Total metabolic tumor volume for predicting cytokine release syndrome and treatment response in patients receiving bispecific antibodies for B-cell lymphoma.

Abstract

7059 Background: CD20xCD3 bispecific antibodies (BsAb) show high response rates with manageable cytokine release syndrome (CRS) in B-cell non-Hodgkin lymphomas (B-NHLs). While emerging data suggest that total metabolic tumor volume (TMTV) may predict treatment response and CRS, these findings have not yet been validated in real-world settings or across diverse histologic subtypes. Methods: We conducted a single center retrospective study evaluating B-NHL patients treated with BsAb. Baseline characteristics, efficacy, and safety outcomes were analyzed. TMTV was calculated from baseline PET scans using a semi-automated method with a threshold of 2x liver SUV mean . Median TMTV separated high/low groups. Results: A total of81 patients are included. The baseline characteristics, efficacy outcomes, and CRS outcomes of each histological subtype are summarized in Table 1. There were no Grade 4 CRS events observed. The median TMTV for DLBCL, FL and MCL was 107.1ml, 61.7ml, and 92.2ml, respectively. In DLBCL, patients with high TMTV were more likely to have bulky disease (>7.5cm) and elevated LDH. Other baseline characteristics as outlined in Table 1 were comparable to low TMTV group. High TMTV was associated with increased risk of CRS of any grade (OR 5.0, 95% CI 1.4-18.1, p=0.017) but was not associated with ORR, CR, PFS or OS. While bulky disease and elevated LDH levels were associated with TMTV, these clinical factors were not predictive of CRS. In contrast, TMTV retained its prognostic significance in multivariate analysis. In FL, high TMTV was associated with a higher rate of bulky disease while LDH level and other baseline characteristics did not differ from the low TMTV group. High TMTV correlated with a lower CR rate (OR 0.1, 95% CI 0.0-0.6, p=0.020) but did not influence CRS risk, PFS or OS. In the small cohort of MCL patients, no significant associations between TMTV and CRS or treatment efficacy were observed. TMTV has no correlation with neurotoxicity in all cohorts. Conclusions: High TMTV predicts for a higher rate of CRS on multivariate analysis in DLBCL, independent of clinical factors. High TMTV was associated with a lower CR rate in FL and further evaluation in MCL is needed with larger sample sizes. Baseline characteristics, efficacy outcomes, and CRS outcomes. DLBCLN=47 FLN=20 MCLN=14 Age, median (yr) 70 66 67 Male, n(%) 27 (57.4) 15 (75.0) 13 (92.9) BsAb regimen, n(%) Epcoritamab Glofitamab Mosunetuzumab 39 (83.0)4 (8.5)4 (8.5) 6 (30.0)3 (15.0)11 (55.0) 2 (14.3)8 (57.1)4 (28.6) Single agent BsAb treatment, n(%) 32 (68.1) 9 (45.0) 8 (57.1) Prior lines of therapy, median (range) 3 (0-10) 3 (1-7) 3 (2-5) Overall response rate (ORR), n(%) 25 (53.2) 16 (80.0) 9 (64.3) Complete response (CR), n(%) 10 (21.3) 8 (40.0) 7 (50.0) CRS, n(%) Grade 1 Grade 2 Grade 3 19 (40.4)13 (68.4)5 (26.3)1 (5.3) 11 (55.0)4 (36.4)7 (63.6)0 (0.0) 10 (71.4)5 (50.0)4 (40.0)1 (10.0)