CAR ‐ HEMATOTOX score to predict risk of poor outcomes with odronextamab in heavily pretreated B‐cell non‐Hodgkin lymphoma

Summary The CAR‐HEMATOTOX model has been shown to help identify patients with relapsed/refractory (R/R) large B‐cell lymphoma (LBCL) at heightened risk of severe neutropenia, serious infection and worse clinical outcomes after chimeric antigen receptor T‐cell (CAR‐T) therapy. We evaluated whether CAR‐HEMATOTOX scores had prognostic value in patients with R/R follicular lymphoma (FL) or diffuse large B‐cell lymphoma (DLBCL) treated with odronextamab (cluster of differentiation [CD]20 × CD3 bispecific antibody) in the ELM‐1 (NCT02290951) and ELM‐2 (NCT03888105) studies. Patients were retrospectively stratified by baseline CAR‐HEMATOTOX score: CAR‐HEMATOTOX score low (HT low [0–1]; FL, n = 115, DLBCL, n = 92) or CAR‐HEMATOTOX score high (HT high [≥2]; FL, n = 24, DLBCL, n = 98). The incidence of grade ≥2 cytokine release syndrome within the first 30 treatment days, and rates of grade ≥3 infection and delayed neutrophil recovery within the first 90 treatment days, were numerically higher in HT high versus HT low patients. In patients with HT high versus HT low scores, progression‐free survival (PFS) was significantly poorer in DLBCL (2.6 vs. 9.6 months [ p < 0.0001]), and overall survival (OS) was significantly worse in FL and DLBCL (FL, 26.5 months vs. not reached [ p = 0.030]; DLBCL, 4.8 vs. 17.2 months [ p < 0.0001]). These results suggest that the CAR‐HEMATOTOX score may provide a readily available measure for identifying patients at risk of inferior clinical outcomes with odronextamab treatment.