Preservation of renal function with surgical revascularization in patients with atherosclerotic renovascular disease

OBJECTIVE: Clinical and anatomic factors predictive of a favorable response to renal revascularization performed for renal function salvage remain poorly defined. To clarify decision making in such patients we reviewed a contemporary experience of surgical renal artery revascularization (RAR) performed primarily for preservation of renal function. METHODS: Between June 1990 and March 2001 (ensuring 1 year minimum follow-up), 96 patients with renal insufficiency (serum creatinine [Cr] concentration >or=1.5 mg/dL) and hypertension underwent RAR for preservation of renal function. Study end points included early and late renal function response, progression to dialysis dependence, and long-term survival. Variables potentially associated with these end points were assessed with univariate analysis, Cox regression analysis, and logistic regression analysis, and survival was assessed with the Kaplan-Meier method. RESULTS: Perioperative failure of RAR occurred in 3 patients (3%), with perioperative mortality in 4 patients (4.1%); thus 92 patients were available for long-term follow-up (mean, 39 months). Mean patient age was 70 +/- 9 years with a mean baseline Cr of 2.6 mg/dL (range, 1.5-7.8 mg/dL). Operative management consisted of aortorenal bypass in 38% of patients, extraanatomic bypass in 38% of patients, and endarterectomy in 24% of patients; 32% of patients required combined aortic revascularization and RAR, and 27% underwent bilateral RAR. At hospital discharge renal function had improved (20% decrement in Cr) in 41 (43%) patients, including 7 patients who were removed from dialysis; remained unchanged in 40 (41%) patients; and declined (20% increase in Cr) in 15 (16%) patients. At last follow-up renal function was either improved or unchanged in 72% of patients. Predictors of a favorable response to RAR at last follow-up included stable Cr at hospital discharge (odds ratio [OR], 7.1; 95% confidence interval [CI], 2.5-21.8; P =.0004) and decreased Cr at hospital discharge (OR, 16; 95% CI, 1.6-307.8; P <.0001); bilateral renal artery repair (OR, 3.1; 95% CI, 0.9-10.6; P =.07) approached clinical significance. Predictors of worsened excretory function at last follow-up included decline of renal function at hospital discharge (OR, 28.9; 95% CI, 5.0-165.4; P =.0002), intervention to treat unilateral renal artery stenosis (OR, 3.8; 95% CI, 0.8-16.6; P =.05), and level of baseline Cr (OR, 3.0; 95% CI, 1.0-4.0; P =.04). Progression to dialysis occurred in 16 (17%) patients. Dialysis-free survival at 5 years was 50%, and overall actuarial survival at 5 years was 59%. Predictors of progression to dialysis during follow-up included treatment of complete renal artery occlusion (OR, 6.2; 95% CI, 1.3-29.5; P =.02), early failure of RAR (OR, 3.1; 95% CI, 0.7-14.2; P =.04) and baseline Cr (OR, 2.9; 95% CI, 1.3-6.1; P =.006). CONCLUSION: Long-term clinical success in the preservation of renal function, noted in 70% of patients, can be predicted by the initial response to RAR and by anatomic factors, in particular, bilateral repair. While extreme (mean Cr >or=3.2 mg/dL) renal dysfunction generally is predictive of poor long-term outcome, a subset of patients will experience favorable results, even to the extent of rescue from dialysis. These results may facilitate clinical decision making in the application of RAR for renal function salvage.