Br. J. Surg. 1980, Vol. 77, September, 1055-1088, P. D. Coleridge-Smith, J. H. Hasty and J. H. Scurr Department of Surgery, University College and Middlesex Schoo! of Medicine, UK Correspondence to: Mr J. H. Scurr, Department of Surgery, University College and Middlesex School of Medicine, Mostimer Streot, London WIN BAA, UK Venous stasis and vein lumen changes during surgery The mechanisms underlying the development of postoperative deep vein thrombosis remain: to be fully elucidated. Previous studies have suggested that peroperative venous distension may be a factor associated with venous thromboembolism. In this study we have obtained high resolution ultrasound images of gastrocnemius and posterior tibial veins in 62 patients undergoing a range of general surgical procedures. From these we determined the changes in vein diameter occurring during the ‘operative procedures, in response to induction of anaesthesia, and afier completion of surgery. Veins showed no evidence of dilatation in response 10 the induction of anaesthesia, but by the end of the operative procedure showed distension of 22-28 per cent. Distension was most prominent in the gastrocnemius veins, and lesser distension was observed in the posterior tibial veins. In a series of patients who received an infusion of 1 litre of saline in addition to basal requirements, distension was 57 per cent compared with 22 per cent in the corresponding control group. Intraoperative venous distension is associated with factors that lead to deep vein thrombosis and may be involved in the mechanisms which result in the commencement of deep vein thrombosis. Keywords: Deep vein thrombosis, pulmonary embolism, venous distension Deep vein thrombosis and subsequent pulmonary embolism remain a significant hazard to hospital patients. In a recent study this pathological process was responsible for 10 per cent of deaths of patients who died in hospital". Postoperative surgical patients accounted for 24 per cent’ of this series, emphasizing the continuing threat that deep vein thrombosis presents, The factors that may predispose patients to risk and the precautions that might prevent this problem are Well knowin? Although much research has beea done into the pathogenesis of this problem, a complete understanding has yet to be achieved. Virchow described the basic aetiology of venous thrombo- embolism more than a century ago’. Recent progress in many aspects of measurement inthis disease has confirmed that these original observations were largely corcect. Clinical studies have demonstrated which patients are at risk of deep. vein thrombosis®®. These advanees in basi science and technology haved to the development ofa number of methods anc clinical regimens for the prevention of deep vein thrombosis and pulmonary embolism. These incorporate both pharmacological ‘and physical approaches intended to reduce the coagulability of blood and ensure that venous stasis is avoided*, Present prophylactic techniques are based on the model of deep vein thrombosis described in Virchow's iad. ‘The role. of anticoagulants in altering the blood coagulation cascade has been studied in detail, and the action of physical agencies in reducing venous stasis has also received attention in scientific investigations over the past four decades”, The importance of the third component of Virchow's triad, vascular endothelial damage, has not been fully investigated and its role in the development of postoperative deep vein thrombosis is not understood Recently Comerota eral. have showa that venous distention ‘ccurs in upper limb veins during the course of surgical procedures, They proposed that this phenomenon may be significant factor in causing vein wall injury and therefore the initiation of deep venous thrombosis in surgical patients®, They have subsequently shown that patients in whom there is distension in the cephalic vein in excess of 20 per cent are at increased risk of postoperative radiofibrinogca-detectable deep vein thrombosis™®. Support for the role of distension in the (007=132/90)091055-05 © 1990 Butirworth-Heinemaan Lid pathogenesis of deep vein thrombosis was found by Scurr et al, in a clinical study of deep vein thrombosis in surgical patients!'