Identification of an optimal absolute lymphocyte count before bispecific antibody therapy for large B-cell lymphoma: A collaborative US bispecific consortium Study

Abstract Background: Currently approved in the third-line or later settings, anti-CD3xCD20 bispecific antibodies (BsAb) are increasingly utilized in the treatment of large B-cell lymphoma (LBCL). These therapies harness CD3+ T cells for anti-tumor activity by recruiting and activating them to target CD20-expressing tumor cells. Thus, T-cell fitness, including both quantitative and qualitative functional impairments, may influence their efficacy, particularly given their use in later lines of treatment. We hypothesized that baseline absolute lymphocyte count (ALC) may be associated with clinical outcomes in patients with relapsed/refractory (R/R) LBCL receiving BsAb therapy and aimed to identify a clinically significant threshold. Methods: Patients with R/R LBCL who received standard-of-care glofitamab or epcoritamab within the Collaborative US Bispecific Consortium (CUBIC) were included in this study. Baseline characteristics, including ALC, were collected prior to BsAb initiation.Cytokine release syndrome (CRS) and immune effector cell-associated neurotoxicity syndrome (ICANS) were graded according to the ASTCT consensus grading system. Treatment response and survival were assessed per Lugano 2014 criteria. A threshold for ALC associated with progression-free survival (PFS) was identified using a maximally selected rank test. Fischer's exact test or Wilcoxon rank-sum test were used for univariate associations of categorical and continuous variables, and logistic regression for multivariate analysis. Results: A total of 262 patients were included in the analysis: 135 (51.5%) treated with epcoritamab and 127 (48.5%) with glofitamab. At time of BsAb initiation, median age was 70 years (range 23-93), 31 (12%) had double- or triple-hit LBCL, 94 (66%) had an international prognostic index score of 3 or higher, median number of prior lines of systemic therapy was 3 (range 1-9), and median ALC was 0.5×10³/μL (range 0-307 × 10³/μL). CRS and ICANS of any grade occurred in 103 (39%) and 36 (14%) patients, respectively, with grade ≥3 occurring in 14 (5%) and 6 (2%) patients. Overall response rates (ORR) and complete response (CR) rate were 47% and 32%, respectively. After a median follow-up of 9.1 months (95% confidence interval [CI] 7.7–12.1), median PFS was 4.6 months. The optimal ALC threshold was identified at 0.19×10³/μL: 26 patients (10%) had ALC below the threshold (low ALC group) and 236 (90%) above the threshold (high ALC group). Median PFS was 1.7 months in the low ALC group and 4.9 months in the high ALC group (p=0.0007). No significant differences in the rates of CRS (p=0.321) of any grade, CRS grade 3 or higher (p=1.000), ICANS of any grade (p=0.635) and ICANS grade 3 or higher (p=0.468) were observed when comparing the two groups. Patients with low ALC experienced significantly lower ORR (23% vs 50%, p=0.012) and CR rate (12% vs 34%, p=0.025) as compared to those with high ALC. As compared to high ALC, factors significantly associated with low ALC on univariate analysis were: lower absolute monocyte count (median 0.4×10 ×10³/μL vs 0.6×10³/μL , p=0.009), lower hemoglobin (median 9.2 g/dL vs 10.8 g/dL, p=0.002), lower platelet count (median 114×10³/μL vs 159×10³/μL, p=0.042), and higher ferritin levels (median 3774 ng/mL vs 401 ng/mL, p=0.048). In multivariate analysis, only low hemoglobin levels maintained an association (odds ratio 1.38, 95% CI 1.06-1.79, p=0.016). Conclusion: An optimal ALC threshold of 0.19×10³/μL could help identify patients with R/R LBCL who may experience deeper responses and longer PFS with epcoritamab or glofitamab. Biological analysis, aimed at investigating clinically relevant T-cell subgroups, and prospective validation of these findings are warranted.