- Email: [email protected]
Prediction of the immediate outcome of femoropopliteal saphenous vein bypass by angiographic runoff score
Eur J Vasc Endovasc Surg 15, 220-224 (1998)
Prediction of the Immediate Outcome of Femoropopliteal Saphenous Vein Bypass by Angiographic Runoff Score A. AIb~ick1% F. Biancari 1, O. Saarinen 2 and M. Lep~intalo 1 IDivision of Vascular Surgery, Department of Surgery, and 2Department of Radiology, Helsinki University Central Hospital, Helsinki, Finland Objectives: To determine the value of the Ad Hoc scoring system (SVS/ISCVS) in predicting the immediate outcome of femoropopliteal saphenous vein grafts. Design: Retrospective study. Materials: One hundred and twenty patients underwent 132 primary femoropopliteal vein bypass procedures, 32 for claudication and lOOfor critical leg ischemia (CLI). Methods: The outflow arteries were graded according to the Ad Hoc scoring system (SVSflSCVS). Postoperative immediate graft patency and leg salvage to the period of the,first 30 days after surgery. Results: Ninety-one per cent of claudicants and 83% of CLI patients had immediate patency. The overall 30-day patency rate was 85%. Leg salvage rate was 91% for the patients with CLI. Patients with score in the highest quartile were found to have a 8.7 times higher risk for immediate graft occlusion (p =0.005). Multivariate analysis showed that the Ad Hoc score was predictive of immediate patency (p = 0.0006) and leg salvage (p = 0.0004). In patients with a score <_7.5 and in those with a score >7.5, the patency rates were 95% and 66% (p=O.O01), and the leg salvage rates were 97% and 80%, (p = 0.004), respectively. Conclusions: The Ad Hoc scoring system is useful in predicting the immediate outcome of femoropopliteaI saphenous vein grafts. Key words: Femoropopliteal bypass; Saphenous vein; Runoff; Scoring system.
Introduction
Knowledge of the stage of a disease enables the clinician to predict the future development of the disease, to plan appropriate treatment, and to compare patient populations and outcome between different centres. In the same way, the decision-making process in the treatment of vascular diseases requires a precise description of the disease and its extension. 1 With this aim, in 1986, the Ad Hoc Committee on Reporting Standards, Society for Vascular Surgery/North American Chapter, International Society for Cardiovascular S u r g e r y ( S V S / I S C V S ) 2 proposed a scheme for grading the runoff status of the lower limb based on angiographic findings which, unfortunately, has received the attention of only a few studies, the results of which are contradictory. 3-8 The aim of this study was, therefore, to retrospectively evaluate the impact of
* Please address all correspondence to: A. Alback, Division of Vascular Surgery, Department of Surgery, Helsinki University Central Hospital, Kasarmikatu 11-13, FIN-00130 Helsinki, Finland.
this scoring system on the immediate outcome of femoropopliteal saphenous vein bypass surgery.
Patients and Methods
Intraoperative monitoring of graft haemodynamics, with its possible ability to improve the 30-day reconstruction outcome, was not part of our routine until 1992. We therefore chose to analyse a sample of 120 consecutive patients who underwent 132 primary femoropopliteal bypass grafting operations for the treatment of claudication (32) or critical leg ischaemia (100) (CLI) from 1986 to 1991. At that time no haemodynamic measurements were made which would have altered the treatment policy by necessitating additional inflow or outflow procedures. 9 All operations were performed using an autogenous saphenous vein. The minimum outer diameter of the grafts was determined. Twenty-seven patients were diabetic. Patient data and
c
1078-5884/98/030220+05 $12.00/0 © 1998 W.B. Saunders Company Ltd.
Angiographic Runoff Score
Table 1. Patient characteristics and potential risk factors for femoropopliteal graft failure.
N. operations Sex M/F Age (range) Diabetes (%) Claudication/CLI Median preoperative ABI Median score in claudicants/CLI In situ vein grafts (%) Reversed vein grafts (%) Median minimumouter graft diameter in mm AK/BKbypasses
132 77/55 64 (57-72) 28 (21%) 32/ 100 0.32 (0.22-0.43) 4.0/6.5 41 (31%) 91 (69%) 4.0 (3.0-4.0) 67/65
AK: femorOpoplitealabove-kneebypasses; BK: femoropoplitealbelow-kneebypasses.
