempt is
made to minimize such trauma. Another conclusion suggested by our data relates to endothelial damage caused by vascular clamps. Although it is universally accepted that anastomotic suture lines are thrombogenic, it is less well appreciated that clamp trauma potentiates the problem. Extensive endothelial loss at clamp sites with underlying hemorrhage and necrosis of myocytes predisposes to deposition of platelets and fibrin with subsequent luminal narrowing. Since clamps must be placed on either side of an anastomosis, a situation not unlike several subcriticai stenoses in series may develop. The extension of endothelial loss past the actual site of injury correlates with the findings of Schwartz et al1- in their study of arterial lesions, in which alterations in microscopic structure and function extended up to 100 cells beyond a point of endothelial injury. Although no readily apparent solution for this problem exists, knowledge of the damage produced by clamps should alert the vascular surgeon to use minimal occlusive force and avoid excessive traction on a vein when using clamps as handles. Unneces¬ sary handling of venous intima and extensive drying during exposure can further aggravate endothelial loss and should be minimized. Repair of anastomotic sites is characterized by early fibrin and platelet deposition followed by a varying degree of thrombus formation. An inverse relationship was noted between the extent of localized thrombus and endothelial bridging of the anastomotic defect. Specimens with mini¬ mal thrombus showed endothelial bridging at one week, and near complete reendothelialization by two to four weeks. Endothelial bridging was incomplete for as long as four weeks in those specimens with considerable fibrin deposition at anastomoses, regardless of surgical tech¬ nique. Benefits of anticoagulation after venous repair may be related to minimization of this fibrin sleeve. In this context, adequate anticoagulation could indirectly promote endothelial healing at venous anastomoses. Tension at anastomotic sites resulted in a histologie picture identical with anastomoses not done under tension, suggesting that the increased thrombosis rate associated with such constricted anastomoses is unrelated to the healing process per se and more probably results from hemodynamic abnormalities. The technique of venous anastomosis has been exten¬ sively reviewed in many articles and monographs."14 Although it is true that experienced vascular surgeons have reported acceptable patency rates using continuous technique, it is our belief that interrupted technique allows for a greater margin of error, particularly in less experi¬ enced hands. Although patency rate was not our primary concern in this investigation and statistical analysis showed no substantial differences, the patency rate for interrupted surgical technique was greater than the paten¬ cy rate for continuous technique (14/15 patent using interrupted technique vs 9/15 patent using continuous technique). Since operative time was similar for the two techniques, it would seem advisable for the surgeon who is
in venous repairs to use inter¬ This is probably most important in veins less than 5 mm in diameter. Venous anastomoses remain a challenge to the vascular surgeon. Low patency rates are influenced by the minimal pulsatility of venous flow and a tendency for in situ thrombosis, which is observed even in uninjured segments. Our investigation demonstrates that moderate amounts of fibrin deposition at anastomoses can retard endothelial healing and contribute to early failure. Although perioper¬ ative anticoagulation may inhibit thrombus formation, judicious placement of vascular clamps and careful handling of venous tissue will consistently minimize endo¬ thelial injury and reduce thrombogenicity. Attention to such technical detail would seem appropriate.
relatively inexperienced rupted suture technique.
H. Thomas Wheeler and Anna Siler
provided technical
assistance.
References 1. Guthrie CC: Blood Vessel Surgery and Its Applications. London, Edward Arnold, 1912. 2. Ney E: Du role des veines dans la circulation collaterale arterielle. Rev Chir 46:903-907, 1912. 3. Propping K: Ueber die ursache der gangrena noch unterbundung grosser arterien. MMW 64:598-599, 1917. 4. Makins GH: Gunshot Injuries to the Blood Vessels. Bristol, England, John Wright & Sons Ltd, 1919. 5. DeBakey ME, Simeone EA: Battle injuries of the arteries in World War 6. with 7.
II. Ann
Surg 123:534-579,
1946.