. The incidence of postoperative deep vein thrombosis was compared in patients receiving intermittent pneumatic compression alone or in combination with graduated ‘compression stockings to prevent thrombosis. The results clearly demonstrated a superior prophylactic effect using both ‘compression methods simultaneously in preventing deep vein thrombosis. It was suggested that vascular endothelial damage, the thitd component of Virchow’s triad, produced by venous distension was prevented by graduated compression stockings but not by intermittent pneumatic compression alone. The aim of this study was to determine whether the deep veins of the lower extremities dilate during the course of surgery under general anaesthesia, Patients and methods Patients undergoing elective surgery at The Middlesex Hospital were centered into this study. Surgical procedures included head and neck, abdominal, and vascular operations, representing @ wide range of surgical practice (Table 1). Vascular procedutes included both Abdominal and lower limb arterial resonstnzctions ae well as varicose vein procedures, Only the unoperated lim was included in the analysis for venous distension, to ensure that the operative manipulation had ‘20 effect on tho limb examined. Patients gave thei informed conser for inclusion in the study. ‘The technique used to measure changes In vein size was high resolution duplex ultrasound imaging. An Actson 128 (Acuson Table 1 Patient population (al subjects) First study Second study Mean age (years) S137) 47 2-79) No. of patents in surgical group Head and nec. 5 10 Vascular 2 4 Abdominal 6 7 Duration of anaesthesia in min 38 (14122) 48 (16-183) ‘Ranges are given in parentheves 1055,Venous distension during surgery: P. D. Coleridge-Smith et al Figure 1 Sconning teclovique: high resolution aliases imaging was used to measure the diameters of the lover limb veins The lind wat supported by an assistant while images ofthe gastrocremius vets were obtained srough this medial clf approach Incorporated, Mountsin View, Calor, USA) equipped with a 5 Mlle inca arcay transducer was used fo all measucements (Figures | tnd 2). Longitudinal sans of the legs were obtained because it was found that ‘these data were the most satfactory for reliable iterprtation. Tt wes found tht several vein groups could be studied sing this technique, and small changes in vein dlameter could be Selected, Both lover limbs were studied unlest the surgical proosdure prevented access (0 the limb. Images were recorded on videotape for Subsequent analysis of vein diameter. The lapes were replayed into an image processing computer (Torch Computers Lid, Great Shelford, Canioridge, UK). A program written specially for this purpose by ‘one ofthe authors (P-D.C-S.] was used fo measure te vein chameters, Taking into account variations inthe sealing of the image. Each vein was measured thre times at San interval fo enable a mean diameter (o be calculated. The image processing computer stored these data for subsequent statistical analysis “The study was conducted in halves. In the Grst part patints were scanned and gastcoenemins vein diameters ware measved in the ‘medial head of gastrocnemius, immediately below the level of the bial tuberosity. Vein diameters were measured on three occasions: (1) on the day before surgsry, usualy the day of admission; (2) after induction of anaesthesia, bet before the start ofthe eurpical procedure; and (3) at the end of the surgical procedure, but before Tecovery from the anaesthetic, withthe patient on the operating table These intervals were selected to permit an assessment of the effet ‘ofanaesthesia onthe vein to be tested. The frst obrervation wise made withthe patient conscious, the last exo withthe patient anaesthetized Vein diameters were measured withthe patient supine and the (im horizontal and ‘with ‘the limb exed at right angles, achieved. by removing the end of the operating tale while an assistant supported the limts. Patients received peroperative wansTesions of intravenous Auids to met thei basal cequirements. No specific ecocd of these was kept, although all anaesthesia was supervised by one anaesthetist and conducted using a standard techni Ta the second pact ofthe study the folowing groups of veins were measured: (I) posterior tibial vin, immediately below the level of the Uibial tuberosity: (2) gastrocnemius vein in the medial head of sastrocnemius at the same level; and (3) posterior tibial vein, atthe midpoint of the cal 1056 Measurements of vein diameter were made on two occasions: (1) ater induction of anaesthesia, but before the start of the surgical procedure; and (2) at the end of the surgical procedure, but before ‘eeavery rom the anaesthetic, with the patient on the operating cable In the second patt ofthis study patents were randomly allocated to recive either no additional therapy or intravenous infusion of 10G0 ml 69 per cent saline in addition fo basal requirements during the operative procedure. In patiens in whom an intravenous line was insected to permit acess forthe administration of anzesthetic agents 1000 mi 09 per