Résumé
Context: R-CHOP is the standard of care for newly diagnosed DLBCL patients; however, 40% remain uncured. Polatuzumab vedotin (Pola), a CD79b-targeting antibody-drug conjugate, has shown promising activity/safety in a Phase Ib/II study (Tilly, et al. Lancet Oncol 2019). We report the Phase III double-blind, placebo-controlled, international POLARIX study (NCT03274492), which compared Pola-R-CHP with R-CHOP in treatment-naïve DLBCL patients with an International Prognostic Index score of 2–5. Design: Patients were randomized (1:1) to six cycles of Pola-R-CHP or R-CHOP and on Day 1 of each cycle received Pola 1.8mg/kg or vincristine 1.4mg/m2, plus intravenous rituximab 375mg/m2, cyclophosphamide 750mg/m2, and doxorubicin 50mg/m2. Patients also received oral prednisone 100mg once daily (Days 1–5) and two further cycles of rituximab. The primary endpoint was investigator-assessed progression-free survival (PFS). Results: Overall, 879 patients (median age 65 [range 19–80] years) were randomized (Pola-R-CHP: n=440; R-CHOP: n=439). At primary data cut-off (28 June 2021), median follow-up was 28.2 months. PFS was superior with Pola-R-CHP vs R-CHOP (hazard ratio [HR] 0.73; 95% confidence interval [CI]: 0.57–0.95; P=0.02) and 2-year PFS rate was improved (76.7% [95% CI: 72.7–80.8] vs 70.2% [95% CI: 65.8–74.6], respectively). Investigator-assessed event-free survival favored Pola-R-CHP vs R-CHOP (HR 0.75; 95% CI: 0.58–0.96; P=0.02) and overall survival was comparable (HR 0.94; 95% CI: 0.65–1.37; P=0.75). While independent review committee-assessed end-of-treatment complete response rate by positron emission tomography-computed tomography was not significantly different with Pola-R-CHP (78.0%) vs R-CHOP (74.0%; P=0.16), disease-free survival suggested more durable responses with Pola-R-CHP vs R-CHOP (HR 0.70; 95% CI: 0.50–0.98). Safety profiles were similar for Pola-R-CHP vs R-CHOP: grade 3–4 adverse event (AE) rates, 57.7% vs 57.5%, respectively; serious AEs, 34.0% vs 30.6%; grade 5 AEs, 3.0% vs 2.3%; AEs leading to dose reduction, 9.2% vs 13.0%; and peripheral neuropathy, any grade, 52.9% vs 53.9%. At data cut-off, fewer patients treated with Pola-R-CHP (23%) vs R-CHOP (30%) had received ≥1 subsequent anti-lymphoma therapy. Conclusions: As the first-line treatment of DLBCL, Pola-R-CHP demonstrated a 27% reduction in the relative risk of disease progression, relapse, or death, and a similar safety profile, compared with R-CHOP.