P-228: The detection of PR minor clones at Single Cell Level identifies High Risk disease in Multiple Myeloma

Introduction: The development of new treatments has improved the outcome of Multiple Myeloma (MM) patients. However, MM remains an incurable disease and a subset of patients has suboptimal response to therapy, short survival, and are defined as high-risk (HR). Despite great efforts to understand the biology of MM, it remains difficult to identify HR disease and prognostic and predictive biomarkers to identify these patients are not well established. In this study, we have evaluated in relapsed/refractory (RR) MM and newly diagnosed (ND) MM patients a new method to identify HR disease based on the presence of highly proliferative (PR) subclones. We also aimed to characterize the transcriptomic signature of the PR cells and identify potential therapeutic targets for HR patients.

Methods: Single-cell RNA analyses were performed on bone marrow sorted CD138⁺ MM cells obtained from 58 ND and 78 RR MM patients with available serial samples. Unbiased mRNA profiling was conducted using the 10x GemCode system and sequencing performed on an Illumina platform. Cell Ranger and Seurat were used for data processing and analysis. Cells were classified as PR and non-PR based on a single-cell GSEA score obtained by using the Zhan gene dataset. Kaplan-Meier survival analysis was performed to evaluate the effect of PR cells on PFS. P values were calculated using the log-rank test.

Results: A PR signature identified in at least 5% of cells was found in 27% of ND and 50% of RR cases and found also in standard-risk patients based on FISH cytogenetics. Importantly, poor PFS was observed in all patients carrying PR subclones. In ND, the median PFS was 30 mos in PR patients and 43 mos in the non-PR patients (p=0.004). As expected, patients with TP53 mutations or chr17p deletion had a shorter PFS (17 mos). In RR, PR patients had a poor PFS of 6 mos when compared to 36 mos in non-PR patients (p=0.003). Of note, all PR patients retained their PR signature over time, and in 44% of patients an enrichment of PR cells was observed at subsequent progression, consistent with Darwinian clonal evolution due to therapeutic pressure. Furthermore, pathway enrichment analysis of PR cells revealed in both groups enrichment of MTORC1 signature, MYC targets, and DNA repair pathways with overexpression of LDH4, PPA1, TPI1, PPIA, ENO1, EIF4A1, and NPM1. Among the identified DEGs, high ENO1 (alpha-enolase) expression was associated with poor PFS in both groups (p=0.005 and 0.002), suggesting a potential role of ENO1 in HR disease.

Conclusions: In this study we have shown that the presence of PR cells identifies HR disease and patients with poor survival. We have also demonstrated that these PR subclones are enriched at time of progression. Therefore, their identification should be considered for better disease prognosis. Lastly, we were able to provide insight into the molecular signature of PR cells and identify some markers that may be used to detect HR disease and guide therapeutic intervention.