Cook FW, Haller JA Jr: Penetrating injuries of the subclavian vessels associated venous complications. Ann Surg 155:370-372, 1962. Rich NM, Hobson RW, Collins GJ, et al: The effect of acute popliteal venous interruption. Ann Surg 183:365-368, 1976. 8. Chandler JG, Knapp RW: Early definitive treatment of vascular injuries in the Vietnam conflict. JAMA 202:960-966, 1967. 9. Drapanas T, Hewitt RL, Weichert RF, et al: Civilian vascular injuries: A critical appraisal of three decades of management. Ann Surg 172:351-360,
1970. 10. Haimovici H, Hoffert PW, Zunicola N, et al: An experimental and clinical evaluation of grafts in the venous system. Surg Gynecol Obstet 131:1173-1186, 1970. 11. Hughes CW: Acute vascular trauma in Korean War casualties. Surg Gynecol Obstet 99:91-100, 1954. 12. Hughes CW: Arterial repair during the Korean War. Ann Surg 147:555-561, 1958. 13. Rich NM, Hughes CW, Baugh JH: Management of venous injuries. Ann Surg 171:724-730, 1970. 14. Rich NM, Hughes CW: Vietnam vascular registry: A preliminary report. Surgery 65:218-226, 1969. 15. Rich NM, Baugh JH, Hughes CW: Acute arterial injuries in Vietnam: 1,000 cases. Trauma 10:359-369, 1970. 16. Simeone FA, Grillo HC, Rendle F: On the question of ligation of the concomitant vein when a major artery is interrupted. Surgery 29:932-951, 1951. 17. Spencer FC, Grewe PV: The management of arterial injuries in battle casualties. Ann Surg 141:304-312, 1955. 18. Sullivan WG, Thornton FH, Baker LH, et al: Early influence of popliteal vein repair in the treatment of popliteal vessel injuries. Am J Surg 122:528-531, 1971. 19. DeWeese JA, Niguidula F: The replacement of short segments of veins with functional autogenous venous grafts. Surg Gynecol Obstet 110:303-308, 1960. 20. Dale WA: Thrombosis and recanalization of veins used as venous grafts. Angiology 12:603-306, 1961. 21. Hobson RW, Croom RD, Rich NM: Influence of heparin and low molecular weight dextran on the patency of autogenous vein grafts in the venous system. Ann Surg 178:773-776, 1973. 22. Ricotta JJ, Schaff HV, Gadacz TR: The effect of aspirin and dipyridamole on the patency of allograft veins. J Surgical Res 26:262-269, 1979. 23. Brody WR, Kosek JC, Angell WW: Changes in vein grafts following aortocoronary bypass induced by pressure and ischemia. J Thorac Cardiovasc Surg 64:847-854, 1972. J
Downloaded From: http://archsurg.jamanetwork.com/ by a University of Arizona Health Sciences Library User on 06/02/2015
24. McCabe M, Cunningham GK: Histological and histochemical examination of autogenous vein grafts. Br J Surg 54:147-155, 1967. 25. Wyatt AP, Taylor GW: Vein grafts: Changes in the endothelium of autogenous free vein grafts used as arterial replacements. Br J Surg 53:943-947, 1966. 26. Ramos JR, Benger K, Mansfield PB, et al: Histologic fate and endothelial changes of distended and undistended vein grafts. Ann Surg 183:205-228, 1976. 27. Urschel HC, Razzuk MA, Wood RF, et al: Factors influencing patency of aortocoronary artery saphenous vein grafts. Surgery 72:1048-1063, 1972. 28. Jones M, Conble DM, Ferrans VJ, et al: Lesions observed in arterial autogenous vein grafts: Light and electron microscopic evaluation. Circulation 47 & 48(suppl III):198-210, 1973. 29. Lie JT, Lourie GM, Morris GC Jr: Aortocoronary bypass saphenous vein graft atherosclerosis: Anatomic study of 99 vein grafts from normal and hyperlipoproteinemic patients up to 75 months postoperatively. Am J Cardiol 40:906-914, 1977. 30. Bulkley BH, Hutchins GM: Pathology of coronary artery bypass graft surgery. Arch Pathol Lab Med 102:273-280, 1978. 31. Karayannocos PE, Hostetler JP, Bond MG, et al: Late failure in vein grafts: Mediating factors in subendothelial fibromuscular hyperplasia. Ann Surg 187:183-188, 1978. 32. Schwartz SM, Haudenschild CC, Eddy EM: Endothelial regeneration: I. Quantitative analysis of initial stages of endothelial regeneration in rat aortic intima. Lab Invest 38:568-580, 1978. 33. Rich NM: Techniques of venous repair, in Rich NM, Hobson RW, Wright CB, et al (eds): Venous Surgery in the Lower Extremities. St Louis, Warren H. Green Inc, 1975, p 243. 34. Hobson RW, Wright CB, Swan KG, et al: Current status of venous injury and reconstruction in the lower extremity, in Bergan JJ, Yao JST (eds): Venous Problems. Chicago, Year Book Medical Publishers, 1978, p 469.
Discussion Makis Tsapogas, MD, MCh, Stony Brook, NY: The report by Dr Krupski and his colleagues is a significant study of a common problem responsible for postoperative complications. Some time ago, we conducted a similar study on the possibly damaging effect of various vascular clamps as compared to soft snares for occluding vessels. We found that the latter method was substantially kinder to the intima of the vessel. We also observed such effects of Fogarty catheters on the endothelium. Overdistention of the balloon was a significant factor contributing to damage of the intima. Effective thrombectomy can be achieved without having an excessively distended balloon. May I ask the authors if they compared the effects of vascular clamps to soft tube snares and also the effects of Fogarty catheters with the varying degree of inflation of the balloon. Peter B. Samuels, MD, Tarzana, Calif: I was very happy to see that in this electronic age Virchow's triad of blood flow, coagula¬ bility, and vessel wall are still intact and to see that the authors noted the extraordinary difficulty of producing a significant venous thrombosis in a healthy, intact vessel. In 1952 at McGill University, we conducted a whole-mount study using supravital staining in veins that have been damaged by mechanical, chemical, and thermal means, and we found that the endothelium was stained at the intercellular lines in a relatively uniform manner. With any form of injury there was a uniform response consisting of minimal injury in which bodies appear only on the intercellular cement lines. With more serious damage you get a spead of fibrin over the vessel wall, but still the spread of fibrin and the propagation of thrombosis is limited to the damaged vein wall and the surrounding area. With maximum trauma, such as the application of a clamp, we do get less of endothelium but strangely enough the subendothelial layer also shows this property of restricting the propagation of thrombosis to merely a covering layer under which repair, as Dr Krupski described it, will proceed. When we did the vein-stripping experiment, it followed the work of J. F. O'Neill who claimed that the vaso vasorum were
important to the maintenance of intact endothelium. We surrounded our veins for up to 72 hours with polyethylene sheathing and we found very small damage. I do not know the reason for the difference in our results from those in the present report. In our series of 150 veins, the only two gross thromboses we had with all types of injury followed the application of a constricting ligature downstream to the injury. Therefore, you have to have at least two of Virchow's triad fulfilled to produce thrombosis, and I think in this case that is provided by the constriction of the anastomosis plus the continuity of the fibrin overlying it. Howev¬ er, the fibrin deposit was referred to as interfering with endothe¬ lial healing. I would like to state that a fibrin tube is present after all vessel anastomoses and endothelial healing is allowed to proceed under cover of this very beneficial fibrin tube. Norman M. Rich, MD, Washington, DC: It was most encourag¬ ing to see the emphasis given to the management of venous problems in this year's program. There has been an increased interest in the management of venous trauma in the last ten years, and the authors, particularly Dr Krupski for his beautiful presentation, are to be congratulated for the valuable study that they have added to this area of interest. Fibrin deposition that they outlined in their study has added additional background and emphasis to the action of adjunctive measures that have been used from heparinization to aspirin to increase patency in venous repair. This will stimulate additional investigation in the future. I had the privilege of reading the manuscript, and the authors give additional information that will be of value to all regarding the action of veins in the arterial system as well as in the venous
system.
