The purpose of this Phase I study was to determine the feasibility of using an anti-erbB-2-encoding adenovirus (Ad21) to treat erbB-2-overexpressing ovarian cancer. Recurrent ovarian cancer patients were treated i.p. with Ad21 in dosages ranging from 1 x 10(9) to 1 x 10(11) pfu. Patients were monitored after treatment for evidence of clinical toxicity and efficacy. Peritoneal aspirates and serum samples were obtained to assess for evidence of gene transfer/expression, for generation of wild-type vector, and antiadenoviral humoral response. Fifteen patients were treated per study specifications. Treatment-specific grade 1/2 fever was experienced by 9 of 15 (60%) patients. Other transient grade 1/2 constitutional, pain, and gastrointestinal symptoms were also experienced. No dose-limiting vector-related toxicity was experienced. Of 13 patients evaluable for response, 5 (38%) had stable disease and 8 (61%) had evidence of progressive disease. One patient with nonmeasurable disease normalized her CA125 at the 8-week evaluation, and one patient with nonmeasurable disease remained without clinical evidence of disease for 6 months after treatment. PCR analysis of peritoneal aspirates demonstrated the presence of Ad21 in 84.6%, 84.6%, and 61.6% of evaluable specimens at days 2, 14, and 56 after treatment, respectively. No wild-type virus was detected. Reverse transcription-PCR analysis demonstrated expression of the anti-erbB-2 sFv-encoding gene in 10 of 14 evaluable patients at day 2. Five of six evaluable patients had an increase in antiadenovirus antibody titer. This study suggests that adenoviral-mediated gene therapy using an anti-erbB-2-directed intrabody is feasible in the context of human ovarian cancer.