risk factors analysed in this study are shown in Table I. All patients underwent preoperative angiography and the outflow arteries were retrospectively graded according to the system proposed by the Ad Hoc Committee on Reporting Standards, SVS/ISCVS. 2 For both femoropopliteal above-knee and femoropopliteal below-knee bypasses, the runoff was graded by the crural scoring system which assigned a score to the anterior tibial artery, posterior tibial artery, and peroneal artery. 3 To rule out technical errors, intraoperative completion angiography was performed by injection of contrast medium into the grafts with the inflow artery cross-clamped. All patients were given 100-250 mg per day of acetylsalicylic acid during the postoperative period. Ankle-brachial index (ABI) measurements were routinely done on the first and seventh postoperative day. Graft patency was determined by pulse palpation and pressure measurements, whereas standard duplex scan and angiography were performed whenever there was a suspicion of graft failure. Postoperative immediate haemodynamic success was defined as graft patency and an improvement of the ABI more than 0.15 until 30 days after surgery. 1° Immediate patency and immediate leg salvage also refer to the period of the first postoperative month. Analysis of data was facilitated by the use of a microcomputer database and the statistical software SPSS for Windows 7.0 (SSPS Inc., 444 North Michigan Avenue, Chicago, IL 60611, USA). The data of nonparametric variables showed a skewed distribution and therefore these were presented as median and interquartile range. The significance of any difference in medians between subgroups were calculated with the Mann-Whitney U-test. For the multivariate analysis of the impact of potential risk factors for graft occlusion and postoperative major amputation, the variables age, sex, diabetes, outflow vessel anatomy,
221
type and size of vein graft, preoperative ABI, Fontane classification of the disease, and runoff score were entered into a logistic regression model. After dividing the sample for runoff score into quartiles, a logistic regression model was also used to estimate the effect of increasing runoff scores on graft occlusion. Receiver operating characteristic (ROC) curves were used to test the adequacy of the runoff scoring system and to define an appropriate cut-off value. An area under the ROC curve for a test greater than 0.5 + twice the standard error (S.E.) was considered statistically significant. Pearson's test was used to correlate runoff score to ABI values. Cases with evidence of a cause for the graft occlusion other than compromised runoff were excluded from the analysis on the predictive value of the runoff score but included in the haemodynamic success, patency and leg salvage rates.
Results
Seven patients (6%) died in the immediate postoperative period. There were 20 graft occlusions (15%) within 30 days after surgery, of which three were in reconstructions done for claudication and 17 in the ones done for CLI. The immediate patency rates were 91% and 83%, respectively, and the overall immediate patency rate was 85%. Seven of the occluded grafts (5%) were reoperated immediately. Reocclusion occurred in six of them, while in one instance the graft remained patent. Only one leg required amputation after reoperation. Of the remaining 13 graft occlusions, five (4%) led to a major amputation. Progression of infection or gangrene was the indication for amputation of four legs with a patent graft. Leg salvage was thus achieved in 91% of CLI patients, whilst one patient operated for claudication had a below-knee amputation as a result of graft occlusion. An immediate haemodynamic success was achieved in 99 reconstructions (75%); 27 of them were in claudicants (84%), and 72 were in CLI patients (72%). In one graft an intact valve cusp was believed to be the probable reason for the occlusion, revalvulectomy was performed and the graft remained patent. In another graft, poor vein quality led to intraoperative thrombosis during the primary operation with consequent thrombectomy, revision of the distal graft and new completion angiography. This reconstruction, with a runoff score of four, occluded the next day and was not reoperated. These two reconstructions were excluded from the analysis of the impact of the runoff score. The median increase from preoperative to postoperative ABI was 0.37 (0.14-0.53). The median Ad Eur J Vasc EndovascSurg Vol 15, March 1998
A. AIb~ick et aL
222
a score _<7.5 and in those with a score >7.5, the patency rates were 95% and 66% (p = 0.001), and the leg salvage rates were 97% and 80% (p = 0.004), respectively. 6 5
©
Discussion
4
