Multidrug resistance (MDR) in a variety of human tumors such as renal cell carcinoma (RCC) is thought to be caused by expression of the mdr1 gene and may be reversed by applying chemosensitizers such as Dexverapamil that inhibit the mdr1 gene product P-glycoprotein. On the basis of our preclinical analysis, we initiated a clinical (GCP) study with vinblastine (VBL), the most effective--if at all--chemotherapeutic agent; dexverapamil; and dexamethasone in patients with RCC. All patients had histologically proven RCC that was metastatic and progressive at study entry. The statistical design featured a preliminary study of two cycles of VBL alone followed by tumor evaluation. If no response was documented, with all patients thus serving as their own control, dexverapamil and dexamethasone were added for a minimum of three cycles of combination therapy. Having obtained institutional permission by the ethical review committee (MEC 124, 106-1993/12), we enrolled 24 patients on this protocol starting on May 3, 1993. In the preliminary study, 1 complete response (CR) was achieved with VBL alone, and myelotoxicity led to an adequate dose reduction from 2 mg/m2 VBL per day given as a 5-day continuous infusion (days 1-5) in 6/10 yet evaluable patients to 1.4 mg/m2 per day. In 8/11 yet evaluable patients, dexverapamil doses reached > or = 3000 mg/day by 7-day oral uptake (days 0-6, supported by 20 mg dexamethasone given twice daily), which is significantly higher than those previously reported. The combination of VBL given at 1.4 mg/m2 per day plus, dexverapamil given at 3000 mg per day was felt to be safe and well tolerated. Nine patients were yet evaluable for response. One partial response and three minor responses were noted in this heavily pretreated study population. It appears that this innovative approach may have some activity in RCC and may eventually lead to a rational treatment modality. Careful evaluation in ongoing studies is warranted.