The use of intravenous or subcutaneous immunostimulants in addition to chemotherapy has not improved the response rate or survival of ovarian cancer patients. The GOG used melphalan with or without subcutaneous administration of Corynebacterium parvum in a randomized trial, and there was no difference between the two groups. Another trial used combination chemotherapy with cisplatin, doxorubicin, and cyclophosphamide with or without BCG (Bacillus Calmette-Guerin) by scarification in patients with suboptimal ovarian cancer, and there was no difference between the two groups. This finding was confirmed in a study by SWOG. The intraperitoneal administration of C. parvum did produce complete responses in some patients with minimal residual disease at second-look laparotomy, suggesting that the regional administration of an immunostimulant, similar to the intralesional injection of melanoma with BCG, might be an effective way to control small amounts of residual cancer confined to the peritoneal cavity.

Based on the response seen with regional nonspecific immunotherapy, several trials of intraperitoneal cytokines have been performed. In a GOG study, recombinant human IFN-α exhibited activity when administered i.p. to patients with minimal residual ovarian cancer. Administration of 25-50 million units three times a week was not tolerated because of persistent general malaise, fever, and gastrointestinal toxicity. Treatment with the same dose once a week was tolerated for 8-16 consecutive weeks. Notably, there was an absence of significant neurotoxicity and renal toxicity. While most of the side effects of single agent recombinant IFN-α seem to complement those of cisplatin, the general malaise and gastrointestinal toxicity produced by both could potentially be additive when these agents are combined. Similar results were reported in another trial of IP IFN-α in the Netherlands. The toxicity encountered in this trial was similar to that seen in the Phase I GOG trial. In a follow-up Phase II trial conducted by the GOG, there was a 28% surgically-documented response rate in 25 patients with platinum-sensitive minimal residual tumors. Overall, 53 surgically evaluated patients have been treated in three trials, with 40% responses and 25% complete responses. All of the responding patients had microscopic or small macroscopic residual disease. Thus, the use of high-dose IP recombinant IFN-α given frequently can result in the regional control of very small-volume disease confined to the peritoneal cavity. Whether survival is affected is not known. Other cytokines have been evaluated. A Phase I-II trial of intraperitoneal interferon-γ also yielded a 30% response rate. Another phase I trial with intraperitoneal IL-2 has been performed, and the optimal dose was defined.

Preclinical studies have examined the interaction of cytokines, such as IFN-a, tumor necrosis factor, and IL-2, with various cytotoxic agents, including cisplatin and doxorubicin Even low doses of cytokines can induce synergistic activity with standard cytotoxic drugs against fresh ovarian cancer cells in vitro. Italian investigators reported a surgically-documented complete response rate of 50% when ovarian cancer patients received weekly treatment IP with cisplatin, alternating with human IFN-α (50 x 10₆ units). Responses were confined to those patients who started their treatment with minimum residual disease. Toxicity was similar to that seen with IFN-α alone. Therefore, the combination of IP cisplatin and IFN-α appeared to be tolerated and resulted in an appreciable response rate. Since response rates were similar to those reported in other studies of single-agent IP cisplatin, it was unclear whether the addition of IFN-α to the cisplatin had any significant benefit. When the GOG conducted a study of combined IP IFN-α and cisplatin, a very low partial response rate of 7% was seen, contrary to the results of other Phase I-II trials of IP cisplatin and IFN-α in patients with residual small-volume ovarian cancer, in which response rates of 20% to 40% have been noted. In the GOG study, however, most evaluable patients had extensive carcinomatosis that was cisplatin-resistant and the maximum tumor diameters were > 1 cm. In a follow-up Phase II study of IFN-α alternating with cisplatin in patients with platinum-sensitive minimal residual disease, the surgically-documented response rate was 28%. In a randomized Phase III trial, Windbichler and colleagues have demonstrated that the use of subcutaneous interferon-γ in women receiving first-line platinum-based chemotherapy in ovarian cancer was well-tolerated. A higher complete clinical response rate (68% vs 56%) and longer disease progression-free survival (48 months vs 17 months, p = .031) was observed in women receiving the interferon-γ plus chemotherapy than in women who were treated with chemotherapy alone. Improved progression-free survival was seen in patients with residual disease greater than 2 cm, as well as in those with optimal resection.

