Eur J Vasc Endovasc Surg 11, 12-18 (1996)

Angioscopy for Quality Control of Saphenous During Bypass Grafting*

Y. G. Wilson 1, A. H. Davies 1, I. C. Currie ~, C. McGrath ~, M. Morgan 1, E. Sheffield 2, R. N. Baird and P. M. Lamont

1Vascular Studies Unit and 2Department of Histopathology, Bristol Royal Infirmary, Bristol, U.K.

Objectives: Although autogenous vein is the conduit of choice for infrainguinal bypass grafting, some 20-30% of vein grafts fail during the first year postoperatively. Many of these failed are now known to have pre-existing pathological changes. Angioscopy enables intraoperative endoluminal visualisation of veins and can reveal anomalies, some previously unsuspected, despite preoperative Duplex ultrasound mapping and normal external appearances. The aim of this study was to compare angioscopic findings with contemporary histological appearances and with subsequent graft outcome and ultimately, to identify those endoluminaI features which might be predictive for failure. Methods: Angioscopic vein inspection was carried out using Olympus 1.4 and 2.2mm angioscopes in patients undergoing femoropopliteal/distal bypass. Severe disease in the veins of five patients led to preferential use of polytetrafluoroethylene (PTFE) for above-knee bypasses. The remaining 38 videotaped sequences were reviewed by two surgeons and scored using a scale of 0 to 3, based on frequency and distribution of angioscopically detected lesions. These included haemorrhagic mural plaques, flimsy intraluminal strands, webs~bands and mobile~adherent thrombus. Vein harvested at operation was assessed by a pathologist according to the level of pre-existing abnormality. Results: There were significant associations between angioscopy/histology scores and graft survival (Z2 = 22.00; dr:3; p < 0.001; X2 = 22.43; dr:3; p < 0.001 respectively). There was a significant correlation between angioscopy and histology scores (R8 = 0.725; p < 0.001). Conclusions: Angioscopy allows immediate identification of the at risk, poor quality vein graft at the time of , without the delays inherent with histological preparation and assessment. Recognition of abnormalities at angioscopy may ultimately improve graft outcome by prospectively eliminating use of poor vein.

Introduction existing intrinsic features may predispose the graft to further intimal hyperplasia. Much has been published on the relative merits of Whilst harvesting techniques 1° and the adaptive autogenous vein and prosthetic materials as conduits responses of veins to arterialisation are relevant to for vascular reconstruction. In the context of infra- success, vein quality is such a vital commodity that the inguinal revascularisation, the main body of evidence proposed graft should be assessed as completely as supports saphenous vein as the conduit of choice, possible prior to completion of the bypass. Duplex even for above-knee grafting. ~-3 However, 20-30% of scanning has superceded saphenous for vein grafts fail during the first 12 months post- pre-operative evaluation of vein and allows a predic- operatively. Inadequate patient selection, sub-optimal tion of size, compliance 9 and anatomy. 11'12 For the vein and technical errors underlie the majority of assessment of vein quality intraoperatively, comple- failures within the first month. 4 Thereafter, neointimal tion flow studies ~3 and completion arteriography 14 hyperplasia is the principal risk, causing progressive have hitherto been the mainstays. With the advent of stenoses. 5 The precise aetiology of intimal hyperplasia angioscopy, the lumen and endoluminal surface can remains uncertain, but vein quality is emerging as a be directly visualised in a search for unsuspected, pre- potent determinant of outcome 6-9 and some pre- existing disease. 15 Angioscopy and Duplex scanning are therefore complementary in the overall assessment *Presented at the 8th annual meeting of the European Society for of quality. Vascular Surger~ Berlin, Germany (September 1994). This prospective study was designed to compare Please address all correspondence to: Miss Y.G. Wilson, Research Fellow, Vascular Studies Unit, Level 7, Bristol Royal Infirmary, angioscopic findings with histological appearances Bristol BS2 8HW, U.K. and graft outcome, with the ultimate aim of identify-

