Gene therapy with viral vectors has shown some promise in nude mice models and in initial Phase I trials of patients with extensive metastatic cancer. A Phase I clinical trial (D. L. Tait et al., Clin. Cancer Res., 3: 1959-1968, 1997) of ovarian cancer patients treated with i.p. retroviral LXSN-BRCA1sv gene therapy reported stable vector, minimal antibody response, and tumor reduction. We initiated a Phase II trial on patients with less extensive disease to evaluate vector pharmacokinetics, immune response, toxicity, and efficacy. Patients received a surgically implanted peritoneal catheter to administer infusions of vector, as well as to retrieve daily samples of peritoneal fluid for analysis. Ovarian cancer patients received four daily i.p. injections of LXSN-BRCA1sv vector therapy for three cycles, 4 weeks apart. Patient peritoneal fluid and plasma were analyzed extensively by PCR, Western blot, complement level (CH50), and chemical and hematological tests. Phase II patients showed no response, no disease stabilization, and little or no vector stability. Because of vector instability and rapid antibody development, which differed dramatically from the Phase I trial data, the trial was terminated after treatment of six patients. Immune system status appears to have played a major role in whether gene therapy was effective. Comparison of Phase I and II patients showed significant differences in tumor burden, immune system status, and response to BRCA1 gene therapy.