Although several reports have confirmed the validity of the intraoperative pre- a n d / o r post-reconstruction angiography, blood flow and peripheral resistance measurements in predicting the outcome of lower limb revascularisations, 7"11-13 there is still lack of preoperative parameters predictive of the fate of these <3 3-5.75 5.76-8.0 >8 grafts. Technical errors, vein quality, biochemical and Score quartiles rheological factors 14'~5 may all have impact on the Fig. 1. Angiographic crural runoff score and risk for immediate outcome of arterial bypass surgery, but inadequate femoropopliteal saphenous vein graft occlusion. ([]) Occlusion. distal runoff leading to poor blood flow is considered the main cause for graft failure. 4'6'7 Since preoperative Hoc runoff score was 4.0 in claudicants and 6.5 in CLI angiograms show a map of ischaemic disease, anpatients (p = 0.01). A statistically significant correlation giographic findings were logically the basis for the between the runoff score and the preoperative (r = 0.3; development of a scoring system which should give p=0.01), and postoperative (r=0.3; p=0.0001) ABI weight to those vessels contributing to the runoff values was found. The correlation between change status. 2 Unfortunately, this method for staging of leg in ABI and runoff score almost reached statistical ischaemia has been tested by only a few studies which significance (r = 0.18; p = 0.058). have given contradictory results, g-s Some of these were Among patients with a patent graft during the im- biased by analysis of results in patients undergoing mediate postoperative period, the median Ad Hoc heterogenous procedures, both for the type and the runoff score was 5.0, whereas in those with graft length of grafting. 3-5"7Furthermore, Takolander et al 4 occlusion the median score was 8.75 (p=0.001). required the grafts to be patent at 1 month after Patients with major amputation were found to have a surgery, as an inclusion criteria in their study. In fact median score of 9.0, whereas patients with leg salvage they experienced eight (15%) immediate occlusions had a median score of 5.0 (p=0.003). Compared to with a mean score not significantly different from the scores in the lowest quartile, scores in the highest patent grafts. They found no predictive value of the quartile had a 8.7 (p=0.005) times higher risk for runoff score for 1-year graft patency. However, despite immediate graft occlusion (Fig. 1). Among the risk duplex surveillance, occlusions due to neointimal factors evaluated by multivariate analysis, o n l y an- hyperplasia and even progression of the athgiographic runoff score had an independent statistical erosclerotic disease might have affected the results. In significance in predicting immediate occlusion of these our opinion, this scoring system can offer only a static grafts (p = 0.0006; odds ratio = 1.44). This was also true picture of the disease at the time of angiography and for postoperative major amputation when the pre- operation and is thus likely to particularly affect the operative ABI was entered after logarithmic trans- immediate outcome. The 30-day graft failure rate has formation (p = 0.0004, odds ratio = 2.24). been recognised as the area for the greatest potential ROC curves for the runoff score in predicting im- improvement in infrainguinal revascularisation and is mediate graft occlusion (area under the curve: 0.76; considered to be predominantly related to the runoff s.~.: 0.23) and postoperative major amputation (area status. 6'14 For these reasons, in the present study the under the curve: 0.8; s.~.: 0.31) were not statistically immediate postoperative outcome was considered an significant, but the best cut-off point of the method appropriate measure for testing this scoring system, was at 7.5. Using such a cut-off point for positive test, thus avoiding bias from intimal hyperplasia and disthe sensitivity, specificity, positive predictive value, ease progression. At the time of the study, grafts were and negative predictive value for immediate graft routinely assessed by intraoperative angiography to occlusion were 75%, 76%, 34%, and 95%, respectively, rule out immediate failure due to technical errors. whereas for major amputation the corresponding figThe reported positive correlation between The Ad ures were 78%, 73%, 20%, and 97%. In patients with Hoc score and runoff resistance, 3 and flow waveform 3
Eur J Vasc Endovasc Surg Vo115,March 1998
Angiographic Runoff Score
223
2 RUTHEI~ORD, RB, FLANIGAN, DE GUPTA, S et al. Suggested pattern s indicates that angiographic grading of lower standards for reports dealing with lower extremity ischemia. J limb arteries can provide a functional assessment of Vasc Surg 1986; 4: 80-94. 3 PETERKIN, GA, MANABE, S, LAMORTE, W~V, MENZOIAN, JO. the disease severity. This was true in our study, as the Evaluation of a proposed standard reporting system for prescore correlated with both preoperative and postoperative angiograms in infrainguinal bypass procedures: anoperative haemodynamics. However, the results from giographic correlates of measured runoff resistance. J Vasc Surg our study, in which a quarter of the reconstructions 1988; 7: 379-385. 4 TAKOLANDERrR, FISCHER-COLBRIE,W, JOGESTRAND,T~ OHLSEN, were performed with a runoff score more than 8, H, OLOESSON,P, SWEDENBORG,J. The "Ad Hoc" estimation of contradict the favourable results with bypass grafts to outflow does not predict patency of infrainguinal rean isolated popliteal artery segment that have preconstructions. Eur J Vasc Endovase Surg 1995; 10: 187-191. 5 TORDOIR, JHM, VAN DER PLAS, JP, JACOBS, JHM, KISTLAAR, viously been reported. 