Weekly induction intraperitoneal chemotherapy after primary surgical cytoreduction in patients with advanced epithelial ovarian cancer

Intraperitoneal (IP) therapy is the direct instillation of chemotherapy into the peritoneal cavity. First proposed by Dedrick et al, [1, 2] IP arm is designed to maximize drug delivery to the tumor while avoiding many of the systemic toxicities associated with intravenous (IV) administration of the drug. It has been hypothesized previously that the high local concentration of cisplatin achievable within the peritoneal cavity after regional administration (10- to 20-fold greater than measured within the systemic compartment) would exert its maximum benefit in patients with microscopic residual disease or very small-volume macroscopic cancer at the time of IP drug delivery [3]. Support for this concept comes from early preclinical data where the depth of penetration of cytotoxic agents directly into the tumor or normal tissue after regional delivery has been measured in millimeter or less from the surface of the peritoneal lining [2, 4‐6]. A number of phase II trials of IP therapy have examined in the salvage setting in women with ovarian cancer have revealed that responses are almost exclusively observed in individuals with microscopic disease only or in those individuals whose maximal tumor diameter measures less than 0.5 to 1 cm [7, 8]. Mature results are also available from two randomized phase III trials conducted by the Gynecologic Oncology Group (GOG) (GOG104, GOG 114) [9, 10]. A landmark study, initiated in the mid 1980s and published in 1996, compared a regimen of IV cyclophosphamide plus IV cisplatin to a regimen of IV cyclophosphamide plus IP cisplatin [9]. In this trial, the median survival time of patients treated with IP cisplatin was improved from 41 to 49 months (P = 0.02) and the death HR was reduced by 24%, in favor of the regional treatment program. Although the authors claimed less toxicity observed in IP drugs delivery, 25% patients received IP therapy less than 4 cycles. In addition, a criticism of this trial is that 73% patients with residual disease ≤ 0.5 cm did not statistically benefit from IP therapy, in other words, just a cohort of 27% women with residual disease between 0.5–2 cm benefited most from this regional treatment program. In the second study (GOG 114), Markman et al [10], compared the new standard regimen of IV cisplatin plus IV paclitaxel to a regimen of moderately high-dose carboplatin followed by IP cisplatin plus IV paclitaxel. They documented the median survival time for women treated with IP therapy was improved from 52 to 63 months, and the death HR was reduced by 19%. One may argue a marginally significant survival improvement (P = 0.05) amounted to a remarkable 11-month prolongation. The results from that trial were challenged on the basis of the unknown impact of 2 cycles of carboplatin that was used before initiation of the IP therapy, concern about 18% of women received cycles of IP therapy. An editorial by McGuire [11] proclaimed that, in the second trial, savage therapies that might influence survival end points were not controlled, and how frequently salvage therapies administered before disease progression that might also influence progression-free survival was not divulged by the authors.

These results, however, do not lead to the widespread practice of IP therapy. Why? Controversies remain both over safety and efficacy of this approach. Consequently, there is likely considerable room for improvement with respect to study design. Traditional IP therapy was delivered simultaneously with IV chemotherapy and repeated every three weeks. With much different to that in the literature, this series was carried out to evaluate the role of weekly IP therapy delivered independently with IV chemotherapy.

We retrospectively reviewed the records of 426 patients who underwent primary surgery for advanced epithelial ovarian cancer between 1990 and 1999 at Fudan University Cancer Hospital. Patients received postoperative IP therapy were compared with those without any IP therapy. IP therapy repeated every week was followed by systemic IV chemotherapy. Follow-up data were available for 409 patients who formed the basis of this study.

Regular IP therapy used independently with IV chemotherapy began from 2–7 days after surgery. Two hundred thirty patients (56.2%) received 2–9 courses (median, 4 courses) of postoperative IP therapy. One hundred and five women (25.7%) received 1–3 courses (mean, 3 courses) of neoadjuvant IP therapy, which was delivered before primary surgery and not considered as an inducing intervention. IP therapies delivered after systemic front-line intravenous chemotherapy such as those for consolidation or salvage intent were not included in this study. IP arm consisted of platinum-based combination therapy administered through a Tenckhoff, or a Port A, or Groshong catheter. Cisplatin, 60 mg/m², and 5-FU, 800–1000 mg/m², and/or mitomycin C 6 mg/m², or etoposide, 100 mg/m², were given in 1.5–2 liters of normal saline and administered IP as fast as possible, usually within 2 hrs. Hydration was given with 1000–1500 ml 5% glucose in 4 hrs.

The cut-off of residual disease in the greatest dimension after primary surgery was 1 cm. Statistical analysis included tests for associations between potential prognostic factors and between prognostic factors and survival. Survival probabilities were estimated by Kaplan-Meier methods, and prognostic factors for survival were evaluated by a log-rank test (univariate) or a Cox proportional hazard model (multivariate).

The present results suggest that weekly IP delivery of chemotherapy for 4 cycles should be a standard care for ovarian cancer. Well-designed randomized controlled trials should be conducted to confirm the results.

This work supported in part by the New Star Project of the Shanghai Health Bureau (R.Y.Z.).