From the Society for Vascular Surgery

Predictors of major amputation despite patent bypass grafts Ann D. Smith, MD, MPH,a Alexander T. Hawkins, MD, MPH,a Maria J. Schaumeier, MD,a Marit S. de Vos, BS,b Michael S. Conte, MD,c and Louis L. Nguyen, MD, MBA, MPH,a,b Boston, Mass; and San Francisco, Calif Objective: Despite patent vein bypass grafts, some patients with critical limb ischemia (CLI) receive major amputations. We analyzed the predictive factors leading to major amputation in the presence of patent lower extremity bypass (LEB) grafts. Methods: Data from the Project of Ex-Vivo vein graft Engineering via Transfection III (PREVENT III), a large prospective randomized trial of 1404 patients who underwent LEB with vein graft for CLI, were queried for outcomes. The primary outcome was major amputation with patent (PMA) LEB compared with patients with patent LEB who achieved limb salvage (PLS). The population excluded those who received amputation for occluded grafts. A Cox proportional hazard model identified independent predictors. Results: Of 1404 LEB patients, 162 (11.5%) had major amputation: 89 (6.3%) with patent and 73 (5.2%) with occluded LEB. For PMA, 21 of 89 (23.6%) developed critical stenosis and 11 of 21 (52.4%) were revised. For PLS, 460 of 1242 (37.0%) developed critical stenosis and 351 of 460 (76.3%) were revised. Predictive patient factors included having preoperative gangrene (vs rest pain; hazard ratio [HR], 3.504; 95% confidence interval [CI], 1.533-8.007; P [ .0029), diabetes (HR, 1.800; 95% CI, 1.006-3.219; P [ .0477), black (vs white) race (HR, 1.779; 95% CI, 1.051-3.011; P [ .0321), baseline creatinine clearance 65 mL/min; HR, 1.759; 95% CI, 1.016-3.048; P [ .0439), prior history of coronary artery bypass grafting (HR, 1.702; 95% CI, 1.080-2.683; P [ .0221), and lower baseline activity quality of life score (HR, 1.401; 95% CI, 1.105-1.778; P [ .0054). Postoperative wound factors included gangrenous changes (HR, 5.830; 95% CI, 1.647-20.635; P [ .0063), surgical wound necrosis (HR, 5.319; 95% CI, 1.478-19.146; P [ .0105), deep (vs superficial) wound infection (HR, 3.815; 95% CI, 1.220-11.927; P [ .0213), and wound healing abnormally (HR, 3.754; 95% CI, 1.061-13.278; P [ .0402). Associated postoperative consequences leading to PMA included having recurrent CLI symptoms (HR, 2.915; 95% CI, 1.816-4.681; P < .0001), a severe (vs mild) adverse event (HR, 2.751; 95% CI, 1.391-5.443; P [ .0036), fewer percutaneous revisions (HR, 2.425; 95% CI, 1.573-3.740; P < .0001), discharge on low-molecular-weight heparin (HR, 2.087; 95% CI, 1.309-3.326; P [ .0020), and decreasing days to critical stenosis/occlusion/revision/amputation (HR, 1.010; 95% CI, 1.007-1.012; P < .0001). Conclusions: Whereas a patent vein graft is important to all vascular surgeons, additional factors should be considered in trying to attain limb salvage for patients with CLI. These factors include intervening surgically before CLI has progressed to a state of gangrene or limited activity and optimizing nutrition, diabetes control, cardiac conditions, and activity level. Revision offers hope for clinical improvement but may be delayed when there is no graft lesion identified. The absence of a graft lesion to revise may also portend amputation despite a patent graft because of nongraft-related factors such as infection. Finally, the experience of a severe (vs mild) adverse event may also result in limb loss despite a patent graft. Systematic efforts to reduce severe adverse events among patients may also lead to increased limb salvage. (J Vasc Surg 2016;-:1-10.) From the Center for Surgery and Public Health, Department of Surgery,a and Division of Vascular and Endovascular Surgery,b Brigham and Women’s Hospital, Harvard Medical School, Boston; and the Division of Vascular and Endovascular Surgery, University of California, San Francisco, San Francisco.c This research was supported by the Arthur Tracy Cabot Fellowship and the Lea du Pont Research and Education Endowment. A.T.H. is also supported by the NIH NHLBI T32 (HL007734) Harvard/Longwood Vascular Surgery Training Program. Author conflict of interest: none. Presented at the 2015 Vascular Annual Meeting of the Society for Vascular Surgery, San Francisco, Calif, May 30-June 1, 2013. Correspondence: Louis L. Nguyen, MD, MBA, MPH, Division of Vascular and Endovascular Surgery, Brigham and Women’s Hospital, Harvard Medical School, 75 Francis St, Boston, MA 02115 (e-mail: llnguyen@ partners.org). The editors and reviewers of this article have no relevant financial relationships to disclose per the JVS policy that requires reviewers to decline review of any manuscript for which they may have a conflict of interest. 0741-5214 Copyright Ó 2016 by the Society for Vascular Surgery. Published by Elsevier Inc. http://dx.doi.org/10.1016/j.jvs.2015.10.101

