Abstract 7663: Identification of high-risk patients in multiple myeloma using a clonal gene signature

Abstract Introduction: Multiple myeloma (MM) is a heterogeneous disease with a small subset of high-risk patients associated with poor prognosis. The identification of these patients is crucial for treatment management and strategic decisions. Methods: We developed a novel computational framework that only requires gene expression profiles to define prognostic gene signatures by selecting genes with expression driven by inferred clonal copy number alterations. We applied this framework to MM and developed a clonal gene signature (CGS) consisting of only 22 genes. This prognostic signature was evaluated in 5 independent datasets with gene expression data of a total of 2, 155 MM samples. For each patient, we calculated a prognostic score based on the expression of the signature genes. Multivariate Cox regression analyses were performed to investigate the association of these signature scores with patient prognosis after adjusting established clinical factors. The performances of CGS and the only commercialized MM gene signatures, GEP70 and SKY92, in identifying high-risk patients were carefully compared by multiple different approaches. Results: Our results indicated that CGS provided significant prognostic values after adjusting for well-established prognostic factors including cytogenetic abnormalities, International Staging System (ISS), and Revised ISS (R-ISS). In addition, it showed significant prognostic associations in patient subsets stratified by these conventional prognostic factors. Importantly, CGS demonstrated a better performance in identifying high-risk patients compared to the GEP70 and SKY92 signatures, which have been recommended for prognostic stratification in MM. Moreover, CGS can further stratify patients into subgroups with significantly differential prognoses when applied to the high-risk and low-risk groups already identified by GEP70 and SKY92. The higher CGS score of patients is significantly associated with poor response to dexamethasone, a commonly used drug for MM treatment. Conclusions: In summary, we proposed a computational framework that only requires gene expression data to identify clonal gene signatures for prognosis prediction. The CGS provides a useful biomarker for improving prognostic stratification in MM, especially for the identification of patients with the highest risk. The computational framework used for defining this signature provides a useful tool, which can be readily applied to many other cancer types. Citation Format: Jian-Rong Li, Christiana Wang, Chao Cheng. Identification of high-risk patients in multiple myeloma using a clonal gene signature abstract. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts) ; 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84 (6Suppl): Abstract nr 7663.