The question that I would ask the authors regards their lack of finding recanalization of thrombus in the venous system. It has
been documented repeatedly that the venous system differs from the arterial system in its rule for recanalization of thrombus. As far as technique is concerned, I was interested and concerned that they did not believe tension affected any of the repairs. I was pleased that the authors emphasized that meticulous techniques are
important in
venous
repair.
Robert L. Kistner, MD, Honolulu: I would like to point out that their results are on 21 mongrel dogs. This study contributes to our understanding of the process of injury and repair in the venous system. I certainly hope that it does not unduly increase our reluctance to operate on veins of humans. It has been demonstrated that we can operate successfully in the venous tree of humans with a minimal rate of thrombosis and with the use of a minimal degree of anticoagulation. In our personal experience with venous surgery, we have found more morbidity from excessive anticoagulation than we have from postoperative thrombosis. Dr Krupski: Dr Tsapogas, we did not use the "soft snare"; we used a standard umbilical tape to create our tourniquets. In answer to your question regarding Fogarty catheters, we did not employ these in our model. Dr Samuels, thank you for your kind response. It was similar research using supervital stains that led us to further investigate the nature of venous healing. Dr Rich, we are currently investigating the efficiency of anticoagulants in promoting the healing process of venous endo¬ thelium. We did not note recanalization in our veins, perhaps because some of our animals were killed at one week or 48 hours and recanalization generally does not occur until later. Regarding the effects of tension on anastomoses, we would like to reiterate that tension does not change the healing process. This does not in any way imply that excess tension is desirable in venous
anastomoses.
Downloaded From: http://archsurg.jamanetwork.com/ by a University of Arizona Health Sciences Library User on 06/02/2015
Response
to Venous
Krupski, MD; Erwin R. Thal, MD; Bruce L. Gewertz, MD; L. Murphy, PhD; Herbert K. Hagler, PhD; William J. Fry, MD
\s=b\This investigation characterized venous endothelial healing after surgical manipulation. Procedures were performed on jugular and femoral veins in 21 mongrel dogs without systemic anticoagulation. Veins were harvested at varying intervals and vessel structure evaluated with light, transmission, and scanning electron microscopy. Veins that were mobilized or stripped of adventitia demonstrated 25% to 50% endothelial loss at one hour. Endothelial damage was rapidly repaired with complete healing observed in some veins at 48 hours. Tourniquets and clamps resulted in prominent medial and endothelial injury at occlusion sites. Eighteen of 24 transected veins remained patent for the study period. Endothelial healing was unaffected by tension at anastomoses. These observations confirm that venous endothelium receives nutrition by luminal diffusion. The healing process of venous anastomoses is characterized by an early fibrin sleeve sealing the anastomotic site; endothelial bridging of defects can be noticeably delayed by excessive fibrin
deposition. (Arch Surg 114:1240-1248, 1979)
Maximilian
Injury
Buja, MD;
recent reports have shown that the incidence of thromboembolism is actually higher in patients whose venous injuries are treated by ligation, thrombosis after peripheral vein repair remains a frequent complication.'-" Early patency rates in canine models range from 55% to 100%; late patency rates are generally higher due to recanalization.1!'-- Demonstration of altered blood coagulation in postoperative patients has led some investigators to advocate the use of anticoagulant and antiplatelet agents after venous anastomosis.21 Although it is generally accepted that venous anasto¬ moses are thrombogenic, the natural healing process of venous endothelium after injury and repair has not been closely studied. A better understanding of this process could allow more judicious use of anticoagulation and the further development of surgical techniques applicable to venous repair. This investigation was specifically designed to characterize the reaction of venous endothelium to
repair.11 Although
·--
modern era of vascular surgery began in 1877 with successful anastomosis of two blood vessels.1 Ironically, this historic achievement involved venous surgery, which has subsequently engendered far less inter¬ est than arterial reconstruction. Continuing experience in the management of arterial trauma has clearly demonstrated the desirability of primary repair. Treatment of peripheral venous injuries is more controversial. Early investigators recommended liga¬ tion for venous injury, even suggesting that limb salvage could be improved in patients with concomitant arterial insufficiency.-*4 Although DeBakey and Simeone' refuted this concept in a careful analysis of venous injuries in World War II, concern remained over the potential for venous thrombosis and pulmonary embolus after venous
TheEck's
for publication July 19, 1979. From the Department of Surgery (Drs Krupski, Thai, Gewertz, and Fry) and Pathology (Drs Buja, Murphy, and Hagler), University of Texas Health Sciences Center, Dallas. Read before the 27th scientific meeting of the International Cardiovascular Society, Nashville, Tenn, June 28, 1979. Reprint requests to Department of Surgery, 5323 Harry Hines Blvd, Dallas, TX 75235 (Dr Thal).
