Optimising MRD-guided treatment: UMRD4 as a half-way treatment timepoint for ibrutinib-venetoclax substantially delays MRD relapse after end of treatment in both IGHV-mutated and IGHV-unmutated CLL.

Abstract Introduction: The FLAIR (ISRCTN01844152) phase III trial for previously untreated fit CLL patients showed that measurable residual disease (MRD)-guided ibrutinib-venetoclax (I+V) significantly improved response, progression-free survival (PFS), and overall survival (OS) compared to ibrutinib-alone (IBR) or fludarabine-cyclophosphamide-rituximab (FCR). Outcomes vary according to immunoglobulin heavy chain (IGHV) gene mutation status. This analysis aims to determine the optimal approach for guiding treatment duration, and to report MRD relapse kinetics based on IGHV mutation status. Methods: Participants were randomly assigned (1:1:1) to FCR, or MRD-guided IBR or I+V. The primary endpoint for I+V vs. IBR was undetectable MRD (uMRD) in the bone marrow (BM) within 2 years. The primary endpoint for I+V vs. FCR and a powered secondary endpoint for I+V vs. IBR was PFS. Other secondary endpoints included OS, the proportion of participants with uMRD4 (<0.01%) at 9 months and longitudinally, pattern of MRD relapse and retreatment, response to therapy (per IWCLL criteria), safety, health-related quality of life, and cost-effectiveness. IGHV mutation status was defined as unmutated (U-CLL), mutated (M-CLL), or subset #2 (S#2-CLL) according to ERIC criteria. MRD was assessed using ERIC flow cytometry in peripheral blood (PB) every 6 months. MRD-guided end-of treatment (MRD-EoT) occurred at twice the time taken to achieve confirmed uMRD4 (defined as first PB uMRD4 sustained for six months confirmed with BM uMRD4), minimum 2 years (yrs) maximum 6yrs of treatment. MRD relapse was defined as >0.01% disease in PB. Results: MRD response and relapse kinetics are available for 1102 participants: 579 FCR, 260 I+V and 263 IBR (not discussed further as there were no BM uMRD4 responses with IBR). In the I+V arm, PB uMRD4 was reached by 23% (44/188) at 6 months and 45% (97/214) at 9 months. Of these, 80% (35/44 and 78/97, respectively) have MRD-EoT at 2yrs. Reaching PB uMRD4 by 9 months was more frequent in U-CLL (52%, 53/102) or S#2-CLL (47%, 7/15), compared to M-CLL (36%, 28/78); within these groups, MRD-EoT was ≤3yrs for 91% (48/53) U-CLL, 86% (6/7) S#2-CLL, and 75% (21/28) M-CLL. High (>1%) MRD levels at 6 months did not preclude MRD-guided cessation (24% [8/34] have MRD-EoT ≤3 years), but if disease remained >1% at 9 or 12 months from randomisation then <10% (2/22 and 0/17 respectively) have MRD-EoT ≤3yrs. MRD relapse kinetics also varied by IGHV status. For U-CLL, 51% (147/291) achieved BM uMRD4 with FCR, versus 75% (92/123) with I+V (MRD-EoT 2yrs=72%, 3yrs=16%, >3yrs=12%). Within uMRD4 responders, the proportion sustaining PB uMRD4 at 3yrs from MRD-EoT was 58% (95% CI 50-65%) for FCR vs. 79% (95% CI 68-90%) for I+V. Median duration of PB uMRD4 was 41 months (95% CI 38-51) for FCR vs. not reached for I+V (median follow-up from EoT: FCR 80 months vs I+V 31 months). Similar outcomes were seen for S#2-CLL: 44% (14/32) BM uMRD4 responses with FCR (median PB uMRD4 duration 41 months, 95% CI 28-53) for FCR vs. 63% (10/16) with I+V with no MRD relapses to date. For M-CLL, 56% (114/203) achieved BM uMRD4 with FCR, compared to 53% (51/97) with I+V (MRD-EoT 2yrs=76%, 3yrs=14%, >3yrs=10%). Within uMRD4 responders, the proportion sustaining PB uMRD4 at 3yrs from MRD-EoT was 70% (95% CI 62-78%) for FCR vs. 93% (95% CI 84-100%) for I+V. Median duration of sustained PB uMRD4 was not reached in either arm. In the FCR arm, the MRD relapse rate steadily increases over 6yrs for U-CLL/S#2 CLL but plateaus in M-CLL between 3-6yrs from MRD-EoT. Median time from MRD relapse to clinical progression was 31 months (95% CI: 27-36) for U-CLL, 36 months (95% CI: 17-39) for S#2-CLL, and 74 months (95% CI: 50-NR) for M-CLL. There are insufficient MRD relapses in the I+V arm to determine if a plateau occurs. Conclusions: MRD-guided I+V improves outcomes, and a simplified approach using PB MRD at 2-3 timepoints is sufficient to optimize treatment duration. The MRD relapse rate at 3yrs post MRD-guided I+V for all IGHV subgroups is lower than in IGHV-mutated CLL treated with FCR, possibly due to deeper remissions or slower doubling times. The plateauing of MRD recurrence in ~40% of IGHV-mutated patients treated with FCR suggests many may be effectively cured. Further follow-up is needed to confirm a similar or higher rate for MRD-guided I+V, but early data are encouraging.