16 PJEHM. Factors determining the outcome of crural and pedal Attempts to improve the resolution of the Ad Hoc revascularisation for critical limb ischaemia. Eur J Vasc Surg scoring system have been done, 3'7 but the limits en1993; 7: 82-86. 6 BLANKESTEIJN,JD, GERTLER, JP, BREWSTER,DC, CAMBI~A, RP, countered by other authors by using this method LAMURAGLIA, GM, ABBOTT, WM. Intraoperative determinants might be related to the intrinsic deficiencies of contrast of infrainguinal bypass graft patency: a prospective study. Eur angiography. Since the demonstration of vessel patJ Vasc Endovasc Surg 1995; 9: 375-382. 7 KARACAGIL,S, ALMGREN,B, BERGSTROM,Rr BOWALD,S, ERIKSSON, ency depends on the delivery of radiopaque contrast I. Postoperative predictive value of a new method of intramaterial to the distal runoff arteries, the loss of an operative angiographic assessment of runoff in femoropopliteal adequate concentration of contrast agent may cause bypass grafting, j Vasc Surg 1989; 10: 400-407. failure in showing vessels contributing to runoff. 17-2° 80KADOME, K, ONOHARA,% YAMAMURA,S, MII, S, SUGIMACHI,K. Evaluation of proposed standards for runoff in femoropopliteal Up to one quarter of vessels demonstrated to be patent arterial reconstructions: correlation between runoff score and on Doppler ultrasonography or at operation may fail flow waveform pattern. A preliminary report, f Cardiovasc Surg 1991; 32: 353-359. to opacify on arteriogramsY -2° When compared with 9 LUTHER, M, LEPANTALO, M. Femoropopliteal reconstructions. intraoperative angiography, preoperative angiography Influence of risks factors on outcome. Ann Chir Gynaecol 1996; failed to show patent runoff vessels in 10%-21% of 85: 238-246. crural arteries, and 40%-86% of combined calf and 10 LEP.KNTALO,M, M~TZKE, S, LUTHER,M. Is it necessary to assess pressures after infrainguinal bypass? In: Greenhalgh RM, Fowkes foot arteries. 2°-22 Magnetic resonance angiography FGR, eds. Trials and Tribulations of Vascular Surgery 7: 277-288. (MRA) has also been shown to demonstrate flow in 11 STIRNEMANN,I~, RIS, HB, Do, D, HKMMERLI, R. Intraoperative flow measurement of distal runoff: a valid predictor of outcome the presence of severe occlusive disease by detecting of infrainguinal bypass surgery. Eur J Surg 1994; 160: 431-436. the intense signal generated by non-pulsatile blood 12 ASCElb E, WHITE, SA, VEITH, FJ, MORIN, L, FREEMAN,K, GUPTA, flow. 23 Studies comparing MRA with contrast angioSK. Outflow resistance measurement during infrainguinal argraphy showed that MRA is able to detect patent terial reconstructions: a reliable predictor of limb salvage. Am J Surg 1987; 154: 185-188. runoff vessels that are not opacified by preoperative 13 LUNDELL,A, BERGQVIST,D. Prediction of early graft occlusion angiography, with a prevalence ranging from 19% to in femoropopliteal and femorodistal reconstruction by measure24% when the actual patency was confirmed by surment of volume flow with a transit time flowmeter and calculation of peripheral resistance. Eur J Vase Surg 1993; 7: 704-708. gical exploration or intraoperative contrast angioDONALDSON,MC, MANNICK, JA, WHITTEMORE, AD. Causes of graphy. 24-2s Therefore, we might expect further 14 primary graft failure after in situ saphenous vein bypass grafting. improvement in the correlation between the Ad Hoc J Vasc Surg 1992; 15: 113-120. runoff score and the outcome of infrainguinal re- 15 WOODBURN,KR, RUMLEY, A, LOWE, GD, LOVE, JG, MURRAY, GD, POLLOCK, JG. Clinical, biochemical, and rheologic factors vascularisation by detecting patent vessels with alaffecting the outcome of infrainguinal bypass grafting. J Vasc ternative imaging techniques. Surg 1996; 24: 639-646.
Acknowledgements The authors wish to thank Jukka Ollgren, MSc, for help with the statistical analysis. The study was supported by a grant from Finska L~ikaresallskapet, which is gratefully acknowledged. References 1 WALDEN, R, MODAN, M, BASS, A, SCHNEIDERMAN,J, ADAR, R, Scoring of vascular disease in the lower extremities. J Cardiovasc Surg 1989; 30: 210-215.
16 KAUFMAN,JL, WHITTEMORE,AD, CoucfI, NP, MANNICK,JA. The fate of bypass grafts to an isolated popliteal artery segment. Surgery 1982; 92: 1027-1030. 17 BEARD,JD, SCOTT,DJA, EVANS,JM, SKIDMORE,Rr HORROCKS,M. Pulse generated runoff: a new method of determining calf vessel patency. Br J Surg 1988; 75: 361-363. 18 SIMMS,HM, HARDMAN,J, SLANEY,G. The relevance of prebypass pedal arch assessment. Br J Surg 1984; 71: 381. 19 CAMPBELL,WB, FLETCHER,EL, HANDS, LJ. Assessment of the distal lower limb arteries: a comparison of arteriography and doppler ultrasound. Ann R Coll Surg Engl 1986; 68: 37-39. 20 Rlcco, JB, PEARCE, WI-I, YAO, JST, FLINN, WR, BERGAN,JJ. The use of operative bypass arteriography and doppler ultrasound recording to select patients for extended femoro-distal bypass. Ann Surg 1983; 198: 646-653. 21 PATEL,PR, SEMEL,L, CLAUSS,RH. Extended reconstruction rate for limb salvage with intraoperative prereconstruction angiography. J Vasc Surg 1988; 7: 531-537. Eur J Vasc Endovasc Surg Vol 15, March 1998
224
A. AIb~ick et aL