Accepted
Fig 1.—Light micrograph of histologie section of vein stripped 48 hours before harvesting. Periadventitia contains damaged and collapsed arteriole of vasa vasorum (large arrowhead). Perladventitial and adventitial (A) tissue is disrupted and is infiltrated with fibrin and neutrophils (small arrowheads). Media (M) and intima are intact ( 120).
Downloaded From: http://archsurg.jamanetwork.com/ by a University of Arizona Health Sciences Library User on 06/02/2015
electron micrographs of stripped veins. Left, One hour after stripping, intimai surface shows multifocal loss of endothelial cells (areas indicated by asterisk) (X650). Right, Two weeks after stripping, intimai surface is completely lined by intact endothelial cells with well-defined intercellular junctions ( 650).
Fig 2.—Scanning
Fig 3.—Transmission electron micrograph of vein that was stripped four weeks before harvest Intimai surface is lined by endothelial cells (E) that are joined by intercellular junctions. Media contains smooth muscle cells (SMC) that are lined by basement membranes and have numerous cytoplasmic filaments, dense bodies, and pinocytotic vesicles. In other areas, intact venous media contains clusters of collagen fibrils as well as smooth muscle cells ( 10,000).
Downloaded From: http://archsurg.jamanetwork.com/ by a University of Arizona Health Sciences Library User on 06/02/2015
electron micrograph of vein one hour after clamping. Intimai surface is devoid of endothelium. Medial smooth muscle cells (SMC) show severe damage. Collagen (C) fibrils and elastic fiber (El) are also present; compare with Fig 3 and 6 ( 10,000).
Fig 4.—Transmission
surgical manipulation and the pattern ing after anastomosis.
of endothelial heal¬
measuring 1 to 1.5 cm was resected and anastomosis was performed under tension using interrupted (N 6) or continuous (N 6) technique. Other than the tension on the anastomosis, =
=
MATERIALS AND METHODS
Twenty-one mongrel dogs weighing from 15 to 30 kg were anesthetized with intravenous pentobarbital. Under sterile condi¬ tions, both external jugular and common femoral veins were carefully dissected. Systemic anticoagulation was not used. The dogs were divided into four groups. In group 1, left jugular and femoral veins (N 18) were =
mobilized over a distance of 5 to 6 cm. Care was taken to preserve the adventitial layer of the vein wall. Right jugular and femoral veins were mobilized over the same distance; all adventitia were meticulously stripped from these vessels. In group 2, veins were mobilized and occluded using a standard vascular clamp (N « 15) and Rommel-type tourniquet (N 9) on each specimen, isolating a segment of vein approximately 3 cm in length. Minimal occlusive force was used in applying clamps and tourniquets. Prior to occlusion, veins were flushed with a small amount of heparinized saline solution (10 units/mL of heparin sodium). After one hour, clamps and tourniquets were removed and the occlusion sites marked with fine vascular suture. In group 3, veins were transect¬ ed and underwent reanastomosis using interrupted (N 9) or continuous (N 9) technique. When using continuous technique, care was taken to place sutures loosely to avoid a "purse-string" effect. Number 6-0 polypropylene (Prolene) and No. 7-0 braided =
=
=
polyester (Ethibond) were randomly assigned to anastomoses. Vascular clamps were applied with minimal occlusive force on either side of the transected veins and dilute heparin solution was instilled proximally and distally. In group 4, a segment of vein
veins in group 4 were treated identically to those in group 3. All incisions were carefully closed and covered with collodion spray. At varying intervals (1 hour, 48 hours, 1-week, 2 weeks, and 4 weeks), the dogs were reanesthetized and vein segments were removed. Specimens were fixed in one of the following two ways: (1) ex vivo preparation in which removed vein segments were promptly opened, rinsed with tissue culture medium, pinned to cardboard under stretch, and immersed in aldehyde solution; or (2) in vivo preparation in which in situ vein segments were cannulated, rinsed with saline at approximately 15 mm Hg of pressure, instilled with aldehyde solution at the same pressure, and fixed in situ for two hours before removal. The ex vivo preparation was used for 15 dogs and the results compared with those obtained by in vivo preparation in six dogs. Fixation was performed using either a modified Karnovsky's solution (2.5% glutaraldehyde and 2% glutaraldehyde in 0.1M phosphate buffer, with a pH of 7.2 to 7.4), or 3% glutaraldehyde in 0.1M phosphate buffer (pH, 7.2 to
7.4). After gross examination, multiple specimens were obtained for microscopic study. Some specimens were submitted for paraffin embedding and preparation of sections stained with hematoxylineosin. Others were postfixed in 1% osmium tetroxide in 0.1M phosphate buffer (pH, 7.2 to 7.4), dehydrated through a graded series of alcohols and propylene oxide, and embedded in epoxy resin (Epon-Araldite). For light microscopy, l-µ thick epoxy sections were cut, stained with toluidine blue, and examined and photographed with a photomicroscope (Zeiss). For transmission
Downloaded From: http://archsurg.jamanetwork.com/ by a University of Arizona Health Sciences Library User on 06/02/2015
Fig 5.—Scanning electron micrograph of clamped veins. Left, One hour after clamping, endothelium has been lost from large area of intimai surface (indicated by *) ( 23). Center, Higher-magnification view shows areas devoid of endothelium (*) and patches of preserved endothelium (E) (x180). Right, Two weeks after clamping, intimai surface has intact endothelial lining ( 530).