1078-5884/96/010012 + 07 $12.00/0 © 1996 W. B. Saunders Company Ltd. Angioscopy and Quality Control 13 ing those endoluminal features which are predictive of completion studies. Whenever possible, saphenous graft failure. vein was harvested for histological examination. In practise, a 2cm length of vein was harvested for histology and usuall)~ this was taken from the distal portion of the vein at the level of transection appro- Patients and Methods priate to the operation and in excess of requirements for the bypass. Fifty-two patients (14 female, 38 male; median age: 70 Postoperatively, all patients entered a Duplex graft years [range : 39-93 years]; 30% diabetic) undergoing surveillance programme, involving scans and Doppler femoropopliteal or femorodistal bypass were recruited ankle pressures at 1 week, 6 weeks and 3, 6, 9 and 12 into a prospective study to evaluate the influence of months postoperatively. Duplex-detected stenoses vein quality on graft patency. Thirty bypasses (58.0%) were considered significant (50% or greater) if the were for critical ischaemia and 19 (37.0%) for severe, ratio of the peak systolic velocity within 2cm of the limiting claudication (walking distance less than 100 stenosis and the peak systolic velocity within the yards). Three patients (5.8%) underwent bypass and narrowed segment was greater than 2.0.17'18 Patients ligation of expanding popliteal aneurysms. were followed up for 12 months postoperatively. Local Research Ethics Committee approval was Definite end points within this period were graft obtained. Experience with angioscopy had been occlusion, intervention for stenosis and death. gained during the 12 months prior to commencement Videotaped sequences of vein inspections during of this study. 16 This represented our all-important angioscopic graft preparation and angioscopic com- learning curve and for the present stud)~ sufficient pletion studies were viewed independently by two experience had been gained to enable meaningful surgeons (YGW and AHD), blinded to patient identify recognition and interpretation of abnormal endolumi- and graft outcome. Lesions observed included flat, nal features at angioscopy. haemorrhagic mural plaques, flimsy fibrinous strands, All patients underwent preoperative Duplex vein organised luminal webs/bands and either free-float- mapping in the Vascular Studies Unit using an ing or adherent thrombi (Fig. 1). Veins were graded on Ultramark 9 HDI scanner (Advanced Technology a scale of 0 to 3 on the basis of lesions observed and Laboratories Ltd., Letchworth, U.K.). Intraoperativel~ their frequency and distribution (Table 1). The pres- the in situ technique of vein grafting was used for ence of diffuse, persistent spasm converted the grade bypass. Nineteen bypasses were onto the above-knee which would have been assigned on the basis of popliteal and 33 were infrageniculate (20 luminal morphology alone to the grade above. below-knee popliteal and 13 tibial). For angioscopy, Vein sent fresh for light microscopy was perfusion Olympus@ 1.4mm and 2.2ram flexible scopes were fixed, serially sectioned and stained with haematox- used, in conjunction with a full Olympus@ angioscopy ylin and eosin and van Gieson stains. Histological stack system, incorporating monitor, light source, grading of pre-existing pathology was carried out by camera and irrigation pump (Keymed Ltd., Essex, an experienced Consultant Histopathologist (ES), also U.K.). Plain Hartmann's solution was used for angio- blinded as to patient identity and outcome. Veins were scopic irrigation in all cases. Twenty-one patients again graded on a scale of 0 to 3 according to the underwent angioscopic vein preparation, with lysis of criteria set out in Table 2. The presence of muscle valves under direct vision, using a modified Mill's hypertrophy caused the grade which would have been valvulotome and identification of tributaries from assigned on the basis of pre-existing intimal hyper- within for their ligation through multiple stab inci- plasia alone to be converted to the grade above. The sions. Patients undergoing traditional vein prepara- current system is a modification of that previously tion underwent blind valvulotomy with' Hall's valvu- devised by Milroy et al. 19 lotomes and ligation of tributaries via a full length incision. All patients underwent angioscopic comple- tion studies of the graft and its anastomoses, as well as haemodynamic flow measurements (Op Dop 130 -- Scimed, Bristol U.K.) and on-table arteriography. On- Data analysis table arteriography was carried out using 15-20 ml of Ultravist 300 (Schering Health Care Ltd., West Sussex, Associations between outcome and angioscopy/his- U.K.), injected via the side-arm of the valved 6 or 8 Fr tology scores were analysed using the chi-square test. introduction catheters used for angioscopy and The correlation between angioscopy and histology inserted into a proximal tributary left long for the scores in those patients who had both parameters

Eur J Vasc EndovascSurg Vol 11, January 1996 14 Y.G. Wilson et aL

Fig. 1.

