A randomized phase III trial of four cisplatin (CIS) containing doublet combinations in stage IVB, recurrent, or persistent cervical carcinoma: A Gynecologic Oncology Group (GOG) study

Reviewer: Christine Hill, MD
Abramson Cancer Center of the University of Pennsylvania
Last Modified: June 1, 2008

Presenter: B. J. Monk
Presenter's Affiliation: University of California, Irvine, Orange, CA
Type of Session: Scientific


  • Cervical cancer represents the third most common gynecologic malignancy in the United States. Although screening via Pap smear has significantly decreased the prevalence of advanced and metastatic cervical cancer in the US, this is not the case worldwide.
  • The Gynecologic Oncology Group (GOG) has performed a series of randomized controlled trials investigating the optimal chemotherapy regimen for patients with advanced or recurrent cervical carcinoma.
    • Over the course of five GOG trials performed in the 1980s and 1990s, cisplatin was determined to be the most effective single-agent treatment for advanced/ recurrent cervical cancer.
    • GOG 169, a subsequent trial, was a randomized, phase III clinical trial that compared single-agent cisplatin to combination cisplatin and paclitaxel in this setting (Moor, 2004).
      • The addition of paclitaxel improved response rate (36% versus 19%, p = 0.002) and progression-free survival (4.8 months versus 2.8 months, p < 0.001), but did not impact median overall survival (9.7 versus 8.8 months, p = NS).
    • Another phase III trial, GOG 179, compared single-agent cisplatin to combination therapy with cisplatin and topotecan (Monk, 2005).
      • The addition of topotecan to cisplatin provided an overall survival benefit (median OS 9.4 versus 6.5 months, p = 0.017), as well as improvement in response rate and progression-free survival.
      • This combination of chemotherapy drugs was subsequently approved by the Food and Drug Association for treatment of advanced and recurrent cervical carcinoma.
  • The trial presented here, GOG 204, was initially opened in 2003, with the intent of comparing cisplatin/ paclitaxel combination chemotherapy to cisplatin/ vinorelbine in the setting of advanced cervical cancer.
  • After GOG 204 had begun accruing patients, the results of GOG 179 became available, demonstrating for the first time a statistically significant overall survival benefit with combination chemotherapy (cisplatin plus topotecan) when compared to single-agent cisplatin.
  • GOG 204 was at that time expanded to include two additional arms: cisplatin plus gemcitabine and cisplatin plus topotecan. The results of this four-armed, randomized, controlled trial intended to investigate the optimal cisplatin-based doublet for treatment of advanced cervical carcinoma are presented here.

Materials and Methods

  • Patients with stage IVB, recurrent, or persistent cervical carcinoma felt not to be amenable to cure were included in this study. Eligibility criteria included ECOG performance status of 0 or 1, acceptable end-organ function, and measurable disease.
  • Patients were randomized to one of four arms:
    • Paclitaxel (135 mg/m2 over 24 hours on day 1) + cisplatin (50 mg/m2 on day 2) every 3 weeks (reference arm) (PC).
    • Vinorelbine (30 mg/m2 on days 1 and 8) + cisplatin (50 mg/ m2 on day 1) every 3 weeks (VC).
    • Gemcitabine (1000 mg/m2 on days 1 and 8) + cisplatin (50 mg/ m2 on day 1) every 3 weeks (GC).
    • Topotecan (0.75 mg/m2 on days 1, 2, and 3) + cisplatin (50 mg/ m2 on day 1) every 3 weeks (TC).
  • An intent-to-treat analysis was planned, with a primary endpoint of 33% increase in overall survival.
  • In order to provide 84.5% power, a total of 600 patients (150 per arm) were planned to be enrolled.
  • An interim futility analysis was planned after approximately 58 deaths per arm.