Fig 6.—Transmission electron micrograph of vein one week after clamping. Intimai surface is lined by endothelium (E) and media has intact smooth muscle cells (SMC). Collagen fibrils are also present ( 10,000).
Downloaded From: http://archsurg.jamanetwork.com/ by a University of Arizona Health Sciences Library User on 06/02/2015
Fig 7.—Scanning electron micrograph of 48-hour-old anastomosis performed with interrupted suture. Left, Low-magnification view shows large gaps between sutures (S) that are focally connected with
strands of thrombus
(arrow). Sutures
are
also covered by thin layer of thrombus (x13). Right, material composed of fibrin and leukocytes partially is devoid of endothelium ( 300).
Higher-magnification view shows thrombotic bridging gap in anastomosis. Adjacent intima
Fig 8.—Scanning electron micrographs of anastomoses performed with interrupted suture technique. Left, One week after
surgery, surface of anastomosis has been extensively covered by endothelium, al¬ though gaps in anastomosis are still pres¬ ent (X150). Right, Two weeks after surgery, this particular anastomosis has become completely lined with endothe¬ lium (x54).
Downloaded From: http://archsurg.jamanetwork.com/ by a University of Arizona Health Sciences Library User on 06/02/2015
Fig 9.—Scanning electron micrographs of vein harvested four weeks after anastomosis was performed using interrupted suture technique. Left, anastomotic site is covered by moderate amount of thrombus (T) ( 41). Right, Higher magnification shows that surface of anastomosis is devoid of endothelium and consists of fibrin with trapped WBCs and RBCs ( 3,167).
Fig 10.—Scanning electron micrographs of anastomoses produced by continuous suture technique. Left, Forty-eight hours after surgery, vein has puckered appear¬ ance, and some thrombus (T) is present along anastomotic site ( 30). Right, Four weeks after surgery, vein shows massive thrombus ( ) ( 26).
Downloaded From: http://archsurg.jamanetwork.com/ by a University of Arizona Health Sciences Library User on 06/02/2015
electron microscopy, ultrathin sections were cut, mounted on copper grids, stained with uranyl acetate and lead citrate, and examined and photographed with a transmission electron micro¬ scope (JEOL 100 C). For scanning electron microscopy, specimens were dehydrated through a graded series of alcohols, critical point dried, mounted on stubs using a colloidal carbon paste, coated with 30 mm of gold palladium alloy using a sputter coater (Dentón), and examined and photomicrographed with a scanning electron
microscope.
RESULTS
Vein specimens fixed ex vivo showed changes identical with those prepared by the more complicated in vivo technique. In group 1, all veins in this group remained patent without gross or microscopic thrombus formation. Both mobilized and stripped veins showed acute disruption of periadventitial and adventitial connective tissue with damage of the vasa vasorum (Fig 1). The acute injury was followed by chronic inflammation and fibrosis in the periadventitia. Both mobilized and stripped veins demon¬ strated focal endothelial loss at one hour (Fig 2, left). Endothelial repair occurred in some veins as early as 48 hours, with all damage repaired by two weeks (Fig 2, right and 3). There was no evidence of subendothelial fibrodys¬ plasia or medial damage in either mobilized or stripped veins at any observation time. Although no intraluminal thrombus was noted in group 2, gross and microscopic examination of the occlusion sites for both vascular clamps and tourniquets showed periad¬ ventitial and adventitial hemorrhage and necrosis. Endo¬ thelial damage was uniformly observed at each occlusion site. The endothelial injury extended over a variable distance and was associated with disruption of medial smooth muscle cells (Fig 4 and 5). Endothelial regeneration at clamp sites began at 48 hours and was complete at two weeks (Fig 6). Aside from the initial disruption of medial elements, no persistent changes were present in the media or adventitia. Complete occlusion was found in only four of 18 speci¬ mens in group 3, resulting in an overall patency rate of 78%. Microscopic evidence of thrombus was uniformly present, characterized by fibrin deposition and trapped blood cells at anastomotic sites (Fig 7). Veins with minimal fibrin deposition demonstrated extensive reendothelialization by two to four weeks (Fig 8). Healing was associated with focal fibrosis and thickening of the wall adjacent to the anastomotic site. In veins with more extensive fibrin deposition or gross thrombus, the healing process, includ¬ ing the amount of repair and endothelial bridging of the defect, was noticeably delayed (Fig 9). Endothelial regen¬ eration remained incomplete for as long as four weeks in these specimens. Excessive fibrin deposition was not asso¬ ciated with a specific anastomotic technique or suture material (Fig 10). Although the occlusion rate was higher for continuous technique (4/9) than interrupted technique (0/9), the sample was too small for valid statistical analysis. Two of the four occluded anastomoses were performed with polypropylene suture and two were performed with braided polyester suture. There was no evidence of recanal¬ ization after occlusion in this series.