Fig. 2.

Eur J Vasc Endovasc Surg Vol 11, January 1996 Angioscopy and Quality Control 15

Table 1. Criteria for angioscopic grading system mended by the Ad Hoc Committee on Reporting Grade Definition* Standards. z° 0 Normal vein with clean endoluminal surface, absence of luminal lesions/ debris. Readily distensible vein wall with angifiscopicirrigation. Isolated focal lesions i.e. haemorrhagic plaques/fibrinous Results strands/webs/thrombus. Or Between February 1993 and June 1994, 52 patients Grade 0 + diffuse spasm. were recruited to the study. This cohort included Multiple focal lesions. Or patients with videotaped sequences of angioscopy Grade I + diffuse spasm. and/or patients with adequate histological sections. Diffuse lesions extending throughout the vein segment. Of these, 38 patients had suitable videotape footage of Or angioscopic inspection and/or completion studies Grade 2 + diffuse spasm. available for review. Forty-three patients had histo- *See also Fig. 1. logical sections of harvested vein which were graded. Twenty-nine patients had both angioscopy video Table 2. Criteria for histological grading system material and histolob~. The median follow-up availa- Grade Definition* ble for this series was 6 months (range: 2 weeks to 12 months). 0 Normal vein with no pre-existingfibro-elastic intimal A further 25 patients undergoing femoropopliteal/ hyperplasia (IH). femorodistal bypass during this time were excluded No muscle hypertrophy. due to instrument and/or operator non-availability Mild focal pre-existingIH, up to 50% of the luminal surface. with respect to angioscop~ particularly for urgent and Or emergency cases and insufficient vein for histology. Grade 0 IH + muscle hypertrophy. An additional five patients undergoing above-knee Moderate diffuse pre-existingIH, greater than 50% of the popliteal bypasses had such severe and widespread, luminal surface. Or pre-existing, intraluminal pathology on angioscopic Grade 1 IH + muscle hypertrophy. inspection that vein was abandoned in favour of Extensive diffuse pre-existingIH, involving the entire PTFE. The disease was unsuspected after Duplex vein luminal surface with musde hypertrophy. mapping and after external inspection during surgery. Or Grade 2 IH + muscle hypertrophy. Fig. 3 shows an angioscopic view of dense, organised webs (representative of the disease observed), and Fig. *See also Fig. 2. 4 represents the equivalent histological section. The angioscopic grade was agreed by both observ- available was analysed using the Spearman rank test. ers in 93% of cases. There was disparity in three Sensitivities, specificities, positive predicitve values patients (by one grade in each case) and a final grade (PPV) and negative predicitive values (NPV) were was agreed by consensus. Not all angioscopic abnor- calculated. Cumulative patency and life-table data malities featured in the scoring system. For example, were analysed according to the methods recom- seven patients each had a retained valve cusp after Fig. 1. Representative series of angioscopic images depicting a blind valvulotomy with the Hall's valvulotome. How- selection of venous anomalies (Olympus 1.4mm angioscope). Top ever, the retained cusps were detected and lysed row (left to right): (a) Flims~ fibrinous, intraluminal strand with during completion angioscopy. It is interesting to note tributary orifice at 3 o'clock; (b) Multiple, fibrous, intraluminal webs; (c) Complex and dense fibrous bands, virtually obscuring that complementary on-table arteriography failed to lumen and trapping thrombotic debris. Middle row (left to right): (a) detect all these retained cusps. Flat, mural, haemorrhagic 'plaque'. This may represent sub- There may be concern that some of the observed endothelial haemorrhage from handling during dissection before insertion of the angioscope. (b) Fine, thrombotic, intraluminal findings are artefactual, that is, created by the passage strand; (c) Multiplet fine, thrombotic, intraluminal strands. Bottom of the angioscope itself. Whilst it has to be conceded row (left to right): (a) Fragment of thrombus, precariously attached that endoluminal manipulation and irrigation could to the vein wall at one point and flapping in the stream of irrigation fluid; (b) Large piece of mobile thrombus, virtually occluding damage the vein, the angioscopes used were smaller lumen; (c) Diffuse, intractable spasm despite distension with than the valvulotomes in current practice and most of irrigation fluid. the abnormal features observed were noted on the first Fig. 2. Photographs illustrating the four histological grades (Elastic passage of the angioscope. Deterioration in vein van Gieson stains; x 40). (a) Grade 1 (top left), (b) Grade 2 (top right), (c) Grade 3 (bottom left), (d) Grade 4 (bottom right). qualit~ by the criteria defined, did not occur by the Reproduced here at 45%. end of the angioscopic instrumentation.