Despite having patent lower extremity bypass (LEB) vein grafts after open surgery, some patients with critical limb ischemia (CLI) still do not achieve limb salvage. In studies that discuss limb salvage in CLI patients, strategies generally involve medical management or procedural intervention with the assumption that a patent vessel or graft will treat the disease. However, there is a subset of patients who experience clinical failure after a procedure despite patent vein bypass grafts. The Project of Ex-Vivo vein graft Engineering via Transfection III (PREVENT III) study was a large, multi-institutional, randomized, double-blinded, placebocontrolled trial of 1404 patients with CLI to test the drug edifoligide for the prevention of vein graft failure in LEB grafts.1 Patients in this study underwent autologous LEB and had a 1-year follow-up period. Because the study showed no difference between the treatment arm and the placebo arm, we are able to analyze the entire cohort in a post hoc analysis to assess predictors of clinical failure that led to amputation despite patent grafts. 1

2 Smith et al

Prior studies have identified the problem of amputation despite patent grafts.2-10 Patients who received amputations despite patent grafts did not experience technical graft failure. Instead, we describe their experience as one of clinical failure. Some small studies had a low volume of patients with amputation despite patent grafts and so describe each patient’s reason for amputation (progression of soft tissue infection/necrosis,2 progressive ischemia and septic shock9). Larger studies have been based on singlecenter or observational data sets and offer predictors for amputation despite patent grafts. Prior predictors offered have included black race,2 diabetes,3,5 dialysis,8 distal targets,2,5 infection,3,10 late revision,2 poor runoff,7,10 preoperative inability to ambulate independently,8 renal failure,2 tissue loss,3,5 and poor wound healing.10 Other studies have found that diabetes2 and dialysis5 were explicitly not predictors for amputation despite patent bypass graft. Our study aimed to clarify the reasons that patients with patent grafts receive amputations by studying a large, multicenter population of patients with CLI who received close follow-up postoperatively for 1 year. METHODS Database and patient selection. We analyzed data from the PREVENT III trial, a prospective, multicenter, double-blinded, randomized controlled trial performed to assess the efficacy of a drug, edifoligide, in preventing graft failure in autologous vein grafts after open LEB surgery in patients with CLI.1 Because edifoligide was found to have no effect on primary outcomes of the surgery, we were able to use the entire cohort to study patients who underwent LEB with autologous vein for CLI. However, we always included edifoligide treatment in all multivariable models to ensure that there were no additional effects that were unaccounted for. Edifoligide remained not statistically significant in multivariable models. The PREVENT III database included 1404 patients who were observed for 1 year postoperatively and received surveillance ultrasound examinations to assess graft patency. The primary outcome was the time to occurrence of nontechnical graft failure requiring revision or amputation. Secondary outcomes included all-cause graft failure (including technical failure), nontechnical primary patency, freedom from clinically significant index graft stenosis, amputation, and index graft failure-free survival.1 Patients were eligible for the PREVENT III study if they were at least 18 years old and were scheduled for LEB for CLI using autologous vein. CLI was defined as having gangrene, nonhealing ischemic ulcer, or ischemic rest pain. Patients were excluded if they were undergoing “revisions of preexisting infrainguinal bypass grafts, reconstructions that used any synthetic or nonautologous component, or vein grafts attached to a nonautologus graft below the femoral bifurcation.”1 Notably, being on dialysis was not an exclusion criterion for PREVENT III. Postoperative duplex ultrasound examinations were performed on PREVENT III patients at 1, 3, 6, and 12 months after LEB. A 9-month ultrasound examination

JOURNAL OF VASCULAR SURGERY --- 2016

was added if a “flow abnormality was observed at 6 months.”1 Additional follow-up studies included anklebrachial index and the Vascular Quality of Life questionnaire.11 Criteria were established for graft reintervention, which included angiographic stenosis of >70%, anklebrachial index 300 cm/s.1 Follow-up quality of life (QoL) scores at 3 months and beyond were excluded from analysis as they were missing from >50% of those with patent LEB grafts who received major amputation (PMA) and >30% of those with patent LEB grafts who achieved limb salvage (PLS). Our study identified predictors for amputation among PREVENT III patients who had patent grafts. We excluded patients with occluded grafts who received amputation from our analysis, resulting in a study cohort of 1331 patients. Time to event was either time to amputation or, in the event of limb salvage, time to death or end of study. Patients were censored for death. Two variables of concern in this study included critical stenosis without revision (SWR) and high-risk grafts. Of the vein grafts used in this study, 24% were considered high risk, defined as a vein with a diameter 60% because of missing data in complete case analysis. An a level of .05, corresponding to P ¼ .05 and confidence interval (CI) of 95%, was used as the criterion for statistical significance. All data manipulation and statistical analyses were performed using Statistical Analysis System (SAS) 9.3 software (SAS Institute, Cary, NC).

JOURNAL OF VASCULAR SURGERY Volume -, Number -

Smith et al 3

Table I. Characteristics of patients with a patent graft and limb salvage (PLS) and a patent graft with major amputation (PMA) Variable Patient descriptors Mean age, years Sex Male Female Race White Black Other Baseline weight, kg Medical history History of MI History of stroke History of CABG Smoking status Never Former or current Hyperlipidemia Hypertension Diabetes Insulin-dependent diabetes Dialysis Renal insufficiency Creatinine clearance >65 mL/min Creatinine clearance 25-65 mL/min Creatinine clearance