In group 4, anastomosis under tension resulted in complete occlusion in three of 12 specimens, resulting in an overall patency rate of 75%. An additional four specimens (33%) were narrowed at anastomotic sites to less than 50% of the original circumference. Histologie findings paral¬ leled those in group 3 with microscopic thrombi at anasto¬ moses, fibrin deposition, and slow endothelial bridging. The incidence of thrombosis in interrupted anastomoses (1/6) and continuous anastomoses (2/6) was similar. Type of suture material played no apparent role in thrombus formation. COMMENT
The increased use of peripheral veins as bypass conduits for aortocoronary, aortorenal, and other arterial bypasses has led to a wealth of data concerning the behavior of venous autografts in arterial systems.-'*-" After initial desquamation in the first week of graft placement, endo¬ thelium rapidly regenerates with half of the total graft surface covered by one month.2"' Despite this restoration of a smooth luminal surface, arterialized veins are plagued by subendothelial alterations, including cellular proliferation and collagen deposition in the intima, media, and adventitia.2T-" Mural edema, inflammatory changes, necrosis of myocytes, and medial and adventitial hemorrhagic foci are frequently noted. In addition to such primary derange¬ ments in the vascular wall, endothelial disruption and rupture of intracellular junctions in venous autografts results in a more favorable environment for deposition of fibrin and platelets. The cause of subendothelial changes remains controver¬ sial; forceful distention prior to placement, wall ischemia, and persistently increased intraluminal pressure have been investigated. The importance of mechanical factors is supported by the demonstration of necrotic smooth muscle cells in the tunica media shortly after graft placement. With time, these medial defects are filled with collagen. Several investigators have suggested that damage to the vasa vasorum during distention of venous autografts is an important mechanism. Subsequent impairment of nutrient blood flow through the media could contribute to subendo¬ thelial fibrodysplasia. Using a somewhat different model for interruption of the vasa vasora, Brody et al-1 found that "ischemie" vein segments demonstrated medial fibrosis whether inter¬ posed in a high-pressure (arterial) or low-pressure (venous) system. In addition, he described noticeable intimai prolif¬ eration and fibrosis, medial fibrosis, and loss of myocytes when "ischemie" vein segments were subjected to high intraluminal pressure. Although our investigation did not address the changes observed in arterialized vein segments, we observed no such alterations in intimai histology nor did we find any medial fibrosis in either stripped or mobilized veins. Since loss of vasa vasora was well documented histologically in both stripped and mobil¬ ized veins, it is suggested that at least in low-pressure systems, venous endothelium, subendothelial intima and media receive sufficient oxygénation and nourishment from luminal diffusion. Furthermore, the identical histo-
Downloaded From: http://archsurg.jamanetwork.com/ by a University of Arizona Health Sciences Library User on 06/02/2015
logic appearance of stripped and mobilized veins indicates that vasa vasora are damaged even when an attempt is
made to minimize such trauma. Another conclusion suggested by our data relates to endothelial damage caused by vascular clamps. Although it is universally accepted that anastomotic suture lines are thrombogenic, it is less well appreciated that clamp trauma potentiates the problem. Extensive endothelial loss at clamp sites with underlying hemorrhage and necrosis of myocytes predisposes to deposition of platelets and fibrin with subsequent luminal narrowing. Since clamps must be placed on either side of an anastomosis, a situation not unlike several subcriticai stenoses in series may develop. The extension of endothelial loss past the actual site of injury correlates with the findings of Schwartz et al1- in their study of arterial lesions, in which alterations in microscopic structure and function extended up to 100 cells beyond a point of endothelial injury. Although no readily apparent solution for this problem exists, knowledge of the damage produced by clamps should alert the vascular surgeon to use minimal occlusive force and avoid excessive traction on a vein when using clamps as handles. Unneces¬ sary handling of venous intima and extensive drying during exposure can further aggravate endothelial loss and should be minimized. Repair of anastomotic sites is characterized by early fibrin and platelet deposition followed by a varying degree of thrombus formation. An inverse relationship was noted between the extent of localized thrombus and endothelial bridging of the anastomotic defect. Specimens with mini¬ mal thrombus showed endothelial bridging at one week, and near complete reendothelialization by two to four weeks. Endothelial bridging was incomplete for as long as four weeks in those specimens with considerable fibrin deposition at anastomoses, regardless of surgical tech¬ nique. Benefits of anticoagulation after venous repair may be related to minimization of this fibrin sleeve. In this context, adequate anticoagulation could indirectly promote endothelial healing at venous anastomoses. Tension at anastomotic sites resulted in a histologie picture identical with anastomoses not done under tension, suggesting that the increased thrombosis rate associated with such constricted anastomoses is unrelated to the healing process per se and more probably results from hemodynamic abnormalities. The technique of venous anastomosis has been exten¬ sively reviewed in many articles and monographs."14 Although it is true that experienced vascular surgeons have reported acceptable patency rates using continuous technique, it is our belief that interrupted technique allows for a greater margin of error, particularly in less experi¬ enced hands. Although patency rate was not our primary concern in this investigation and statistical analysis showed no substantial differences, the patency rate for interrupted surgical technique was greater than the paten¬ cy rate for continuous technique (14/15 patent using interrupted technique vs 9/15 patent using continuous technique). Since operative time was similar for the two techniques, it would seem advisable for the surgeon who is
in venous repairs to use inter¬ This is probably most important in veins less than 5 mm in diameter. Venous anastomoses remain a challenge to the vascular surgeon. Low patency rates are influenced by the minimal pulsatility of venous flow and a tendency for in situ thrombosis, which is observed even in uninjured segments. Our investigation demonstrates that moderate amounts of fibrin deposition at anastomoses can retard endothelial healing and contribute to early failure. Although perioper¬ ative anticoagulation may inhibit thrombus formation, judicious placement of vascular clamps and careful handling of venous tissue will consistently minimize endo¬ thelial injury and reduce thrombogenicity. Attention to such technical detail would seem appropriate.
relatively inexperienced rupted suture technique.