Eur J Vasc Endovasc Surg Vol 11, January 1996 16 Y.G. Wilson et aL

Fig. 4. Histological section of intrahiminal webs, associated with valve remnants. (Elastic van Gieson stain; × 40). Reproduced here at 80%. Table 3. The association between angioscopy grade and graft outcome. (X2 test compares 'failed' grafts (stenoses and occlusions) Fig. 3. Angioscopic image of fibrous webs. (The presence of this sort with normal grafts) of lesion throughout the veins of five patients led to preferential use 0 1 2 3 of PTFE). Occlusion <30 days 0 0 3 8 Stenosis/occlusion >30 days 0 1 3 1 Patent graft 7 9 6 0

X2 = 22.00; df : 3; p < 0.001.

Table 4. The association between histology grade and graft outcome. (X2 test compares 'failed' grafts (stenoses and occlusions) with normal grafts) 0 1 2 3

Occlusion <30 days 0 0 4 5 Stenosis/occlusion >30 days 0 0 " 6 3 Patent graft 4 13 7 1 X2 = 22.42; df : 3; p < 0.001.

and 100%. There was a significant correlation between angioscopy and histology grades for the 29 patients with both parameters available (Spearman's rank correlation coefficient R8 = 0.725; p < 0.001). Fig. s. "Double lumen" phenomenon indicative of varicosity. (This In this cohort of 52 patients, the cumulative primary appearance is often seen when approaching a discrete varicosity. and secondary patencies at 12 months were 57% and The varicosity itself billows out between the two luminal edges under the pressure of the irrigatior~ fluid). 80% respectively (log rank test : p < 0.05). This compares favourably with previously published patency data from this Unit for an earlier series of Tables 3 and 4 show the associations between graft peripheral graft patients. 18 The 30 day failure rate in outcome and angioscopy/histology scores. There this cohort was 19.2% as against 12.0% for the earlier were significant correlations between outcome and series. 21 angioscopy scores 0(2 = 22.00; df:3; p < 0.001) and between outcome and histological scores 0(2 = 22.42; df: 3; p < 0.001). Taking angioscopy scores of 2 and 3 as representative of poor quality vein, the respective Discussion sensitivities, specificities, PPV's and NPV's of angio- scopy in relation to graft outcome were: 93%, 73%, These results demonstrate that angioscopy is a useful 70% and 94%. The equivalent values for histology technique for intraoperative quality control in the scores with respect to outcome were : 100%, 68%, 69% selection of vein prior to bypass. There is a need for a