  • Between May, 2003 and April, 2007, 513 patients were enrolled on this study. As noted previously, patients were initially randomized to PC versus VC. Upon availability of results from the GOG 179 trial in 2005, the trial was expanded and patients were randomized to receive PC, VC, GC, or TC.
    • The four arms were well-balanced, with median age in all four arms of approximately 50 years. Performance status was 0 in the majority of cases in all arms. The proportion of patients with grade 2 versus grade 3 disease was approximately 1:1 in all arms.
  • In April 2007, a planned interim futility analysis was performed, and recommended early closure of this study on the basis that none of the four arms was predicted to provide significant advantage over PC by the end of the study.
  • Tumor response rates were not significantly different across the four arms. Rates of complete response were 29.1%, 25.9%, 22.3%, and 23.4% for PC, VC, GC, and TC, respectively (p = NS).
  • No significant overall survival difference was demonstrated across the four treatment arms, although non-statistically significant benefit to PC was observed.
    • Median survival with PC was 12.9 months, versus 10 – 10.3 months in the other three study arms (p = NS).
    • Hazard ratios for death were 1.15, 1.32, and 1.26 with VC, GC, and TC when compared to PC (p = NS).
  • No significant difference in progression-free survival was demonstrated. Hazard ratios for disease progression were 1.35, 1.39, and 1.27 with VC, GC, and TC, respectively, when compared to PC (p = NS). Median progression-free survival was less than 6 months in all treatment arms.
  • Quality of life (QOL) was assessed before treatment, during treatment, and after completion of 6 cycles, with regard to physical well-being and functional well-being.
    • No significant differences in QOL were observed between the four arms; however, QOL appeared to demonstrate a non-statistically significant trend towards improvement with treatment with PC versus the remaining arms.
    • Similar results were seen when pain was assessed via the brief pain index; patients receiving PC had a non-statistically significant improvement in pain with treatment when compared to patients enrolled on other trial arms.
    • No difference in neurotoxicity was observed across the four arms.
  • Toxicities did not vary significantly across the four arms:
    • The lowest incidence of myelosuppression was seen in the PC arm, although this difference was not significant.
    • No difference in fatigue, myalgias, dermatitis, or other toxicities were observed between the four arms.
  • On multivariate analysis, poor performance status was identified as a poor prognostic factor (hazard ratio for death 1.80 for patients with ECOG performance status 1 versus 0). Disease recurrence in a previously irradiated region also portended poorer prognosis (hazard ratio for death 1.408 for patients with recurrent disease in an irradiated region versus those with elsewhere recurrence only). Finally, Hispanic race appeared to be a positive prognostic factor (hazard ratio for death 0.807). No difference in prognosis appeared present based on Caucasian or African-American race.

Author's Conclusions

  • The authors conclude paclitaxel/ cisplatin should remain the doublet of choice in the treatment of advanced or recurrent cervical cancer. 
  • The authors observed non-statistically significant trends in improved overall survival, tumor response, and quality of life with PC versus other doublets examined in this study.
  • They also note that ethnicity and performance status may be important prognostic factors for patients with advanced or recurrent cervical carcinoma.

Clinical/Scientific Implications

  • This trial represents a continuation of systematic exploration of conventional chemotherapies for the treatment of advanced and recurrent cervical cancer by the Gynecologic Oncology Group (GOG).
  • Unfortunately, no improvement in disease response was seen in any of the experimental arms; in fact, non-statistically significant benefits were seen in the reference arm consisting of paclitaxel and cisplatin.
  • Overall, treatment options for advanced and recurrent cervical cancer remain dismal, with median progression-free survival less than 6 months in all arms evaluated in this study, and median overall survival ranging from 10 months to one year.
    • Given these circumstances, all systemic chemotherapeutic options for treatment of advanced or recurrent cervical cancer should be viewed as palliative. 
    • Impact on quality-of-life should be considered during choice of any palliative regimen; although no significant difference in QOL was observed across the arms investigated in this study, QOL issues should certainly be considered in the decision to begin palliative treatment versus best supportive care.
  • The authors mention an upcoming GOG trial which is planned to randomize patients with recurrent/ advanced cervical carcinoma to PC +/- bevacizumab or topotecan/paclitaxel +/- bevacizumab. Results of treatment with the biologic agent bevacizumab will hopefully prove to be interesting and important.
  • The results of GOG 204 demonstrate minimal changes in tumor response or overall survival with multiple conventional chemotherapeutic regimens. As future trials are designed, the addition of novel biologic agents will hopefully provide benefit to this population of patients. To this end, potential enrollment on phase II trials examining these agents may allow more rapid access to novel agents for patients with advanced and recurrent cervix cancer, and may be considered as treatment possibilities in the future.

View Patient Summary