Phase II, single-arm, open-label, multicenter study: Efficacy of adjunctive Bruton's tyrosine kinase inhibitor (BTKi) zanubrutinib and chimeric antigen receptor (CART) in aggressive B-cell non-hodgkins lymphoma (aNHL)

Abstract

Abstract Background: Resistance to CART has been linked to T-cell exhaustion with suppressive effects on the tumor microenvironment (TME), resulting in poor outcomes in aNHL with progression within 6 months. By contrast, survival is improved in pts with complete response (CR) beyond 6 months (mo) post CART with 12-mo duration of CRs of ~ 62% and PFS of 65% (Kamdar et al., Lancet 2022). Maintenance post CART may enhance CART function and result in more durable CRs. Zanubrutinib (Z) can reverse T-cell exhaustion and trigger apoptosis of lymphoma cells. To explore whether Z may augment CART efficacy in aNHL, we conducted a single-arm, phase 2 multi-institutional study administering Z as lead-in (ZLI) prior to apheresis followed by maintenance therapy (ZM) in pts with CR or PR post CART. Methods: Adult R/R aNHL pts who received CD19 CART as ≥ 2nd line were treated from 2021-2024. ZLI 180 mg bid was administered for 7-14 days prior to apheresis. Bridging was allowed. Pts with PR/CR on D29 PET/CT post CART received ZM180 mg bid x 13 cycles (12 mo) and were evaluable for primary endpoint. Pts with SD or PD were taken off study. The primary endpoint was the day 180 (D180) CR rate by 2014 Lugano criteria. Using a one-sided exact binomial test, we evaluated whether the observed D180 CR rate with adjunctive Z exceeded a historical benchmark of 44%, derived from prior CART studies. The null hypothesis would be rejected if ≥14/23 pts (69%) achieved a CR at D180. Secondary endpoints were PR to CR conversion, PFS, OS, immune subset changes with ZLI and ZM (using a high-dimensional spectral flow cytometry panel), minimal residual disease (MRD by Signatera assay), and toxicity. PFS and OS were estimated using Kaplan-Meier method; subgroup comparisons were done using log-rank tests or Cox regression. Significant changes in immune subsets were analyzed with two-way ANOVA followed by multiple-comparison correction using Benjamini-Hochberg adjustment (p<0.05). MRD analyses are underway. Results: 32 pts were treated; 23 were response evaluable. Median age was 61 (range 52-71), 61% were male, 100% had ECOG 0-1, 74% had advanced stage, 13% had elevated LDH, 35% were non-GCB. Histologies included de novo DLBCL (43%), transformed indolent NHL (48%) THRLBCL (9%); 13% had double hit (DHL). 48% received bridging, 52% received CART 2nd line and 48% ≥3rd line; 78% received axi-cel and 22% liso-cel. Median length of ZLI was 7 days (range 7-10). On D29 post CART, 7 pts had PR (30%) and 16 (70%) CR; all received ZM. Median number of ZM cycles was 10 (range 5-13); with ZM, 3 PR (43%) pts converted to CR by D90. At D180, CR rate was 70%, exceeding the prespecified threshold to meet our primary endpoint. The 12- and 18-mo duration of CR were 79 and 68% and 12 and 18-mo duration of PR were 57 and 38%, respectively. Median follow-up was 17.3 mo. 18-mo PFS and OS were 64 and 83%, respectively (6 mo post completion of ZM). Elevated LDH, non-GCB, histology, 2nd vs 3rd line CART, did not impact PFS or OS. Compared to baseline, pts with 1 week of ZLI showed increased circulating CD8+ T cells but lower frequencies of CD4+ Treg, memory B cells and noncanonical monocytes. Terminally differentiated TEMRA (effector memory T cells re-expressing CD45RA) cells and exhausted-like cells (PD1+CD39+) in gamma delta (γδ) T cells were reduced. PD-L1 expression on MDSCs was also decreased. This is in line with the potential immunomodulatory effects of ZLI on T cell exhaustion and suppressive cells. With 3 mo of ZM, pts with higher frequency of central memory but lower frequency of TEMRA cells in CD8+ T cells were more likely to maintain CR beyond 6 mo post CART. With ZLI (n=31), 15 pts (48%) had treatment related adverse events (TRAEs), mostly hematologic and gastrointestinal [G3, 13% - neutropenia (n=1), lymphopenia (n=1)]. With ZM (n=23), 20 pts (87%) had TRAEs (G3-G4 60%) with no deaths. G3-4 events in ≥5% of pts were lymphopenia (26%), neutropenia (26%), leukopenia (17%), infection (9%); G3 A-flutter occurred in 1 pt. Conclusions: In pts with R/R aNHL, ZLI prior to CD19 CART apheresis followed by ZM post CART resulted in 6-mo CR rates and 12-18 mo duration of CRs above historical rates reported for CART alone. Additional studies are needed to clarify the impact of Z exposure on changes in the TME and therapeutic effects of CART. The safety profile of ZM was manageable with low rates of G3 toxicities. Our results support evaluation of this combination in a larger cohort of aNHL.