Phase III Trial of Carboplatin and Paclitaxel Compared With Cisplatin and Paclitaxel in Patients With Optimally Resected Stage III Ovarian Cancer: A Gynecologic Oncology Group Study

Reviewer: Ryan P. Smith, MD
The Abramson Cancer Center of the University of Pennsylvania
Last Modified: August 10, 2003

Authors: Ozols RF, Bundy BN, Greer BE, et al.
Source: JCO, Vol 21 (17), 1-7, 2003


  • Ovarian cancer is the fourth most common cancer in women and among the most aggressive
  • A large proportion of patients will present with advanced (stage III) disease, with spread within the peritoneal cavity
  • After optimal debulking, adjuvant chemotherapy has been shown to increase survival in these patients
  • The regimen of cisplatin and paclitaxel has been shown to be superior to other regimens, and is considered standard of care for the adjuvant treatment in ovarian cancer
  • Carboplatin, generally better tolerated than cisplatin, has been reported to have comparable activity when compared to cisplatin, though this has been questioned in patients with small volume disease (<1 cm of disease after surgery)
  • This study investigated the use of carboplatin plus paclitaxel chemotherapy in women with advanced ovarian cancer


  • 792 women were entered and randomized on the study
  • Patients were required to have pathologically verified epithelial ovarian cancer, with residual disease after surgical debulking of no more than 1.0 cm
  • Patients were balanced in terms of important prognostic factors (cell type, tumor grade, presence of residual disease)
  • Patients were randomized between two arms:
  • Arm 1 (standard): cisplatin 75 mg/m2 (1 mg/min) and paclitaxel 135 mg/m2 (over 24 hours) every 3 weeks for a total of six courses
  • Arm 2 (investigational): carboplatin AUC 7.5 and paclitaxel 175 mg/m2 (over 3 hours) every 3 weeks for a total of six courses
  • Patients were given G-CSF if they had treatment delays or the development of neutropenic fever
  • The study was run as a noninferiority trial (designed essentially to statistically exclude the possibility that the carboplatin arm is inferior to the cisplatin arm) with overall survival (OS) and progression free survival (PFS) as the endpoints evaluated in an intent to treat basis
  • 90% of patients were followed for at least 4 years or until their death


  • Approximately 55% had grade 3 tumors and 65% had residual disease after surgery
  • 85% of patients completed the cisplatin regimen and 87% completed the carboplatin regimen
  • Median doses of chemotherapy agents that patients received were almost exactly that which was designed
  • For the cisplatin arm, PFS was 24%, with a median PFS of 19.4 months. For the carboplatin arm, PFS was 27%, with a median PFS of 20.7 months (p=NS)
  • For the cisplatin arm, OS was 42%, with median OS of 48.7 months. For the carboplatin arm, OS was 47%, with a median OS of 57.4 months
  • Patients who had complete visual removal of tumor had longer survival times compared to those with residual disease. Within these groups, there was no effect of the chemotherapy regimen used
  • Median survival after recurrence was 23 months, with no difference between the two arms
  • Grade 3-4 toxicity profiles differed between the two arms: The cisplatin arm had a higher incidence of leukopenia (though with few admissions for neutropenic fever), GI toxicity (23% vs. 10%), and metabolic and genitourinary toxicity (though numbers for both of these areas were very small). The carboplatin arm had a higher incidence of thrombocytopenia (39% vs. 5%), though there were no reports of significant bleeding or the need for platelet transfusion.

Author's Conclusions

  • The combination of carboplatin and paclitaxel is not inferior to cisplatin and paclitaxel with regard to PFS and OS in patients with small-volume stage III epithelial ovarian cancer.
  • Carboplatin has a more favorable toxicity profile than cisplatin. Therefore, carboplatin plus paclitaxel is considered the preferred regimen for patients with ovarian cancer.


The vast majority of patients with ovarian cancer will require adjuvant chemotherapy. The standard had been cisplatin and paclitaxel, though based on the toxicity profiles, many physicians have been using carboplatin plus paclitaxel. This study demonstrates that the regimen of carboplatin plus paclitaxel is as efficacious as cisplatin plus paclitaxel. Though the study was not statistically designed to detect a benefit of carboplatin, the 16% risk of death noted in this study with carboplatin plus paclitaxel is intriguing. Based on the regimens used, patients received more paclitaxel in the investigational arm and likely received more platinum agent, with the use of carboplatin AUC 7.5. Though this may affect the outcomes, it also speaks to the fact that patients can tolerate higher doses with carboplatin. Both arms were fairly well tolerated, as the majority of patients completed all 6 cycles with the doses planned. From a toxicity profile, carboplatin is thought to be better tolerated, mainly from a renal and gastrointestinal perspective. This study continues to support this thought, though the actual numbers of patients that developed renal toxicity (as represented by the metabolic and genitourinary toxicities) were small. In addition, there was no information given on the use of G-CSF, which presumably was higher in the carboplatin arm, as it has a history of inducing greater myelosuppression. Perhaps the most striking figure that is derived from this important paper is the fact that over 70% of patients continue to relapse, and the vast majority of these relapsed patients will ultimately succumb to their disease. Although carboplatin and paclitaxel may now represent the standard of adjuvant chemotherapy in patients with ovarian cancer, obviously further investigations are required to improve on these outcomes.