22 SCARPATO,R, GEMBAROWICZ,R, FARBER,Set al. Intraoperative occlusive disease: accuracy and reliability of time-of-flight MR prereconstruction arteriography. Arch Surg 1981; 116: 1053-1055. angiography. Radiology 1994; 192: 351-357. 23 MULLIGAN,SA, MATSUDA,T, LANZER,Pet al. Peripheral arterial 27 OWEN, RS, CARPENTER,JP, BAUM, RA, PERLOI% LJ, COPE, C. occlusive disease: prospective comparison of MR angiography Magnetic resonance imaging of angiographically occult runoff and color duplex US with conventional angiography. Radiology vessels in peripheral arterial occlusive disease. N Engl J Med 1991; 178: 695-700. 1992; 326: 1577-1581. 24 CARPENTER,JP, OWEN,RS, BAUM,RA et aL Magnetic resonance 28 CAILPENTEI~JP, GOLDEN,MA, BARKER,CE HOLLAND,GA, BAUM, angiography of peripheral runoff vessels. J Vasc Surg 1992; 16: RA. The fate of bypass grafts to angiographically occult runoff 807-815. vessels detected by magnetic resonance angiography. J Vasc Surg 25 AMERICANCOLLEGEOF RADIOLOGYNEW TECHNOLOGYASSESS1996; 23: 483-489. MENT GROUP. Multicenter trial to evaluate peripheral vascular magnetic resonance angiography. JAMA 1995; 274: 875-880. 26 McCAULEY,TR, MONIB,A, DICKEY,KW et al. Peripheral vascular Accepted 24 September 1997
Eur J Vasc Endovasc Surg Vol 15, March 1998
Prediction of the Immediate Outcome of Femoropopliteal Saphenous Vein Bypass by Angiographic Runoff Score A. AIb~ick1% F. Biancari 1, O. Saarinen 2 and M. Lep~intalo 1 IDivision of Vascular Surgery, Department of Surgery, and 2Department of Radiology, Helsinki University Central Hospital, Helsinki, Finland Objectives: To determine the value of the Ad Hoc scoring system (SVS/ISCVS) in predicting the immediate outcome of femoropopliteal saphenous vein grafts. Design: Retrospective study. Materials: One hundred and twenty patients underwent 132 primary femoropopliteal vein bypass procedures, 32 for claudication and lOOfor critical leg ischemia (CLI). Methods: The outflow arteries were graded according to the Ad Hoc scoring system (SVSflSCVS). Postoperative immediate graft patency and leg salvage to the period of the,first 30 days after surgery. Results: Ninety-one per cent of claudicants and 83% of CLI patients had immediate patency. The overall 30-day patency rate was 85%. Leg salvage rate was 91% for the patients with CLI. Patients with score in the highest quartile were found to have a 8.7 times higher risk for immediate graft occlusion (p =0.005). Multivariate analysis showed that the Ad Hoc score was predictive of immediate patency (p = 0.0006) and leg salvage (p = 0.0004). In patients with a score <_7.5 and in those with a score >7.5, the patency rates were 95% and 66% (p=O.O01), and the leg salvage rates were 97% and 80%, (p = 0.004), respectively. Conclusions: The Ad Hoc scoring system is useful in predicting the immediate outcome of femoropopliteaI saphenous vein grafts. Key words: Femoropopliteal bypass; Saphenous vein; Runoff; Scoring system.
Introduction
Knowledge of the stage of a disease enables the clinician to predict the future development of the disease, to plan appropriate treatment, and to compare patient populations and outcome between different centres. In the same way, the decision-making process in the treatment of vascular diseases requires a precise description of the disease and its extension. 1 With this aim, in 1986, the Ad Hoc Committee on Reporting Standards, Society for Vascular Surgery/North American Chapter, International Society for Cardiovascular S u r g e r y ( S V S / I S C V S ) 2 proposed a scheme for grading the runoff status of the lower limb based on angiographic findings which, unfortunately, has received the attention of only a few studies, the results of which are contradictory. 3-8 The aim of this study was, therefore, to retrospectively evaluate the impact of
* Please address all correspondence to: A. Alback, Division of Vascular Surgery, Department of Surgery, Helsinki University Central Hospital, Kasarmikatu 11-13, FIN-00130 Helsinki, Finland.
this scoring system on the immediate outcome of femoropopliteal saphenous vein bypass surgery.
Patients and Methods
Intraoperative monitoring of graft haemodynamics, with its possible ability to improve the 30-day reconstruction outcome, was not part of our routine until 1992. We therefore chose to analyse a sample of 120 consecutive patients who underwent 132 primary femoropopliteal bypass grafting operations for the treatment of claudication (32) or critical leg ischaemia (100) (CLI) from 1986 to 1991. At that time no haemodynamic measurements were made which would have altered the treatment policy by necessitating additional inflow or outflow procedures. 9 All operations were performed using an autogenous saphenous vein. The minimum outer diameter of the grafts was determined. Twenty-seven patients were diabetic. Patient data and
c
1078-5884/98/030220+05 $12.00/0 © 1998 W.B. Saunders Company Ltd.
Angiographic Runoff Score
Table 1. Patient characteristics and potential risk factors for femoropopliteal graft failure.
N. operations Sex M/F Age (range) Diabetes (%) Claudication/CLI Median preoperative ABI Median score in claudicants/CLI In situ vein grafts (%) Reversed vein grafts (%) Median minimumouter graft diameter in mm AK/BKbypasses
132 77/55 64 (57-72) 28 (21%) 32/ 100 0.32 (0.22-0.43) 4.0/6.5 41 (31%) 91 (69%) 4.0 (3.0-4.0) 67/65
AK: femorOpoplitealabove-kneebypasses; BK: femoropoplitealbelow-kneebypasses.