H. Thomas Wheeler and Anna Siler
provided technical
assistance.
References 1. Guthrie CC: Blood Vessel Surgery and Its Applications. London, Edward Arnold, 1912. 2. Ney E: Du role des veines dans la circulation collaterale arterielle. Rev Chir 46:903-907, 1912. 3. Propping K: Ueber die ursache der gangrena noch unterbundung grosser arterien. MMW 64:598-599, 1917. 4. Makins GH: Gunshot Injuries to the Blood Vessels. Bristol, England, John Wright & Sons Ltd, 1919. 5. DeBakey ME, Simeone EA: Battle injuries of the arteries in World War 6. with 7.
II. Ann
Surg 123:534-579,
1946.
Cook FW, Haller JA Jr: Penetrating injuries of the subclavian vessels associated venous complications. Ann Surg 155:370-372, 1962. Rich NM, Hobson RW, Collins GJ, et al: The effect of acute popliteal venous interruption. Ann Surg 183:365-368, 1976. 8. Chandler JG, Knapp RW: Early definitive treatment of vascular injuries in the Vietnam conflict. JAMA 202:960-966, 1967. 9. Drapanas T, Hewitt RL, Weichert RF, et al: Civilian vascular injuries: A critical appraisal of three decades of management. Ann Surg 172:351-360,
1970. 10. Haimovici H, Hoffert PW, Zunicola N, et al: An experimental and clinical evaluation of grafts in the venous system. Surg Gynecol Obstet 131:1173-1186, 1970. 11. Hughes CW: Acute vascular trauma in Korean War casualties. Surg Gynecol Obstet 99:91-100, 1954. 12. Hughes CW: Arterial repair during the Korean War. Ann Surg 147:555-561, 1958. 13. Rich NM, Hughes CW, Baugh JH: Management of venous injuries. Ann Surg 171:724-730, 1970. 14. Rich NM, Hughes CW: Vietnam vascular registry: A preliminary report. Surgery 65:218-226, 1969. 15. Rich NM, Baugh JH, Hughes CW: Acute arterial injuries in Vietnam: 1,000 cases. Trauma 10:359-369, 1970. 16. Simeone FA, Grillo HC, Rendle F: On the question of ligation of the concomitant vein when a major artery is interrupted. Surgery 29:932-951, 1951. 17. Spencer FC, Grewe PV: The management of arterial injuries in battle casualties. Ann Surg 141:304-312, 1955. 18. Sullivan WG, Thornton FH, Baker LH, et al: Early influence of popliteal vein repair in the treatment of popliteal vessel injuries. Am J Surg 122:528-531, 1971. 19. DeWeese JA, Niguidula F: The replacement of short segments of veins with functional autogenous venous grafts. Surg Gynecol Obstet 110:303-308, 1960. 20. Dale WA: Thrombosis and recanalization of veins used as venous grafts. Angiology 12:603-306, 1961. 21. Hobson RW, Croom RD, Rich NM: Influence of heparin and low molecular weight dextran on the patency of autogenous vein grafts in the venous system. Ann Surg 178:773-776, 1973. 22. Ricotta JJ, Schaff HV, Gadacz TR: The effect of aspirin and dipyridamole on the patency of allograft veins. J Surgical Res 26:262-269, 1979. 23. Brody WR, Kosek JC, Angell WW: Changes in vein grafts following aortocoronary bypass induced by pressure and ischemia. J Thorac Cardiovasc Surg 64:847-854, 1972. J
Downloaded From: http://archsurg.jamanetwork.com/ by a University of Arizona Health Sciences Library User on 06/02/2015
24. McCabe M, Cunningham GK: Histological and histochemical examination of autogenous vein grafts. Br J Surg 54:147-155, 1967. 25. Wyatt AP, Taylor GW: Vein grafts: Changes in the endothelium of autogenous free vein grafts used as arterial replacements. Br J Surg 53:943-947, 1966. 26. Ramos JR, Benger K, Mansfield PB, et al: Histologic fate and endothelial changes of distended and undistended vein grafts. Ann Surg 183:205-228, 1976. 27. Urschel HC, Razzuk MA, Wood RF, et al: Factors influencing patency of aortocoronary artery saphenous vein grafts. Surgery 72:1048-1063, 1972. 28. Jones M, Conble DM, Ferrans VJ, et al: Lesions observed in arterial autogenous vein grafts: Light and electron microscopic evaluation. Circulation 47 & 48(suppl III):198-210, 1973. 29. Lie JT, Lourie GM, Morris GC Jr: Aortocoronary bypass saphenous vein graft atherosclerosis: Anatomic study of 99 vein grafts from normal and hyperlipoproteinemic patients up to 75 months postoperatively. Am J Cardiol 40:906-914, 1977. 30. Bulkley BH, Hutchins GM: Pathology of coronary artery bypass graft surgery. Arch Pathol Lab Med 102:273-280, 1978. 31. Karayannocos PE, Hostetler JP, Bond MG, et al: Late failure in vein grafts: Mediating factors in subendothelial fibromuscular hyperplasia. Ann Surg 187:183-188, 1978. 32. Schwartz SM, Haudenschild CC, Eddy EM: Endothelial regeneration: I. Quantitative analysis of initial stages of endothelial regeneration in rat aortic intima. Lab Invest 38:568-580, 1978. 33. Rich NM: Techniques of venous repair, in Rich NM, Hobson RW, Wright CB, et al (eds): Venous Surgery in the Lower Extremities. St Louis, Warren H. Green Inc, 1975, p 243. 34. Hobson RW, Wright CB, Swan KG, et al: Current status of venous injury and reconstruction in the lower extremity, in Bergan JJ, Yao JST (eds): Venous Problems. Chicago, Year Book Medical Publishers, 1978, p 469.
