B05 plasma cell leukemia-like status has independent prognostic value in the context of the second revision of the international staging system

Introduction: Even though the majority of newly diagnosed multiple myeloma (NDMM) patients has benefited significantly from the introduction of novel agents, it remains an unmet need to identify the best upfront treatment strategy for high-risk disease. Risk-adapted trial designs (e.g., the OPTIMUM/MUKnine trial, NCT03188172, and the GMMG-CONCEPT trial, NCT03104842) aim to answer this question, which in turn drive research efforts to improve risk assessment in NDMM. Recently, our group has published two novel prognostic tools for NDMM: (I) The Second Revision of the International Staging System (R2-ISS), which identifies four different risk categories by combining conventional prognostic markers (D’Agostino et al. – J Clin Oncol 2022). (II) Plasma Cell Leukemia-like (PCL-like) status, which identifies high-risk disease by combining tumor transcriptomic data of 54 genes reflecting high levels of circulating tumor cells (Hofste op Bruinink et al. – J Clin Oncol 2022). Since both tools may detect different types of aggressive disease, we hypothesized that combining PCL-like status with the R2-ISS classification would improve prognostic accuracy in NDMM. Methods: Baseline characteristics including serum markers and high-risk cytogenetic aberrations were collected from NDMM patients enrolled in the HOVON-65/GMMG-HD4 (EudraCT 2004-000944-26), HOVON-87/NMSG-18 (EudraCT 2007-004007-34), EMN02/HO95 (EudraCT 2009-017903-28) and MMRF CoMMpass studies (NCT01454297) and used to determine the R2-ISS classification. PCL-like status was calculated from transcriptomic profiles of CD138-enriched bone marrow tumor cells. Hazard ratios (HRs) for PCL-like status, R2-ISS classification or a combination thereof were estimated using a Cox proportional hazards model stratified by study cohort and corrected for age ≤ 65 years. Results: For 865 NDMM patients of known age, both the R2-ISS classification and PCL-like status could be determined, with a median follow-up time of 55 months. 18% were classified as R2-ISS low (R2-ISS I), 30% as R2-ISS low-intermediate (R2-ISS II), 43% as R2-ISS intermediate-high (R2-ISS III) and 9% as R2-ISS high (R2-ISS IV); 10% as PCL-like MM and 90% as intramedullary MM (i-MM). In univariate analyses, PCL-like status was associated with both an inferior PFS (HR, 1.8; 95% confidence interval (CI), 1.4 to 2.3; P < 0.0001) and OS (HR, 2.4; 95% CI, 1.8 to 3.3; P < 0.0001). This also applied to the R2-ISS classification when comparing R2-ISS IV to R2-ISS I, II and III (HR for PFS, 2.0; 95% CI, 1.5 to 2.6; P < 0.0001 and HR for OS, 3.0; 95% CI, 2.2 to 4.2; P < 0.0001). PCL-like status retained its prognostic significance in a multivariate model when combined with the R2-ISS classification, both in terms of PFS (HR, 1.6; 95% CI, 1.2 to 2.1; P = 0.0004) and OS (HR, 2.1; 95% CI, 1.6 to 2.9; P < 0.0001). Of note, 2% of patients classified both as PCL-like MM and R2-ISS IV, translating into a median PFS of 12 months and a median OS of 15 months. Patients with PCL-like MM and R2-ISS III (5%) had a comparable median OS to those with i-MM and R2-ISS IV (7%) (47 and 38 months, respectively). Validation is planned in other European Myeloma Network trial cohorts. Conclusions: 1. PCL-like status combined with the R2-ISS classification improves prognostic accuracy in NDMM. 2. The presence of both PCL-like MM and R2-ISS IV may confer exceptionally high risk. 3. Patients with PCL-like MM and R2-ISS III have a comparable OS to those with i-MM and R2-ISS IV.