risk factors analysed in this study are shown in Table I. All patients underwent preoperative angiography and the outflow arteries were retrospectively graded according to the system proposed by the Ad Hoc Committee on Reporting Standards, SVS/ISCVS. 2 For both femoropopliteal above-knee and femoropopliteal below-knee bypasses, the runoff was graded by the crural scoring system which assigned a score to the anterior tibial artery, posterior tibial artery, and peroneal artery. 3 To rule out technical errors, intraoperative completion angiography was performed by injection of contrast medium into the grafts with the inflow artery cross-clamped. All patients were given 100-250 mg per day of acetylsalicylic acid during the postoperative period. Ankle-brachial index (ABI) measurements were routinely done on the first and seventh postoperative day. Graft patency was determined by pulse palpation and pressure measurements, whereas standard duplex scan and angiography were performed whenever there was a suspicion of graft failure. Postoperative immediate haemodynamic success was defined as graft patency and an improvement of the ABI more than 0.15 until 30 days after surgery. 1° Immediate patency and immediate leg salvage also refer to the period of the first postoperative month. Analysis of data was facilitated by the use of a microcomputer database and the statistical software SPSS for Windows 7.0 (SSPS Inc., 444 North Michigan Avenue, Chicago, IL 60611, USA). The data of nonparametric variables showed a skewed distribution and therefore these were presented as median and interquartile range. The significance of any difference in medians between subgroups were calculated with the Mann-Whitney U-test. For the multivariate analysis of the impact of potential risk factors for graft occlusion and postoperative major amputation, the variables age, sex, diabetes, outflow vessel anatomy,
221
type and size of vein graft, preoperative ABI, Fontane classification of the disease, and runoff score were entered into a logistic regression model. After dividing the sample for runoff score into quartiles, a logistic regression model was also used to estimate the effect of increasing runoff scores on graft occlusion. Receiver operating characteristic (ROC) curves were used to test the adequacy of the runoff scoring system and to define an appropriate cut-off value. An area under the ROC curve for a test greater than 0.5 + twice the standard error (S.E.) was considered statistically significant. Pearson's test was used to correlate runoff score to ABI values. Cases with evidence of a cause for the graft occlusion other than compromised runoff were excluded from the analysis on the predictive value of the runoff score but included in the haemodynamic success, patency and leg salvage rates.
Results
Seven patients (6%) died in the immediate postoperative period. There were 20 graft occlusions (15%) within 30 days after surgery, of which three were in reconstructions done for claudication and 17 in the ones done for CLI. The immediate patency rates were 91% and 83%, respectively, and the overall immediate patency rate was 85%. Seven of the occluded grafts (5%) were reoperated immediately. Reocclusion occurred in six of them, while in one instance the graft remained patent. Only one leg required amputation after reoperation. Of the remaining 13 graft occlusions, five (4%) led to a major amputation. Progression of infection or gangrene was the indication for amputation of four legs with a patent graft. Leg salvage was thus achieved in 91% of CLI patients, whilst one patient operated for claudication had a below-knee amputation as a result of graft occlusion. An immediate haemodynamic success was achieved in 99 reconstructions (75%); 27 of them were in claudicants (84%), and 72 were in CLI patients (72%). In one graft an intact valve cusp was believed to be the probable reason for the occlusion, revalvulectomy was performed and the graft remained patent. In another graft, poor vein quality led to intraoperative thrombosis during the primary operation with consequent thrombectomy, revision of the distal graft and new completion angiography. This reconstruction, with a runoff score of four, occluded the next day and was not reoperated. These two reconstructions were excluded from the analysis of the impact of the runoff score. The median increase from preoperative to postoperative ABI was 0.37 (0.14-0.53). The median Ad Eur J Vasc EndovascSurg Vol 15, March 1998
A. AIb~ick et aL
222
a score _<7.5 and in those with a score >7.5, the patency rates were 95% and 66% (p = 0.001), and the leg salvage rates were 97% and 80% (p = 0.004), respectively. 6 5
©
Discussion
4
Although several reports have confirmed the validity of the intraoperative pre- a n d / o r post-reconstruction angiography, blood flow and peripheral resistance measurements in predicting the outcome of lower limb revascularisations, 7"11-13 there is still lack of preoperative parameters predictive of the fate of these <3 3-5.75 5.76-8.0 >8 grafts. Technical errors, vein quality, biochemical and Score quartiles rheological factors 14'~5 may all have impact on the Fig. 1. Angiographic crural runoff score and risk for immediate outcome of arterial bypass surgery, but inadequate femoropopliteal saphenous vein graft occlusion. ([]) Occlusion. distal runoff leading to poor blood flow is considered the main cause for graft failure. 