Eur J Vasc Endovasc Surg Vol 11, January 1996 Angioscopy and Quality Control 17 complementary modality to supplement existing the presence of intraluminal webs/bands, thought to screening methods. Angioscopy fulfils this role and represent sites of incomplete recanalisation, following despite being invasive, there is no evidence from the thrombophlebitis. 34 In a subsequent stud~ the same published series that it adversely influences graft group reported on the high incidence of intraluminal patency.~2"~3 This is the first prospective study to disease in arm veins and found that angioscopy was compare angioscopic assessment of saphenous vein not only a sensitive technique for detecting disease, quality with graft outcome. but by monitoring endoluminal and surgical inter- The concept of vein quality was first advocated by ventions to upgrade the quality of the vein, it also Logerfo et al. 6 and Szilagyi et al. 7 as a highly significant contributed to improved early graft patency. 35 determinant of graft patency. However, vein quality The prospective application of angioscopy for detec- came to be regarded by some as synonymous with tion of occult disease and for selection of the optimum size. There are many opinions about the optimum conduit for bypass is not new. is Potentially, patency graft diameter for vein bypass. 24 Moreover, size is just rates could be enhanced by discarding poor quality one facet of the morphology and function of vein and vein or allowing upgrade of mediocre vein. Sales et al. with respect to the long saphenous vein, a spectrum of carried out angioscopic examination of saphenous anatomical variation and pathological change has vein remnants following bypass surgery and by been described. ~5'25'26 Recently, these intrinsic changes comparison with histological features, found that 27% have been linked to graft patency.8'27'28 The apprecia- of veins were abnormal by angioscopic criteria. 15 The tion of pre-existing pathological changes justifies same group described angioscopic appearances attrib- efforts for detailed assessment of the conduit prior to utable to vein wall calcification, but admitted a bypass. difficulty with detection of varicosities and thick- Preoperative, non-invasive, Duplex scanning is well walled veins. In the authors' experience, calcification established for mapping the course of the saphenous has not been an obvious feature, although varicosities vein and gives an estimate of internal diameter, whilst can be recognised by the "double lumen" phenome- detecting varicosities and sites of major tributaries. 11"12 non (Fig. 5) and a subjective impression of wall However, the outer extent of the vein wall merges thickness/compliance can be gained by observing ultrasonically with interstitial tissue, making it diffi- distensibility with irrigation at different flow rates. cult to measure wall thickness accurately. Moreover, Angioscopy has been slow to gain recognition36 and Duplex failed to detect the intraluminal defects later has been treated with some scepticism. Angioscopic identified at angioscopy, including gross disease in the assessment is qualitative and interpretation is sub- five patients whose vein was rejected because the jective, but, this new application attempts to provide a grafts would have failed. Panetta et aI. also showed that Duplex is not infallible; preoperative scanning semi-quantitative dimension and may help the tech- failed to detect 38% of abnormal veins in their nology to realise its full potential and achieve more series. 8 widespread acceptance. This study has demonstrated Visual inspection, palpation and arteriography have that poor quality vein can be identified by the all been used for the intraoperative evaluation of the angioscope and has confirmed the poor outcome conduit. 8 Measurement and observation of arterial associated with such veins. On this basis, angioscopy flow in the open graft, before completion of the distal has a role in quality control for assessing the potential anastomosis, may also provide clinically reassuring vein conduit prior to its use in either lower limb evidence of the functional adequacy of the conduit. ~1 revascularisation or coronary artery surgery. Further However, these methods are limited for the detection prospective studies are needed to clarify whether of subtle intraluminal changes which might otherwise using angioscopy for vein selection does result in go undetected. The diagnostic ability of angioscopy improved graft patency rates. during inspection of grafts, anastomoses and native is well established 29'3° and its superiority over arteriography has been confirmed in two recent prospective trials. 3L32 Neville et al. evaluated Acknowledgements angioscopy against arteriography and for the management of intimal flaps in a We are grateful to KeymedLtd. for their technicalsupport during canine model. ~3 Both endovascular techniques proved this study and for financial assistance with the costs of colour superior to arteriography for diagnosis of intimal reproduction of the illustrations. We also acknowledgethe United Bristol HealthcareNHS Trust MedicalResearch Committee for their defects and were accurate in assessing compromised financial assistance which enabled purchase of angioscopy luminal diameter. Miller et al. were the first to describe equipment.

