Background
Phase III RCT (GOG 120). 526 patients with FIGO stage IIB–IVA squamous, adenosquamous, or adenocarcinoma of the cervix with negative para-aortic lymph nodes. Three-arm trial comparing radiosensitizing chemotherapy regimens delivered concurrently with pelvic radiation (45 Gy + brachytherapy). Enrolled through the Gynecologic Oncology Group and Southwest Oncology Group.
Interventions and follow up
Arm A: Pelvic RT + cisplatin 40 mg/m² IV weekly (max 6 cycles)
Arm B: Pelvic RT + cisplatin 50 mg/m² day 1 + 5-FU 4 g/m² continuous infusion days 1–4 + hydroxyurea 2 g/m² oral twice weekly
Arm C: Pelvic RT + hydroxyurea 3 g/m² oral twice weekly (control)
Primary endpoint: Progression-free survival and overall survival
mFollow up: Initial report 35 months (Whitney JCO 1999); long-term update (Rose JCO 2007)
Arm B: Pelvic RT + cisplatin 50 mg/m² day 1 + 5-FU 4 g/m² continuous infusion days 1–4 + hydroxyurea 2 g/m² oral twice weekly
Arm C: Pelvic RT + hydroxyurea 3 g/m² oral twice weekly (control)
Primary endpoint: Progression-free survival and overall survival
mFollow up: Initial report 35 months (Whitney JCO 1999); long-term update (Rose JCO 2007)
Results
Relative risk of progression vs HU (Arm C): Cis+5FU+HU (Arm B): RR 0.57, P<.001; Cis weekly (Arm A): RR 0.51, P<.001
Relative risk of death vs HU: Arm B: RR 0.66, P=.004; Arm A: RR 0.61, P=.002
3-year PFS (Rose 2007 update): ~63% Cis only vs ~59% Cis combo vs ~47% HU
Relative risk of death vs HU: Arm B: RR 0.66, P=.004; Arm A: RR 0.61, P=.002
3-year PFS (Rose 2007 update): ~63% Cis only vs ~59% Cis combo vs ~47% HU
Adverse events
Main adverse events: Grade 3–4 hematologic toxicity significantly higher in Arm B (triple drug) vs Arm A (single agent cisplatin): 30% vs 18% grade ≥3 leukopenia. Cisplatin alone (Arm A) was better tolerated than the combination arm. Hydroxyurea alone (Arm C) had the lowest toxicity but worst efficacy.
Conclusions
Both cisplatin-based regimens (weekly cisplatin alone and cisplatin+5-FU+HU) produced significantly superior progression-free and overall survival compared with hydroxyurea alone, confirming concurrent cisplatin as the preferred radiosensitizer for locally advanced cervical cancer. Weekly cisplatin alone (40 mg/m²) achieved comparable efficacy to the triple-drug regimen with substantially less hematologic toxicity, establishing it as the preferred regimen.
Key Limitations
Key Limitations: Hydroxyurea-only comparator arm is no longer used, so the triple vs single cisplatin comparison is more clinically relevant. GOG 120 did not compare cisplatin alone vs no concurrent chemotherapy, which was addressed by RTOG 9001 and GOG 109. The Rose 2007 long-term update showed convergence of the two cisplatin arms, reinforcing weekly cisplatin as the simpler, less-toxic standard. Carboplatin-based or gemcitabine-containing regimens remain under investigation.
Clinical Context
GOG 120 established weekly cisplatin 40 mg/m² during pelvic RT as the preferred radiosensitizer for locally advanced cervical cancer — simpler and less toxic than the combination regimen with equivalent efficacy. This regimen remains standard of care globally. The study, along with RTOG 9001 and GOG 109, was among five concurrent chemoradiation trials that prompted the 1999 NCI Clinical Announcement advocating for cisplatin-based CRT. Immunotherapy (pembrolizumab) added to first-line concurrent CRT is now being evaluated (CALLA, ENGOT-cx11/KEYNOTE-A18).
References