RESUMEN
Administration of chimeric-antigen receptor T-cell (CAR-T) therapy is complex and associated with unique toxicities. Identifying patients at risk for inferior outcomes is important for individualized management. The Glasgow-prognostic score (GPS) is a simple score shown to be highly prognostic of outcomes in the setting of traditional chemotherapy or checkpoint inhibitor administration. We sought to evaluate the value of the GPS to predict outcomes of patients with relapse refractory multiple myeloma (RRMM) receiving anti-BCMA CAR-T therapy. We included all patients treated with commercial CAR-T therapy for RRMM between 5/1/2021 and 2/1/2023 at the Moffitt Cancer Center. The GPS (CRP >1 mg/dL, 1 point; albumin <3.5, 1 point) was calculated for all patients at lymphodepletion (day -6) and patients were grouped as high-risk GPS (score = 2) or low-risk GPS (0 or 1). The primary endpoint was overall survival (OS) at day 100. A total of 139 pts were included, with a median follow-up of 6.7 months (95% CI, 6.2 to 8.9 months). Pts were treated with either idecabtagene vicleucel (83%) or ciltacabtagene autoleucel (17%). In total, 14% were classified with high-risk GPS, with significantly increased risk for grade 3 cytokine release syndrome (P = .003) and ICANS of any grade (P < .001). Patients in the high-risk GPS group had significantly lower day-100 OS (68.4% versus 97.3%, P < .001), OS at 6 months (56% versus 91.8% P = .0019) and PFS at 6 months (38.3% versus 72.3%, P = .03). The association of GPS with day-100 OS remained significant in a multivariable model. In conclusion, the GPS identifies a group of high-risk patients with RRMM receiving CAR-T therapy who experience increased rates of immune-mediated toxicity and are at higher risk for early mortality.