cont saline was infused during the course of the surgical procedure. In patents undergoing procedure where Aud infasion wae Considered necessary of volume replacement was required, 1000 mi 09 per cent saline was infused In addition to the ealewated basal and Transfusion requiernens. The purpose of this was to investigate the effect of crystalloid infusion, a fetor that may ineresee the risk of deep vein thrombosis. All pabenis recived appropriate prophylactic ‘teaiment (0 prevent deep vein thrombosis, where indicate. Low-dose fuboutaneoos heparin was ‘ured Tor’ thi, avoiding compression treatments that might influence venovs distension, For each measurement, ultasourel images ofeach se of veins were recorded with the patent Iying supine on the examination bed of Figure 2 Longitudinal viow of gastrocneinius veins obtained by iratoond iaging Inthe epi postion a the start a and end b ofthe ‘operation. An Increase inthe vein diameter can be seen conspared with lhe preoperative views Br. J. Surg., Vol. 77, No. 8, September 1990P<0.02 —— P< 0.01 “ll s : as Before operation P< 0.001 | Mean vein diameter (cm) Induction After operation Figure 3. First rudy’ mean vein dlomeers of gastrocnemius befor, a the start of, and after operation, wh the Limb supine () and perived to hang dependent (C). Thee isa significant increage in diameter af thesexeins ony atthe and ofthe operative procedaren.s.,notsighiican! 0 oa fore operation Induction After operation Increase In vein lameter (8) Figure Percentage inerate in vin dhameer compared withthe supine position before operation in the same patents at tn Figure 3; My, Lind Supine, Cl lind dependent opeccting table. Measurements were then repeated with the knee Nosed at right angles Tn the operating theatre this was achieved by removing the ond of the operating table while the limbs ware supported by an fsistant Each limb was then examined in torn, rst inthe supine position, then with the fm permitted to Dee at the knee while the ther was maintained supine. The period required to collect the data from eich limb for supine and dependent examinations wat Approximately 1-2 min, 35 indicated by the time information recorded fon videocasiete with the ullessound images. The techaigue. of permiting the limb to hang dependest was wed as a means of ralsing, the venous pressure in the leg by 2 controlled amount, but avoiding the use of a tourniquet. The intention was to assess the distensibi of the wins by this manoeuvre, and was not in any way intended 10 simulate the normal events during surgery. Since the peciod of Gepondency war brit it was assumed that this would aot aes the subsequent measurement of ela duimeterat the end athe operation, Statistical analyse of data was undertaken using Student’ ts Peized tests were uted for comparison of measurements trom the same Tim a cileren times, Two-sample tests were employed to compare vals from diferent patient groups. Pearson's correlation coefficient (was calculated to assess correlation between variables The use of statistical tate that assume a Gaussian ditribution of samples was sified by caleulating the skewaess and kurtosis of the orginal data ‘There was no substantial deviation from Gaussian dieebation in any of the data analysed, Calculations were made with the assistance of The computer software SPSSPC, SPSS Incorporated, Chicago, Minos, USA} running on 4 mierosompiter Br. J. Surg., Val. 77, No. 9, September 1980 Venous distension during surgery: P. D. Coleridge-Smith ot al Results In the fist series of observations (Figure 3) made only on the gastrocnemius veins, there was no change in the vein diameter immediately after the induction of anaesthesia. The veins retained their ability to distend on dependency and the diameter with the limbs supine was unchanged. It was only at the end ‘of the operative procedure that a statistically significant increase in diameter was observed in the supine position. There was no significant further increase in the diameter of these veins in response to moving to the dependent position, suggesting that they had already become maximally dilated (Figure 4) In the subsequent studies where two sections of the postecior tibial vein were observed in addition to the gastrocnemius veins, the gastrocnemius veins were significantly smaller in diameter than the posterior tibial veins (Figure 5). ‘The gastrocnemius veins participate inthe calf muscle pump and might be expected to be smaller than the major axial limb vessels, These veins showed a much greater ability to distend in response to a rise in venous pressure than the posterior tibial veins, and this may be in keeping with their pump function, On moving (0 the ependent position the gastrocnemius veins distended by 37 per cent (P-<005) compared with 8 per cént (n.s.) for the tmidcalf posterior tibial veins. During the operative procedure the mean distension observed in the gastrocnemius veins was 22 pot cent (P<005) but vein diameter did not change significantly in either of the posterior tibial measurements. In the group of pationts who received intraoperative saline transfusion the distension was much greater than in the corresponding control group (Figure 6). At the end of the operative procedure there was a 48 per cent distension of the gastrocnemius veins in the supine position compared with the preoperative measurement (P =:0003). This effect was confined to the gastrocnemius veins, and did not reach statistical significance in the posterior tibial veins. The response of the sastrocnemius veins in the control and saline transfusion groups 's compared in Figure 7, There was no correlation between the time taken for the surgical procedure and the fractional increase in diameter of the gastrocnemius veins. This did not appear to be a factor in ‘causing augmentation of the distension process in the patients studied. Discussion TThe problem of stasis in the deep veins ofthe lower extremities ‘uring surgical procedures undertaken with genecal anaesthesia [i PEE EEEEEEEeE EEE tino? Supine Dependent” ~ Supine Dependent pesition Start of operation End of operation igure $. Second study: venous diameter and response to inereased enous pressure in the three ovine where measweneents were mae. Pots are from the control group; BA, gastrocnemius; 8, upper posterior tibial veins: midelf posterior tbial veins: ns. significant 10879 during surgery: P. D. Coleridge-Smith et al i 17-0908 CT Set Z z Z E 4g bs wy z Z| Z i Al : Aliw § Zig sos Yi : yi i yi vi Line?" Supina Dependant Supine Dapencont positon Start of operation End of operation Figure 6 Second study: the effect of saline transfusion during surgery, comparing the effect onthe thee sets of veins; My, gatrocnerian, upper posterior bil eins: Cl, midelf posterior tibial eins: ns not significant hnas been documented in several well controled independent investigations!*-". These experiments have demonstrated the slowing of venous blood flow during anaesthesia, showing both ‘decrease in blood flow velacity and anincreasein the residence time of blood in the deep veins ofthe call. Ths is especially pronounced in the sinuses of the venous valves. A number of factors may result in the stagnation of venous blood in the lower extremities during surgery. The loss of normal muscle tone and hence venous pump function during the course of anaesthesia may result in substantial alterations in the usual flow patterns in the veins of the lower limb, particularly i muscle tone is farther reduced by neuromuscular blocking agents!*. Virchow’s triad suggests that stasis in these veins is likely to result in deep ein thrombosis “The present study provides direct evidence that, concomitant with intraoperative blood flow changes shown by other investigators, there are significant increases in the diameter of the deep eins, We propose that these may lead to the potential for endothelial damage, a further factor suggested by Virehow et al. and consistent with recent clinical and ultrastructucal Studies!#*- OF the vessels examined, the gastrocnemius vein ‘appears to undergo the greatest dilatation fom induction of anaesthesia to end of operation, However, additional work with the improved scanning capabilites of more recent duplex ultrasound equipment which provides better image resalution is warranted before final comparisons of venous distension at Gifferent anatomical sites can be made. ‘Our findings support the observation of Tripolitis et al. that ‘enous capacitance changes inthe surgical patient”. Our data show that as the venous diameter increases during. the operation, the ability to increase futher in response to a rise in venous pressure i reduced “The effect of saline transfusion on venous distension followed the same trond as in eoatrols, but produced a significantly sgreater increase in mean vein diameter than in the control group by the end of the operation. This observation suggests that fluid adminstrtion may be one of the intraoperative factors tat resulted in the observed inerease in vein diameter in the first study. The relationship between this elect and thrombotic events is not clear from these data, although an association between fluid administration and postoperative deep vein thrombosis has been observed previously". tn adcition, the roles of graduated elastic compression and intermiitent pneumatic compression in