Eur J Vasc Endovasc Surg Vol 11, January 1996 18 Y.G. Wilson et aL

References 19 MILROYCM, SCOTTDJA, BEARDJD, HORROCKSM/BRADFIELD J~WB. Histological appearances of the long saphenous vein. ] Pathol 1989; 159: 311-316. 1 MICHAELSJA. Choice of material for above-knee femoropopliteal 20 RUTHERFORD RB, FLANIGAN DP, GUPTA SK et aI. Suggested bypass graft. Br J Surg 1989; 76: 7-14. standards for reports dealing with lower extremity ischaemia. ] 2 VEITH FJ, GUPTA SK, ASCER E et al. Six-year prospective Vasc Surg 1986; 4: 80-94. multicentre randomized comparison of autologous saphenous 21 DAVIESAH, MAGEETR, BAIRDRN, HORROCKSM. Intra-operative vein and expanded polytetrafluoroethylene grafts in infra- measurement of vascular graft resistance as a predictor of early inguinal arterial reconstructions. J Vasc Surg 1986; 3: 104-114. outcome. Br ] Surg 1993; 80: 854-857. 3 DONALDSONMC, WFLiTTEMOREAD, MANNICKJA. An argument in 22 MILLERA, STONEBRIDGEPA, JEPSEN SJ et al. Continued experience favour of all-autogenous tissue for vascular bypasses below the with intraoperative angioscopy for monitoring infrainguinal inguinal ligament. Advances in Surgery 1991; 24: 69-90. bypass grafting. Surgery 1991; 109: 286-293. 4 STEPTLL, FLINNWR~ MCCARTHYWJ, BARTLETTS% BERGANJJ, YAO 23 CLAIRDG, GOLDENMA, MANNICKJA, WHITTEMOREAD, DONALD- JST. Technical defects as a cause of early graft failure after SON MC. Randomized prospective study of angioscopically femoro-distal bypass. Arch Surg 1987; 122: 599-604. assisted in situ saphenous vein grafting. J Vasc Surg 1994; 19: 5 SZILAGYZDE, ELLIOTrJP, HAGEMANJH, SMITHRF, DALL'OLMoCA. 992-1000. Biologic fate of autogenous vein implants as arterial substitutes: 24 DAVIES AH, MAGEE TR, HORROCKSM. Vein graft factors in the clinical, angiographic and histopathologic observations in outcome of femorodistal bypass. Cur J Vasc Surg 1994; 8: femoro-popliteal operations for atherosclerosis. Ann Surg 1973; 249-256. 178: 232-246. 25 CHEANVECHALC, EFFLERDB, HOOPERJR. The structural study of 6 LOGERFO FW, CORSONJW, MANNICKJA. Improved results with the saphenous vein. Ann Thorac Surg 1975; 20: 636~45. femoropopliteal vein grafts for limb salvage. Arch Surg 1977; 112: 26 SHAH DM, CHANG BB, LEOPOLD PW, CORSON JW, LEATHERRP, 567-570. KARMODY AM. The anatomy of the greater saphenous venous 7 SZILAGYIDE, HAGEMANJH, SMITHRE, ELLIOTTJP, BROWNF, DIETZ system. J Vasc Surg 1986; 3: 273-283. P. Autogefious vein grafting in femoropopliteal atherosclerosis : 27 MAP,IN ML, VEITH FJ, PANETTA TF. Saphenous vein biopsy : a the limits of its effectiveness. Surgery 1979; 86: 836-851. predictor of vein graft failure. J Vasc Surg 1993; 18: 407-415. 8 PANETTATF, MARIN ME, VEITH FJ et al. Unsuspected preexisting 28 DAVIESAH, MAGEETR, SHEFFraLDE, BAIRDRN, HORROCKSM. The saphenous vein disease : an unrecognised cause of vein bypass aetiology of vein graft stenoses. Cur J Vasc Surg 1994; 8: failure. ]Vasc Surg 1992; 15: 102-112. 389-394. 