A randomized phase 3 trial of Gemcitabine or Nab-paclitaxel combined with cisPlatin as first-line treatment in patients with metastatic triple-negative breast cancer

Abstract

Platinum is recommended in combination with gemcitabine in the treatment of metastatic triple-negative breast cancer (mTNBC). We conduct a randomized phase 3, controlled, open-label trial to compare nab-paclitaxel/cisplatin (AP) with gemcitabine/cisplatin (GP) in mTNBC patients (ClinicalTrials.gov NCT02546934). 254 patients with untreated mTNBC randomly receive AP (nab-paclitaxel 125 mg/m² on day 1, 8 and cisplatin 75 mg/m² on day 1) or GP (gemcitabine 1250 mg/m² on day 1, 8 and cisplatin 75 mg/m² on day 1) intravenously every 3 weeks until progression disease, intolerable toxicity or withdrawal of consent. The primary endpoint is progression-free survival (PFS); secondary endpoints are objective response rate (ORR), safety and overall survival (OS). The trial has met pre-specified endpoints. The median PFS is 9.8 months with AP as compared to 7.4 months with GP (stratified HR, 0.67; 95% CI, 0.50-0.88; P = 0.004). AP significantly increases ORR (81.1% vs. 56.3%, P < 0.001) and prolongs OS (stratified HR, 0.62; 95% CI, 0.44-0.90; P = 0.010) to GP. Of grade 3 or 4 adverse events, a significantly higher incidence of neuropathy in AP and thrombocytopenia in GP is noted. These findings warrant further assessment of adding novel agents to the nab-paclitaxel/platinum backbone due to its high potency for patients with mTNBC.

Conflict of interest statement

The authors declare no competing interests.

© 2022. The Author(s).

Figures

Fig. 1. Consort diagram of the GAP study.

Patients may have discontinued the assigned treatment for multiple reasons. This figure shows the numbers of patients included in the intention-to-treatment and safety analyses.

Fig. 2. Kaplan–Meier estimates of survival curves for AP and GP groups.

Kaplan–Meier plot of PFS (A) and OS (B). Data for the ITT population. Treatment effects were compared using the log-rank test; HRs and corresponding 95% CIs were estimated using Cox-proportional-hazard regression model. P values are one-sided with no adjustment for multiplicity. HR = hazard ratio.

Fig. 3. Subgroup analysis of PFS.

Data are median for ITT population in clinically relevant subgroups. Subgroup included patients with n = 45) and ≥40 years (n = 208); ECOG = 0 (n = 40) and ECOG = 1 (n = 213); premenopausal status (n = 113) and postmenopausal status (n = 140); de novo stage IV (n = 48), disease-free survival <1 year (n = 29) and ≥1 year (n = 176); No. of metastatic organ site = 1 (n = 94), =2 (n = 81) and ≥3 (n = 78); visceral metastasis (n = 163) and no visceral metastasis (n = 90); previous treatment with anthracycline (n = 173) and with no anthracycline (n = 80); previous treatment with taxane (n = 170) and with no taxane (n = 83). HRs and corresponding 95% CIs were estimated using Cox-proportional-hazard regression model. Data are presented as HR and 95% CI. HR hazard ratio, ECOG Eastern Cooperative Oncology Group.