There has been interest in the use of adoptive immunotherapy in ovarian cancer. In a Phase I/II trial performed at the National Cancer Institute, IL-2 and lymphokine-activated killer (LAK) cells were administered intraperitoneally to 22 patients with ovarian cancer, many of whom had minimal residual disease confined to the peritoneal cavity. In this trial, the toxicity was high, especially with the development of extensive peritoneal fibrosis. There were, however, eight surgically documented partial responses (27%). Improvements in the specificity of this therapy, as well as the development of methods to reduce regional toxicity, could make adoptive immunotherapy an applicable technique in these patients.

Monoclonal antibodies raised against antigens on the surface of the epithelial ovarian cells have been used in clinical trials. Most of these antibodies have been linked to radiolabeled agents, such as ₁₂₅I or ₁₈₆Rh, and have been used for radioimmunoscintigraphy. Epenetos and colleagues and others, have used the intraperitoneal injection of ₁₃₁I-labeled monoclonal antibody directed toward placental alkaline phosphatase labeled and human milk fat globulin protein (HMFG). Intraperitoneal administration of ₉₀Y-labeled anti-HMFG protein significantly prolonged survival relative to historical controls. A multicenter randomized controlled study has compared IP treatment with ₉₀Y-labeled antibodies to no additional therapy in patients without evidence of disease following second-look laparoscopy. Accrual has been completed, but results have not yet been analyzed. Other investigators have evaluated ₁₂₅I-labeled OC125 antibody directed toward the CA125 antigen, and have shown good radiolocalization in patients with advanced bulky disease.355, 356 Herceptin (trastuzumab) is an antibody that reacts with the tyrosine kinase growth factor receptor HER-2, a receptor that is overexpressed in 30% of breast cancers. In patients with recurrent breast cancer, trastuzumab has been shown to improve the response rate to paclitaxel and doxorubicin. Trials of this antibody in ovarian cancers that overexpress HER-2 are ongoing. Only 15% to 30% of ovarian cancers overexpress HER-2 and a response of 12% has been observed in a Phase II trial of trastuzumab. Consequently, relatively few ovarian cancer patients may benefit from this approach.

The therapeutic potential of an anti-idiotypic response to an anti-CA125 antibody was discovered in a retrospective analysis of a diagnostic study with MAb-B43.13, where a large group of patients who had been imaged enjoyed unexpectedly long survival. When the clinical and immunologic profiles of 60 patients exposed to labeled MAb-B43.13 were compared to those of a contemporaneous historical cohort of 247 patients, those in the immunoscintigraphy group were 2.7 times less likely to die from ovarian cancer than were controls who had not been imaged (p < .001). Median survival time was doubled from 30 months in controls to 59 months in the immunoscintigraphy group. A randomized multicenter study compared vaccination with MAb-B43.13 (Ovarex) to placebo in 342 women with stage III/IV ovarian cancer who were disease free at the conclusion of frontline treatment. Overall, there was no difference in time to relapse, but in a subpopulation of 99 women defined by more successful response to frontline surgery and chemotherapy, time to relapse was extended from 10.3 months in the placebo group to 20.2 months in those who received MAb-B43.3 (p = .029). A prospective study has been planned to confirm the efficacy of MAb-B43.3 in this more optimal population.

Hormonal Therapy
Epithelial ovarian tumors frequently contain elevated levels of estrogen and androgen receptors. Some patients with epithelial ovarian tumors have had responses to endocrine therapy, although it appears that the overall response rate is approximately 10% to 20%. Progestational agents and estrogen antagonists, such as tamoxifen, have been used either singly or in combination with cytotoxic chemotherapy in patients with advanced disease. Other studies have used gonadotropin agonists,for example. leuprolide acetate alone or with tamoxifen, or aromatase inhibitors. In a Gynecologic Oncology Group study, patients with advanced epithelial ovarian cancer were treated with tamoxifen and a 13% partial response rate was reported. There is a lack of correlative studies comparing the response to endocrine manipulations and the presence of hormone receptors. Hormonal therapy for patients with advanced ovarian cancer should primarily be reserved for those who have failed chemotherapeutic regimens and who are not candidates for aggressive salvage drug regimens or investigative approaches. These agents are well tolerated and in circumstances where quality of life issues are paramount, such therapies can be employed.