BACKGROUND: The course of advanced renal cell carcinoma is extremely variable, ranging from spontaneous remission to disease progression refractory to chemotherapy. Immunotherapy has held promise of improved outcomes based on uncontrolled studies and randomized controlled trials generally limited by small size and low power. OBJECTIVES: To evaluate immunotherapy for advanced renal cell carcinoma by comparing: (1) high dose interleukin-2 to other options and (2) interferon-alfa to other options. The primary outcome of interest was overall survival at one year, with remission as the main secondary outcome of interest. SEARCH STRATEGY: A systematic search of the CENTRAL, MEDLINE, and EMBASE databases was conducted for the period 1966 through end of December 2003. Handsearches were made of the proceedings of the periodic meetings of the American Urologic Association, the American Society of Clinical Oncology, ECCO - the European Cancer Conference, and the European Society of Medical Oncology for the period 1995 to June 2004. SELECTION CRITERIA: Randomized controlled trials that selected (or stratified) patients with advanced renal cell carcinoma, utilized an immunotherapeutic agent in at least one study arm, and reported remission or survival by allocation. Fifty-three identified studies involving 6117 patients were eligible and all but one reported remission; 32 of these studies reported the one-year survival outcome. DATA COLLECTION AND ANALYSIS: Two reviewers independently abstracted each article by following a prospectively designed protocol. Dichotomous outcomes for treatment remission (partial plus complete) and for deaths at one year were used for the main comparisons. Survival hazard ratios were also used for studies of interferon-alfa versus controls, and for two randomized studies of the value of initial nephrectomy prior to interferon-alfa in fit patients with metastases detected at the time of diagnosis. MAIN RESULTS: Combined data for a variety of immunotherapies gave an overall chance of partial or complete remission of only 12.9% (99 study arms), compared to 2.5% in 10 non-immunotherapy control arms, and 4.3% in two placebo arms. Twenty-eight percent of these remissions were designated as complete (data from 45 studies). Median survival averaged 13.3 months (range by arm, 6 to 27+ months). The difference in remission rate between arms was poorly correlated with the difference in median survival so that remission rate is not a good surrogate or intermediate outcome for survival for advanced renal cancer. We were unable to identify any published randomized study of high-dose interleukin-2 versus a non-immunotherapy control, or of high-dose interleukin-2 versus interferon-alfa reporting survival. It has been established that reduced dose interleukin-2 given by intravenous bolus or by subcutaneous injection provides equivalent survival to high dose interleukin-2 with less toxicity. Results from four studies (644 patients) indicate that interferon-alfa is superior to controls (OR for death at one year = 0.56, 95% confidence interval 0.40 to 0.77). Using the method of Parmar 1998, the pooled overall hazard ratio for death was 0.74 (95% confidence interval 0.63 to 0.88). The weighted average median improvement in survival was 3.8 months. T he optimal dose and duration of interferon-alfa remains to be elucidated. The addition of a variety of enhancers, including lower dose intravenous or subcutaneous interleukin-2, has failed to improve survival compared to interferon-alfa alone. Two recent randomized studies have examined the role of initial nephrectomy prior to interferon-alfa therapy in highly selected fit patients with metastases at diagnosis and minimal symptoms: despite minimal improvement in the chance of remission, both studies of up-front nephrectomy improved median survival by 4.8 months over interferon-alfa alone. Recent studies have been examining anti-angiogenesis agents. A landmark study of bevacizumab, an anti-vascular endothelial growth factor antibody, was associated with significant prolongation of the time to progression of disease when given at high dose compared to low-dose or placebo therapy though frequency of remissions or survival were not improved. AUTHORS' CONCLUSIONS: interferon-alfa provides a modest survival benefit compared to other commonly used treatments and should be considered for the control arm of future studies of systemic agents. In fit patients with metastases at diagnosis and minimal symptoms, nephrectomy followed by interferon-alfa gives the best survival strategy for fully validated therapies. The need for more effective specific therapy for this condition is apparent.