Discussion Makis Tsapogas, MD, MCh, Stony Brook, NY: The report by Dr Krupski and his colleagues is a significant study of a common problem responsible for postoperative complications. Some time ago, we conducted a similar study on the possibly damaging effect of various vascular clamps as compared to soft snares for occluding vessels. We found that the latter method was substantially kinder to the intima of the vessel. We also observed such effects of Fogarty catheters on the endothelium. Overdistention of the balloon was a significant factor contributing to damage of the intima. Effective thrombectomy can be achieved without having an excessively distended balloon. May I ask the authors if they compared the effects of vascular clamps to soft tube snares and also the effects of Fogarty catheters with the varying degree of inflation of the balloon. Peter B. Samuels, MD, Tarzana, Calif: I was very happy to see that in this electronic age Virchow's triad of blood flow, coagula¬ bility, and vessel wall are still intact and to see that the authors noted the extraordinary difficulty of producing a significant venous thrombosis in a healthy, intact vessel. In 1952 at McGill University, we conducted a whole-mount study using supravital staining in veins that have been damaged by mechanical, chemical, and thermal means, and we found that the endothelium was stained at the intercellular lines in a relatively uniform manner. With any form of injury there was a uniform response consisting of minimal injury in which bodies appear only on the intercellular cement lines. With more serious damage you get a spead of fibrin over the vessel wall, but still the spread of fibrin and the propagation of thrombosis is limited to the damaged vein wall and the surrounding area. With maximum trauma, such as the application of a clamp, we do get less of endothelium but strangely enough the subendothelial layer also shows this property of restricting the propagation of thrombosis to merely a covering layer under which repair, as Dr Krupski described it, will proceed. When we did the vein-stripping experiment, it followed the work of J. F. O'Neill who claimed that the vaso vasorum were
important to the maintenance of intact endothelium. We surrounded our veins for up to 72 hours with polyethylene sheathing and we found very small damage. I do not know the reason for the difference in our results from those in the present report. In our series of 150 veins, the only two gross thromboses we had with all types of injury followed the application of a constricting ligature downstream to the injury. Therefore, you have to have at least two of Virchow's triad fulfilled to produce thrombosis, and I think in this case that is provided by the constriction of the anastomosis plus the continuity of the fibrin overlying it. Howev¬ er, the fibrin deposit was referred to as interfering with endothe¬ lial healing. I would like to state that a fibrin tube is present after all vessel anastomoses and endothelial healing is allowed to proceed under cover of this very beneficial fibrin tube. Norman M. Rich, MD, Washington, DC: It was most encourag¬ ing to see the emphasis given to the management of venous problems in this year's program. There has been an increased interest in the management of venous trauma in the last ten years, and the authors, particularly Dr Krupski for his beautiful presentation, are to be congratulated for the valuable study that they have added to this area of interest. Fibrin deposition that they outlined in their study has added additional background and emphasis to the action of adjunctive measures that have been used from heparinization to aspirin to increase patency in venous repair. This will stimulate additional investigation in the future. I had the privilege of reading the manuscript, and the authors give additional information that will be of value to all regarding the action of veins in the arterial system as well as in the venous
system.
The question that I would ask the authors regards their lack of finding recanalization of thrombus in the venous system. It has
been documented repeatedly that the venous system differs from the arterial system in its rule for recanalization of thrombus. As far as technique is concerned, I was interested and concerned that they did not believe tension affected any of the repairs. I was pleased that the authors emphasized that meticulous techniques are
important in
venous
repair.
Robert L. Kistner, MD, Honolulu: I would like to point out that their results are on 21 mongrel dogs. This study contributes to our understanding of the process of injury and repair in the venous system. I certainly hope that it does not unduly increase our reluctance to operate on veins of humans. It has been demonstrated that we can operate successfully in the venous tree of humans with a minimal rate of thrombosis and with the use of a minimal degree of anticoagulation. In our personal experience with venous surgery, we have found more morbidity from excessive anticoagulation than we have from postoperative thrombosis. Dr Krupski: Dr Tsapogas, we did not use the "soft snare"; we used a standard umbilical tape to create our tourniquets. In answer to your question regarding Fogarty catheters, we did not employ these in our model. Dr Samuels, thank you for your kind response. It was similar research using supervital stains that led us to further investigate the nature of venous healing. Dr Rich, we are currently investigating the efficiency of anticoagulants in promoting the healing process of venous endo¬ thelium. We did not note recanalization in our veins, perhaps because some of our animals were killed at one week or 48 hours and recanalization generally does not occur until later. Regarding the effects of tension on anastomoses, we would like to reiterate that tension does not change the healing process. This does not in any way imply that excess tension is desirable in venous
anastomoses.
Downloaded From: http://archsurg.jamanetwork.com/ by a University of Arizona Health Sciences Library User on 06/02/2015