4'6'7 Since preoperative Hoc runoff score was 4.0 in claudicants and 6.5 in CLI angiograms show a map of ischaemic disease, anpatients (p = 0.01). A statistically significant correlation giographic findings were logically the basis for the between the runoff score and the preoperative (r = 0.3; development of a scoring system which should give p=0.01), and postoperative (r=0.3; p=0.0001) ABI weight to those vessels contributing to the runoff values was found. The correlation between change status. 2 Unfortunately, this method for staging of leg in ABI and runoff score almost reached statistical ischaemia has been tested by only a few studies which significance (r = 0.18; p = 0.058). have given contradictory results, g-s Some of these were Among patients with a patent graft during the im- biased by analysis of results in patients undergoing mediate postoperative period, the median Ad Hoc heterogenous procedures, both for the type and the runoff score was 5.0, whereas in those with graft length of grafting. 3-5"7Furthermore, Takolander et al 4 occlusion the median score was 8.75 (p=0.001). required the grafts to be patent at 1 month after Patients with major amputation were found to have a surgery, as an inclusion criteria in their study. In fact median score of 9.0, whereas patients with leg salvage they experienced eight (15%) immediate occlusions had a median score of 5.0 (p=0.003). Compared to with a mean score not significantly different from the scores in the lowest quartile, scores in the highest patent grafts. They found no predictive value of the quartile had a 8.7 (p=0.005) times higher risk for runoff score for 1-year graft patency. However, despite immediate graft occlusion (Fig. 1). Among the risk duplex surveillance, occlusions due to neointimal factors evaluated by multivariate analysis, o n l y an- hyperplasia and even progression of the athgiographic runoff score had an independent statistical erosclerotic disease might have affected the results. In significance in predicting immediate occlusion of these our opinion, this scoring system can offer only a static grafts (p = 0.0006; odds ratio = 1.44). This was also true picture of the disease at the time of angiography and for postoperative major amputation when the pre- operation and is thus likely to particularly affect the operative ABI was entered after logarithmic trans- immediate outcome. The 30-day graft failure rate has formation (p = 0.0004, odds ratio = 2.24). been recognised as the area for the greatest potential ROC curves for the runoff score in predicting im- improvement in infrainguinal revascularisation and is mediate graft occlusion (area under the curve: 0.76; considered to be predominantly related to the runoff s.~.: 0.23) and postoperative major amputation (area status. 6'14 For these reasons, in the present study the under the curve: 0.8; s.~.: 0.31) were not statistically immediate postoperative outcome was considered an significant, but the best cut-off point of the method appropriate measure for testing this scoring system, was at 7.5. Using such a cut-off point for positive test, thus avoiding bias from intimal hyperplasia and disthe sensitivity, specificity, positive predictive value, ease progression. At the time of the study, grafts were and negative predictive value for immediate graft routinely assessed by intraoperative angiography to occlusion were 75%, 76%, 34%, and 95%, respectively, rule out immediate failure due to technical errors. whereas for major amputation the corresponding figThe reported positive correlation between The Ad ures were 78%, 73%, 20%, and 97%. In patients with Hoc score and runoff resistance, 3 and flow waveform 3
Eur J Vasc Endovasc Surg Vo115,March 1998
Angiographic Runoff Score
223
2 RUTHEI~ORD, RB, FLANIGAN, DE GUPTA, S et al. Suggested pattern s indicates that angiographic grading of lower standards for reports dealing with lower extremity ischemia. J limb arteries can provide a functional assessment of Vasc Surg 1986; 4: 80-94. 3 PETERKIN, GA, MANABE, S, LAMORTE, W~V, MENZOIAN, JO. the disease severity. This was true in our study, as the Evaluation of a proposed standard reporting system for prescore correlated with both preoperative and postoperative angiograms in infrainguinal bypass procedures: anoperative haemodynamics. However, the results from giographic correlates of measured runoff resistance. J Vasc Surg our study, in which a quarter of the reconstructions 1988; 7: 379-385. 4 TAKOLANDERrR, FISCHER-COLBRIE,W, JOGESTRAND,T~ OHLSEN, were performed with a runoff score more than 8, H, OLOESSON,P, SWEDENBORG,J. The "Ad Hoc" estimation of contradict the favourable results with bypass grafts to outflow does not predict patency of infrainguinal rean isolated popliteal artery segment that have preconstructions. Eur J Vasc Endovase Surg 1995; 10: 187-191. 5 TORDOIR, JHM, VAN DER PLAS, JP, JACOBS, JHM, KISTLAAR, viously been reported. 16 PJEHM. Factors determining the outcome of crural and pedal Attempts to improve the resolution of the Ad Hoc revascularisation for critical limb ischaemia. Eur J Vasc Surg scoring system have been done, 3'7 but the limits en1993; 7: 82-86. 6 BLANKESTEIJN,JD, GERTLER, JP, BREWSTER,DC, CAMBI~A, RP, countered by other authors by using this method LAMURAGLIA, GM, ABBOTT, WM. Intraoperative determinants might be related to the intrinsic deficiencies of contrast of infrainguinal bypass graft patency: a prospective study. Eur angiography. Since the demonstration of vessel patJ Vasc Endovasc Surg 1995; 9: 375-382. 