LACK OF RETROPERITONEAL LYMPHADENOPATHY PREDICTS SURVIVAL OF PATIENTS WITH METASTATIC RENAL CELL CARCINOMA

Abstract

Patients with metastatic renal cell carcinoma have a reported 5-year survival of 0% to 20%. The ability to predict which patients would benefit from nephrectomy and interleukin-2 (IL-2) therapy before any treatment is initiated would be useful for maximizing the advantage of therapy and improving the quality of life. A retrospective analysis of the x-rays and charts of patients treated at the National Institutes of Health Surgery Branch between 1985 and 1996, who presented with metastatic renal cancer beyond the locoregional area and the primary tumor in place, was performed. Preoperative computerized tomography or magnetic resonance imaging, or radiological reports if no scans were available, were used to obtain an estimate of the volume of retroperitoneal lymphadenopathy. Operative notes were used to evaluate whether all lymphadenopathy was resected or disease left in situ, or if any extrarenal resection, including venacavotomy, was performed. Mean survival rate was calculated from the time of nephrectomy to the time of death or last clinical followup. If patients received IL-2 therapy, the response to treatment was recorded. Mean survival and response rate for IL-2 were compared among patients in 3 separate analyses. Patients without preoperatively detected lymphadenopathy were compared with those with at least 1 cm.3 retroperitoneal lymphadenopathy. Also, the patients who had detectable lymphadenopathy were divided into subgroups consisting of all resected, incompletely resected, unresectable and unknown if all disease was resected. Each subgroup was compared with patients without detectable preoperative lymphadenopathy. Patients with less than were compared to those with greater than 50 cm.3 retroperitoneal lymphadenopathy. Patients undergoing extrarenal resection at nephrectomy (complex surgery) due to direct invasion of the tumor into another intra-abdominal organ were compared with those undergoing radical nephrectomy alone, regardless of lymph node status. Statistical analysis was done with the Mantel-Cox test for comparison of survival on Kaplan-Meier curves and with Fisher's exact test for response rates for IL-2. A total of 154 patients with metastatic renal cell carcinoma underwent cytoreductive nephrectomy as preparation for IL-2 based regimens. There were 82 patients with metastatic renal cell carcinoma and no preoperative retroperitoneal lymphadenopathy who survived longer (median 14.7 months) than the 72 with lymphadenopathy (median 8.5, p = 0.0004). Patients with incompletely resected, unresectable or an unknown volume resected had decreased survival compared to those with no retroperitoneal lymphadenopathy. A multivariate analysis of survival was performed evaluating the known prognostic indicators, performance status and tumor burden, as represented by the number of organs involved with metastases, and the new prognostic factor, lymphadenopathy. Lymphadenopathy was more closely associated with survival than performance status, and appeared to be a new prognostic variable. Patients with and without retroperitoneal lymphadenopathy at initial presentation had similar rates for treatment with IL-2 (54% for both groups). Of the 82 patients with no lymphadenopathy 11 (13%) had long-term survival greater than 5 years. Of the 6 complete responses to IL-2, 5 occurred in this group. Only 1 other patient with incompletely resected retroperitoneal lymphadenopathy survived longer than 5 years. No significant difference in survival was seen between patients who did or did not undergo complex surgery. Patients who presented with metastatic renal cancer and retroperitoneal lymphadenopathy had a shorter survival than those with no detectable retroperitoneal lymphadenopathy. It is warranted to continue to perform complex extrarenal resection during nephrectomy since no significant difference in the response rate for IL-2 or mean survival compared with those of patients undergoing nephrectomy alone is currently detectable.