9 DAVIESAH, MAGEETR, BAIRDRN, SHEFFIELDE, HoP,ROCKS M. Vein 29 GRUNDPESTWl ~, LITVACK]~ CLICKD et al. Intraoperative decisions compliance : a pre-operative indicator of vein morphology and based on angioscopy in peripheral . Circulation of veins at risk of vascular graft stenosis. Br J Surg 1992; 79: 1988; 78(suppl I): 1-13-I-17. 1019-1021. 30 MILLER A, CAMPBELL DR, GIBBONS GW et al. Routine intra- 10 ADCOCKOT, ADCOCKGLD, WHEELERJR, GREGORYRT, SNYDERSO, operative angioscopy in lower extremity revascularisation. Arch GAYLERG. Optimal techniques for harvesting and preparation of Surg 1989; 124: 604-608. reversed autogenous vein grafts for use as arterial substitutes : a 31 GILBERTSON JJ, WALBH DB, ZWOLAK RM et al. A blinded review. Surgery 1984; 96: 886-893. comparison of angiograph~ angioscopy, and duplex scanning in 11 RuoPP BA, RANLEY JJ, HANNAN LA et aI. Real time duplex the intra-operative evaluation of in-situ saphenous vein bypass ultrasound mapping of the greater saphenous vein before in-situ grafts. J Vasc Surg 1992; 15: 121-129. infra-inguinal revascularisation. J Vasc Surg 1979; 138: 211-218. 32 MrLLERA, MARCACCIOEJ, TANNENBAUMGA et al. Comparison of 12 LEOPOLD PW, SHANDALLA, KUPINSKIAM et al. Role of B-mode angioscopy and for monitoring infrainguinal venous mapping in infra-inguinal in-situ saphenous vein bypass vein grafts : results of a prospective randomized trial. J bypass. Am ] Surg 1986; 152: 206-210. Vasc Surg 1993; 13: 382-398. 13 BEARD JD, SCOTT DJA, SKIDMORE RI BAIRD RN, HORROCKS M. 33 NEVILLE RF, YASUHARAH, WATANABEBL CANADYJ, DURAN W, Operative assessment of femoro-distal bypass grafts using a new HOBSON RW. Endovascular management of arterial intimal Doppler flowmeter. Br J Surg 1989; 76: 925-928. defects : an experimental comparison by arteriograph~ angio- 14 MILLSJL, FUJITANIRL"V[, TAYLOR SM. Contribution of routine intra- scopy and intravascular ultrasound. J Vasc Surg 1991; 13: operative completion arteriography to early infra-inguinal 496-502. bypass patency. Am ] Surg 1992; 164: 506-511. 34 MILLER A, JEPSEN SJ~ STONEBR1DGEPA et aL New angioscopic 15 SALESCM, MARINME, VEITHFJ et aI. Saphenous vein angioscopy: findings in graft failure after infrainguinal bypass. Arch Surg a valuable method to detect unsuspected venous disease. J Vasc 1990; 125: 749-755. Surg 1993; 18: 198-206. 35 MARCACCIO EJ, MILLER A, TANNENBAUMCA. Angioscopically 16 DAVIESAH, MAGEETR, THOMPSONJF et al. Preliminary experience directed interventions improve arm vein bypass grafts. ] Vasc of angioscopy in femoro-distal bypass. Ann RCS Eng 1993; 75: Surg 1993; 17: 994-1004. 178-180. 36 STONEBPaDGEPA, Mur~E JA. Vascular : useful tool or 17 Gr~GG MJ, NICOLAIDES AN, WOLFE JHN. Femorodistal vein interesting toy. Br J Surg 1994; 81: 633-634. bypass graft stenoses. Br ] Surg 1988; 75: 737-740. 18 DAVIESAH, MAGEE TR, TENNANTSGW, LAMONTPM, BAIRD RN, HOP,RocKS M. Criteria for identification of the "at-risk" infra- Accepted 8 December 1994 inguinal bypass graft. Cur J Vasc Surg 1994; 8: 315-319.

Cur J Vasc Endovasc Surg Vol 11, January 1996