Validation and refinement of the new high-risk multiple myeloma (HRMM) criteria in relapsed or refractory patients treated with BCMA CAR-T.

Abstract Introduction: The consensus definition of high-risk multiple myeloma (HRMM) was recently revised by the International Myeloma Society (IMS) and International Myeloma Working Group (IMWG) to improve risk stratification in the era of contemporary therapies (Avet-Loiseau JCO 2025). The new definition of HRMM was constructed from clinical and genomic data in newly diagnosed MM patients treated with triplet or quadruplet regimens combining proteasome inhibitors, immunomodulatory agents, and/or anti-CD38 monoclonal antibodies. However, it is unknown if the HRMM definition can be used to risk stratify relapsed or refractory MM patients treated with BCMA CAR-T. Methods: We identified consecutive MM patients treated with BCMA CAR-T between 2017-2024. Patients with HRMM had at least one of the following criteria: (1) del(17p) and/or TP53 mutation with cancer cell fraction ≥20%; (2) t(4;14;), t(14;16), or t(14;20) co-occurring with 1q+ and/or del(1p32); (3) monoallelic del(1p32) with 1q+ or biallelic del(1p32); or (4) high B2M (≥5.5 mg/L) with normal creatinine (1.2 mg/dL) (Avet-Loiseau JCO 2025). Testing for del(1p32) and TP53 mutations was only available in a subset of patients due to the recent incorporation of these genetic aberration into testing platforms. Patients not meeting criteria for HRMM were considered to have standard-risk MM (SRMM). Extramedullary disease (EMD) was defined by the presence of plasmacytomas in visceral organs and/or soft tissue that was distinct from bones. Treatment responses were assessed per the IMWG. Univariable and multivariable hazard regression models were used to identify predictors of progression-free survival (PFS) and overall survival (OS). Results: A total of 158 patients were treated with BCMA CAR-T (ide-cel: n=77, cilta-cel: n=46; investigational: n=35), and 73 patients (46%) met the revised criteria for HRMM. Baseline EMD and high LDH (ULN) were present in 61 (39%) and 60 (38%) patients, respectively. After a median follow-up of 25 months (95% CI 23-30), the median PFS and OS for all patients were 15 months (95% CI 11-18) and 42 months (95% CI 32-NR), respectively. On univariable analysis, patients with revised HRMM had significantly shorter PFS (10 vs 18 months; HR 1.81, p=0.003) and OS (18 months vs NR; HR 2.89, p0.001). Revised HRMM remained independently associated with shorter PFS (HR 1.72, p=0.01) and OS (HR 2.90, p0.001) after adjusting for EMD, LDH, ferritin, and BCMA CAR-T drug product on multivariable analysis. EMD (PFS: HR 1.98, p=0.002; OS: HR 1.89, p=0.02) and high LDH (PFS: HR 1.72, p=0.01; OS: HR 2.11, p=0.005) were also associated with shorter PFS and OS on multivariable analysis. We then investigated the impact of incorporating EMD and high LDH as additional adverse prognostic factors into the HRMM criteria. Among patients with revised HRMM (n=73), 29 (40%), 30 (41%), and 14 (19%) patients had 0 (HRMM-0), 1 (HRMM-1), and 2 (HRMM-2) additional adverse risk factors. Patients with SRMM, HRMM-0, HRMM-1, and HRMM-2 had a median PFS of 18, 20, 7, and 7, respectively (p0.0001), and estimated 24-month OS of 78%, 71%, 21%, and 41%, respectively (p0.0001). There was no significant difference in PFS and OS between patients with SRMM versus HRMM-0 and HRMM-1 versus HRMM-2 (p0.05 for all comparisons). Based on these findings, we combined SRMM + HRMM-0 (reclassified-SRMM SRMM-R) and HRMM-1 + HRMM-2 (reclassified-HRMM HRMM-R) into two new risk groups. Compared to patients with SRMM-R, patients with HRMM-R had significantly shorter PFS (7 vs 18 months; p0.0001) and OS (13 months vs NR; p0.0001). The estimated 12-month PFS (24% vs 65%) and 36-month OS (19% vs 69%) were also significantly shorter with HRMM-R. Conclusion: Our data validate the recently revised HRMM criteria in MM patients treated with BCMA CAR-T. However, we also show that patients with HRMM in the absence of EMD or high LDH have survival outcomes similar to SRMM. Incorporation of EMD and LDH into the HRMM criteria can further refine the risk stratification and identify the ultra-high-risk patients (HRMM-R) for whom over 75% relapse within one year of receiving BCMA CAR-T. Patients with HRMM-R should be considered for the ongoing clinical trials investigating consolidation or maintenance strategies after BCMA CAR-T. This study is limited by its retrospective nature and lack of uniform testing for del(1p32) and TP53 mutations, and the findings need to be validated in a prospective, multi-institutional study.