Long-term immune reconstitution and final 1-year follow-up after fixed-duration venetoclax-obinutuzumab (VenO) in first-line (1L) chronic lymphocytic leukemia (CLL): Results from the Phase III CRISTALLO trial

Abstract

Abstract Background: Venetoclax + an anti-CD20 antibody has shown deep remissions and long-term immune recovery in patients (pts) with relapsed/refractory CLL (Kater et al. Hemasphere 2024). Immunological effects of this combination in 1L CLL are not fully characterized. The CRISTALLO Phase III trial compared the efficacy and safety of fixed duration (FD) VenO vs fludarabine-cyclophosphamide-rituximab/bendamustine-rituximab (FCR/BR) in fit pts with treatment (Tx)-naïve CLL (Sharman et al. ASH 2024). Immune dynamics during and after VenO Tx and correlation of immune recovery patterns with minimal residual disease (MRD) status were explored. Methods: Pts were randomized 1:1 to VenO or FCR/BR. The primary endpoint was undetectable MRD (uMRD; <10-4) rate at Month 15 in peripheral blood (PB). Key secondary endpoints included progression-free survival. Immune profiling was performed at baseline, during Tx, and every 3 months (mos) up to 48 mos post-Tx in pts stratified by MRD status at Month 15 (<10-4) in PB. A flow cytometry panel was developed to identify non-malignant B cells (CD3/CD24/CD56-CD19+ and excluding CD19+CD20dim/negCD5+CD23+CD79bdim/neg) vs CLL cells (CD19+CD20dim/negCD5+CD23+CD79bdim/neg) in PB. Serial measurements included lymphocyte subsets (CD4+/CD8+ T cells and NK cells), immunoglobulin (Ig) levels (IgG, IgA, IgM), and detailed B-cell compartment analysis. Results: Overall, 80 and 86 pts were randomized to VenO and FCR/BR, respectively; 81.3% (VenO) and 54.7% (FCR/BR) achieved uMRD at Month 15. Non-malignant B-cell and CLL cell profiling was performed for 80 pts at baseline and on-Tx in the VenO arm, and 84 pts at baseline in the FCR/BR arm. Lymphocyte subsets (n=72) and Ig profiles (n=77) were assessed pre- and post-VenO up to 48 mos follow-up (FU). At baseline, no significant differences in levels of CLL cells, non-malignant B cells, CD4+ T cells, CD8+ T cells, NK cells, and Igs were observed between Tx arms in both uMRD and MRD+ groups. Analysis of CD19+ cells (capturing CLL and non-malignant B cells) during VenO Tx showed a marked depletion at 12 mos FU. In the uMRD group, CLL cells remained suppressed up to 36 mos FU (median: 3.9 CLL cells/µL). In contrast, in the MRD+ group, CLL cells were not fully depleted at 12 mos FU (median: 28.9 CLL cells/µL) and rebounded rapidly, showing a 10-fold increase by 18 mos FU (median: 315.1 CLL cells/µL). Non-malignant B cells were relatively less depleted with Tx vs CLL cells (median: 66.9 and 51.8 cells/µL in the uMRD and MRD+ groups, respectively). Notably, the MRD+ group showed a faster recovery of non-malignant B cells over time, with median counts rising to 373.3 cells/µL at 24 mos FU vs 127.9 cells/µL in the uMRD group. However, recovery was partial and did not return to baseline in the uMRD group (1127.7 cells/µL). Similar to B-cell recovery kinetics, NK cells showed gradual immune reconstitution after transient depletion at 6 mos FU in the VenO arm. However, recovery of CD4+ and CD8+ T cell counts was delayed and did not return to baseline by 48 mos FU, irrespective of MRD status. Ig levels, including IgG, IgM, and IgA, showed a significant but transient decline after VenO Tx, with the lowest levels seen at 3 mos FU, followed by signs of recovery from 9 mos FU. IgM levels were 0.22 g/L at baseline (n=63), 0.14 g/L at 3 mos FU (n=13), 0.19 g/L at 12 mos FU (n=55), 0.315 g/L at 24 mos FU (n=48), and 0.335 g/L at 36 mos FU (n=28), mirroring non-malignant B-cell recovery patterns. Consistent with the facilitated non-malignant B-cell recovery in the MRD+ group, median IgG, IgA, and IgM levels were all numerically higher in the MRD+ vs the uMRD group across all time points. Additional assessment on the impact of Igs on grade ≥3 infection will be presented. At final data cutoff (19 Mar 2025; median FU: 41 mos), fewer pts had progressed/died with VenO vs FCR/BR (14 vs18; p=0.399). There were no new deaths since the primary analysis and no new safety signals were identified with VenO. Conclusions: CRISTALLO demonstrated that FD VenO causes differential effects on malignant and non-malignant B-cell populations, with sustained CLL cell depletion in pts achieving uMRD alongside B-cell recovery, delayed T-cell recovery and effective Ig reconstitution. Findings extend previous observations and confirm that immune reconstitution can occur while maintaining deep remissions. Furthermore, final analysis results continue to support the efficacy and safety of FD VenO.