7 KARACAGIL,S, ALMGREN,B, BERGSTROM,Rr BOWALD,S, ERIKSSON, ency depends on the delivery of radiopaque contrast I. Postoperative predictive value of a new method of intramaterial to the distal runoff arteries, the loss of an operative angiographic assessment of runoff in femoropopliteal adequate concentration of contrast agent may cause bypass grafting, j Vasc Surg 1989; 10: 400-407. failure in showing vessels contributing to runoff. 17-2° 80KADOME, K, ONOHARA,% YAMAMURA,S, MII, S, SUGIMACHI,K. Evaluation of proposed standards for runoff in femoropopliteal Up to one quarter of vessels demonstrated to be patent arterial reconstructions: correlation between runoff score and on Doppler ultrasonography or at operation may fail flow waveform pattern. A preliminary report, f Cardiovasc Surg 1991; 32: 353-359. to opacify on arteriogramsY -2° When compared with 9 LUTHER, M, LEPANTALO, M. Femoropopliteal reconstructions. intraoperative angiography, preoperative angiography Influence of risks factors on outcome. Ann Chir Gynaecol 1996; failed to show patent runoff vessels in 10%-21% of 85: 238-246. crural arteries, and 40%-86% of combined calf and 10 LEP.KNTALO,M, M~TZKE, S, LUTHER,M. Is it necessary to assess pressures after infrainguinal bypass? In: Greenhalgh RM, Fowkes foot arteries. 2°-22 Magnetic resonance angiography FGR, eds. Trials and Tribulations of Vascular Surgery 7: 277-288. (MRA) has also been shown to demonstrate flow in 11 STIRNEMANN,I~, RIS, HB, Do, D, HKMMERLI, R. Intraoperative flow measurement of distal runoff: a valid predictor of outcome the presence of severe occlusive disease by detecting of infrainguinal bypass surgery. Eur J Surg 1994; 160: 431-436. the intense signal generated by non-pulsatile blood 12 ASCElb E, WHITE, SA, VEITH, FJ, MORIN, L, FREEMAN,K, GUPTA, flow. 23 Studies comparing MRA with contrast angioSK. Outflow resistance measurement during infrainguinal argraphy showed that MRA is able to detect patent terial reconstructions: a reliable predictor of limb salvage. Am J Surg 1987; 154: 185-188. runoff vessels that are not opacified by preoperative 13 LUNDELL,A, BERGQVIST,D. Prediction of early graft occlusion angiography, with a prevalence ranging from 19% to in femoropopliteal and femorodistal reconstruction by measure24% when the actual patency was confirmed by surment of volume flow with a transit time flowmeter and calculation of peripheral resistance. Eur J Vase Surg 1993; 7: 704-708. gical exploration or intraoperative contrast angioDONALDSON,MC, MANNICK, JA, WHITTEMORE, AD. Causes of graphy. 24-2s Therefore, we might expect further 14 primary graft failure after in situ saphenous vein bypass grafting. improvement in the correlation between the Ad Hoc J Vasc Surg 1992; 15: 113-120. runoff score and the outcome of infrainguinal re- 15 WOODBURN,KR, RUMLEY, A, LOWE, GD, LOVE, JG, MURRAY, GD, POLLOCK, JG. Clinical, biochemical, and rheologic factors vascularisation by detecting patent vessels with alaffecting the outcome of infrainguinal bypass grafting. J Vasc ternative imaging techniques. Surg 1996; 24: 639-646.
Acknowledgements The authors wish to thank Jukka Ollgren, MSc, for help with the statistical analysis. The study was supported by a grant from Finska L~ikaresallskapet, which is gratefully acknowledged. References 1 WALDEN, R, MODAN, M, BASS, A, SCHNEIDERMAN,J, ADAR, R, Scoring of vascular disease in the lower extremities. J Cardiovasc Surg 1989; 30: 210-215.
16 KAUFMAN,JL, WHITTEMORE,AD, CoucfI, NP, MANNICK,JA. The fate of bypass grafts to an isolated popliteal artery segment. Surgery 1982; 92: 1027-1030. 17 BEARD,JD, SCOTT,DJA, EVANS,JM, SKIDMORE,Rr HORROCKS,M. Pulse generated runoff: a new method of determining calf vessel patency. Br J Surg 1988; 75: 361-363. 18 SIMMS,HM, HARDMAN,J, SLANEY,G. The relevance of prebypass pedal arch assessment. Br J Surg 1984; 71: 381. 19 CAMPBELL,WB, FLETCHER,EL, HANDS, LJ. Assessment of the distal lower limb arteries: a comparison of arteriography and doppler ultrasound. Ann R Coll Surg Engl 1986; 68: 37-39. 20 Rlcco, JB, PEARCE, WI-I, YAO, JST, FLINN, WR, BERGAN,JJ. The use of operative bypass arteriography and doppler ultrasound recording to select patients for extended femoro-distal bypass. Ann Surg 1983; 198: 646-653. 21 PATEL,PR, SEMEL,L, CLAUSS,RH. Extended reconstruction rate for limb salvage with intraoperative prereconstruction angiography. J Vasc Surg 1988; 7: 531-537. Eur J Vasc Endovasc Surg Vol 15, March 1998
224
A. AIb~ick et aL
22 SCARPATO,R, GEMBAROWICZ,R, FARBER,Set al. Intraoperative occlusive disease: accuracy and reliability of time-of-flight MR prereconstruction arteriography. Arch Surg 1981; 116: 1053-1055. angiography. Radiology 1994; 192: 351-357. 23 MULLIGAN,SA, MATSUDA,T, LANZER,Pet al. Peripheral arterial 27 OWEN, RS, CARPENTER,JP, BAUM, RA, PERLOI% LJ, COPE, C. occlusive disease: prospective comparison of MR angiography Magnetic resonance imaging of angiographically occult runoff and color duplex US with conventional angiography. Radiology vessels in peripheral arterial occlusive disease. N Engl J Med 1991; 178: 695-700. 1992; 326: 1577-1581. 24 CARPENTER,JP, OWEN,RS, BAUM,RA et aL Magnetic resonance 28 CAILPENTEI~JP, GOLDEN,MA, BARKER,CE HOLLAND,GA, BAUM, angiography of peripheral runoff vessels. J Vasc Surg 1992; 16: RA. The fate of bypass grafts to angiographically occult runoff 807-815. vessels detected by magnetic resonance angiography. J Vasc Surg 25 AMERICANCOLLEGEOF RADIOLOGYNEW TECHNOLOGYASSESS1996; 23: 483-489. MENT GROUP. Multicenter trial to evaluate peripheral vascular magnetic resonance angiography. JAMA 1995; 274: 875-880. 26 McCAULEY,TR, MONIB,A, DICKEY,KW et al. Peripheral vascular Accepted 24 September 1997
Eur J Vasc Endovasc Surg Vol 15, March 1998