Phase III trial of four cisplatin-containing doublet combinations in stage IVB, recurrent, or persistent cervical carcinoma: a Gynecologic Oncology Group study

Abstract

Purpose: Assess toxicity and efficacy of cisplatin (Cis) doublet combinations in advanced and recurrent cervical carcinoma.

Patients and methods: Patients were randomly assigned to paclitaxel 135 mg/m(2) over 24 hours plus Cis 50 mg/m(2) day 2 every 3 weeks (PC, reference arm); vinorelbine 30 mg/m(2) days 1 and 8 plus Cis 50 mg/m(2) day 1 every 3 weeks (VC); gemcitabine 1,000 mg/m(2) day 1 and 8 plus Cis 50 mg/m(2) day 1 every 3 weeks (GC); or topotecan 0.75 mg/m(2) days 1, 2, and 3 plus Cis 50 mg/m(2) day 1 every 3 weeks (TC). Survival was the primary end point with a 33% improvement relative to PC considered important (85% power, alpha = 5%). Quality-of-life data were prospectively collected.

Results: A total of 513 patients were enrolled when a planned interim analysis recommended early closure for futility. The experimental-to-PC hazard ratios of death were 1.15 (95% CI, 0.79 to 1.67) for VC, 1.32 (95% CI, 0.91 to 1.92) for GC, and 1.26 (95% CI, 0.86 to 1.82) for TC. The hazard ratios for progression-free survival (PFS) were 1.36 (95% CI, 0.97 to 1.90) for VC, 1.39 (95% CI, 0.99 to 1.96) for GC, and 1.27 (95% CI, 0.90 to 1.78) for TC. Response rates (RRs) for PC, VC, GC, and TC were 29.1%, 25.9%, 22.3%, and 23.4%, respectively. The arms were comparable with respect to toxicity except for leucopenia, neutropenia, infection, and alopecia.

Conclusion: VC, GC, and TC are not superior to PC in terms of overall survival (OS). However, the trend in RR, PFS, and OS favors PC. Differences in chemotherapy scheduling, pre-existing morbidity, and toxicity are important in individualizing therapy.

Fig 1.

CONSORT flow diagram. Initially, 41 patients were registered onto a randomized phase III trial comparing two arms, paclitaxel + cisplatin (PC) versus vinorelbine + cisplatin (VC). On January 26, 2004, the trial was amended to include two additional arms: gemcitabine + cisplatin (GC) and topotecan + cisplatin (TC). The initial 41 patients were excluded from the primary analysis. Additional analyses with these patients and with all of the ineligible patients yielded the same qualitative conclusions with regard to treatment efficacy. prog, progression.

Fig 2.

(A) Overall survival Kaplan-Meier plots for the 434 patients in the study sample and (B) hazard ratios with 95% CIs adjusted for multiplicity, using Dunnett's procedure. (C) Progression-free survival Kaplan-Meier plots for the 434 patients in the study sample and (D) hazard ratios with 95% CIs adjusted for multiplicity, using Dunnett's procedure. Cis + Pac, cisplatin + paclitaxel; Cis + Vin, cisplatin + vinorelbine; Cis + Gem, cisplatin + gemcitabine; Cis + Top, cisplatin + topotecan; Rel Haz, relative hazard; Var, Variance; ln, natural logarithm; HR, hazard ratio.

Fig 3.

(A) Hazard ratios for overall survival with 95% CIs adjusted for treatment regimen and disease status, investigating the potential prognostic impact of race (black not statistically significant), performance status (patients with performance status = 1 indicate worse prognosis), disease site (patients with target lesion in a previously irradiated zone indicate worse prognosis), and ethnicity (patients of Hispanic origin indicate improved prognosis). (B) Hazard ratios of overall survival for disease status with 95% CIs, adjusted for treatment, race, ethnicity, performance status, and disease site. The status of disease was classified into three categories: advanced disease, recurrent disease, and persistent disease. Persistent disease was classified as evidence of disease following initial therapy. Recurrent disease was further broken down by the progression-free interval since the date of diagnosis. All hazard ratios are relative to patients with advanced disease. Rel Haz, relative hazard; Var, Variance; ln, natural logarithm; HR, hazard ratio; PFI, progression-free interval.