A post hoc safety analysis of fixed-duration acalabrutinib-venetoclax combinations vs chemoimmunotherapy: Results from the phase 3 AMPLIFY trial

Abstract

Abstract The AMPLIFY trial assessed fixed-duration (FD) acalabrutinib + venetoclax (AV) or AV + obinutuzumab (AVO) vs fludarabine + cyclophosphamide + rituximab/bendamustine + rituximab (FCR/BR) in patients (pts) with treatment-naive (TN) chronic lymphocytic leukemia (CLL); at a median follow-up of 40.8 mo, AV and AVO yielded prolonged progression-free survival (PFS) vs FCR/BR and manageable (Brown et al. NEJM 2025;392:748-62). This post hoc analysis evaluated the safety of FD AV, AVO, and FCR/BR. AMPLIFY (ACE-CL-311; NCT03836261) is an ongoing, randomized, open-label, phase 3 trial of pts with TN CLL aged ≥18 y with ECOG performance status ≤2 and without del(17p) or TP53 mutation. Pts were randomized for treatment (tx) with AV (acalabrutinib [A], cycles [C] 1–14; venetoclax [V], C3–14), AVO (AV as described; obinutuzumab, C2−7), or chemoimmunotherapy (FCR/BR, C1−6). Exposure-adjusted event rates (EAERs; events/100 person-mo) were calculated for tx-emergent adverse events (AEs) of clinical interest (ECI), including cardiac events, neutropenia, hypertension, infections, and tumor lysis syndrome (TLS). Cumulative incidences of cardiac events and infections were analyzed by Kaplan–Meier (KM) analysis (descriptive P-values). TLS risk at baseline and C3 (post-debulking/pre-V) was evaluated (high risk=lymph node [LN]≥10 cm, or absolute lymphocyte count [ALC]≥25×109/L and LN≥5 cm; medium risk=LN≥5–<10 cm or ALC ≥25×109/L; low risk=LN<5 cm and ALC <25×109/L). Time and cause of early tx discontinuation or death (all tx or any study drug in the regimen) were analyzed; pts who completed tx were censored. PFS and time-to-next-treatment (TTNT) rates, including impact of dose modification (DM), were assessed using KM methods. PFS and TTNT were analyzed as time from end of tx (14 mo). Pts who did not complete tx were excluded. In total, 834 pts received ≥1 dose of AV (n=291), AVO (n=284), or FCR/BR (n=259). Median (range) duration of A exposure was 12.9 mo (1–18) in AV and 12.9 mo (0–18) in AVO; tx exposure was 5.6 mo (1–11) in FCR/BR. Incidences of any-grade ECIs were 76.3% (AV), 85.2% (AVO), and 71.4% (FCR/BR) (P=0.229 and P<0.0001 for AV and AVO vs FCR/BR, respectively); EAERs were lower with AV (25.3) and AVO (36.1) vs FCR/BR (57.8). Incidence of cardiac events was higher with AV (9.3%) and AVO (12.0%) vs FCR/BR (3.5%; P≤0.006, both comparisons); EAERs were similar across arms (0.83, 1.11, and 0.86, respectively). Any-grade atrial fibrillation, ventricular tachyarrhythmias, and hypertension EAERs were similar across arms (AV, 0.05, 0.05, 0.30; AVO, 0.16, 0.08, 0.32; FCR/BR, 0.14, 0, 0.57, respectively). Any-grade neutropenia EAERs were lower for AV (5.4) and AVO (8.0) vs FCR/BR (17.2). Any-grade infection incidences were higher with AV (50.9%) and AVO (53.9%) vs FCR/BR (31.7%; P<0.0001, both comparisons); EAERs were similar across arms (6.47, 8.38, and 8.77, respectively). Cumulative infection incidence was similar for AV and AVO vs FCR/BR (HR 0.75, 95% CI 0.56–1.01 and HR 0.84, 95% CI 0.63–1.12, respectively). Most pts (83%) with high TLS risk at baseline (AV, n=93; AVO, n=75; FCR/BR, n=86) shifted to medium (60.2%, 21.3%, 8.1%) or low (19.4%, 61.3%, 77.9%) TLS risk at C3. TLS in the AV arm (n=1; during V ramp-up) and the AVO arm (n=1; during A lead-in) was limited to laboratory TLS. Discontinuation of tx or death occurred in 8.4% (AV), 5.1% (AVO), and 18.6% (FCR/BR); time-to-event (discontinuation or death) analysis favored AV and AVO vs FCR/BR (HR 0.07, 95% CI 0.03–0.15; P<0.0001; HR 0.16, 95% CI 0.07–0.30; P<0.0001, respectively). AEs were the most common reason for discontinuation of any tx (AV, 7.9%; AVO, 20.1%; FCR/BR, 10.8%). Any AE leading to dose withholding of any tx occurred in 49.8% (AV), 64.8% (AVO), and 31.3% (FCR/BR) of pts; AEs leading to dose reduction of any tx occurred in 14.1% (AV), 20.8% (AVO), and 11.2% (FCR/BR) of pts. At 24 mo from the end of tx, the estimated PFS rate was 77.0% with DM (n=125) and 80.1% without DM (n=127) in the AV arm; in the AVO arm, the 24-mo estimated PFS rate was 91.2% with DM (n=124) and 94.4% without DM (n=93). The same pattern was seen for TTNT rates. Overall, no new safety signals with FD AV and AVO were identified. EAERs for infections and cardiac events were similar across all arms. Over 80% of pts at high TLS risk at baseline transitioned to low/medium risk with initial tumor debulking. While DM was largely due to AEs, these did not seem to impact efficacy of AV or AVO.