providing prophylaxis against deep 1088 Pz0.019 039 @ my : i 1 Supine Dependent Supine Dependent Besition Start of operation End of operation igure 7 Second study: response of gatrocnemis veins inthe conirol ‘and saline tranefusion groupe compared cn the samie graph. Data are Jom the same patiens as 42 Figutes Sand 6: Mi control: Ch saline ‘ansfision: me not significant vein thrombosis are more clearly understood. These findings emphasize the importance of avoiding compression profiles that ‘would tend to trap venous blood in the distal veins when using compression to prevent deep vein thrombosis, This might promote the venous distension that has been observed in this study. Careful application of antiembolism stockings or the use of sequential graduated intermittent pneumatic compression is indicated asthe latter has been demonstrated not to trap venous blood distally"™", This study provides diroct anatomical evidence demons: trating that events during an operative procedure results in venous distension, in addition to the slowing of flow that previous authors have shown, We suggest that this may Contribute to the endothelial damage part of Virchow's triad, promoting the formation of deep vein thrombosis. Acknowledgem« ‘This work was supported in part by Kendall Healthcare. Products Company, Manstcid, Massachusetts, USA, and the Special Trustees ‘of The Middlesex Hospital. References 1. Sandler DA, Martin JF. Autopsy proven pulmonary embolism in hoeptal’ patients: are we detecting enough deep. vein ‘hrombosis? J Soe Med 1989; 82: 202-5. 2, Nicolaides AN, Irving D. Clinical actors and the risk of deep sein thromboris, In: Nicolaides AN, ed. Thromboembolin: Aetiology, Advances in Prevention and Monagarient. Lancaster ‘MTP Lid, 1975: 193-203, 3. Janssen HE, Schachaer J, Hubbard J, Hartman JT. The risk of ‘ep venous thrombosis:< compateriznd epidemiologic approach Surgery 1987; 101: 205-11 4. Coldit GA, Tuden RL, Oster'G. Rates of venous thrombosis, after venous surgery: combined results of randomised clinical Lancer 1986; fe 13-6 5. Virchow R. Gesammelte Abandlungen cur Wissenschaptichen ‘Medizin. Frankfurt-am-Main: von Meilingee Goln, 1856 6, Nationa Insitute of Heath Consensus Development Conference Prevention of venous thrombosis and pulmonary embolism, YAMA 1986: 256: 744-8, 7. Kakkar VV. The cutest status of low-dose heparin inthe prophylaxis ‘of thrombophlebitis and. pulmonary embolism World J Surg 1978; 2: 3-18 8, Caprini JA, Scurr 3H, Hasty JH, Role of compression modalities in & prophylactic program for deep vein thrombosis. Semit ‘Diramb Homost 1968; 1(Suppl): 77-87 9. Comerots Al, Stewart GI, White IV. Combined ditydeo- crgotamine and heparin prophylaxis of postoperative deep vein tirombosis: propesed mechanism of ation. drm J Surg 1985; 50 3-4, Br. J. Surg. Vol. 77, No. 8, September 1990410, Comerota AJ, Stewart GI, Alburgee PD, Smalley K, White JV, Operative venodilation: «previously unsuspected Tetor in the {2Use of postoperative deep vein thrombosis. Surgery 1987; 106: 301-9, 11. Scuer JH, Coleridge Smith PD, Hasty JH, Regimen for improved clletiveness of intermittent pneumatic compression in. dasp ‘yenous thrombosis prophylaxs. Surgery 1987; 102: 816-20, 12. Janwin SB, Davies G, Greentalgh RM. Postoperative deep vein thrombosis caused by intravenous fuids during eurgery. Br J ‘Surg 1980; 6: 600-3. 13. Lewis CE, Antoine J, Mueller C, Talbot WA, Swaroop R, Eawards WS. Elastic compression in the prevention of venous stasis. Ane J Surg 1976; 138: 139-43. 14. Nicolaides AN. Venous stasis inthe lower limb. In: Nicolaides AN, ed. Thromboembotion: Aetiology, Aiances in Prevention ‘and Management, Lancaster: MTP Lid, 1975; (93-204 15, Doran FSA, Drury M, Sivyer A.A simple way to combat the venous stasis which oceurs in the lower limbs duting surgical Br. J. Surg.. Vol. 77, No. 8, September 1990 Venous distension during surgery: P. D. Coleridge-Smith et al 16, cperations. BJ Surg 1964; SU: 486. Schaub RG, Lynch PR, Stevart G3. The response of enaine veins ta thee types of abdominal surgery: scanning and transmission slectron microscopic study. Surgery 1978; &3: 41-24 ‘AI, Bodily KG, Blackshear WM et ai, Venous 1 and outflow in the postoperative patient. Aan Sirg 1979; 190: 643, ‘Kamm R, Butcher R, Froelich Jet of. Optimisation of indies cof external pneumatic compression for prophylais against deep vein thrombosis: radionuclide gated imaging studies. Card! Reo 1986; 20: 588-96, Nicoidides AN, Feenandes JF, Pollock AY. Iniermitent sequential pacurnatic compression in the legs in the preveation Df venous stasis and postoperative deep venous thrombosis, ‘Surgery 1983; 7; 69-16 Paper accepted 9 April 1990 1089