A proliferation score defined by PCLI-anchored transcriptomics predicts outcomes in multiple myeloma

Cell proliferation is a fundamental driver of disease aggressiveness in multiple myeloma (MM); however, transcriptional programs underpinning proliferative capacity remain incompletely defined. Using Affymetrix microarray data from 136 Mayo Clinic patients with MM with paired plasma cell labeling index (PCLI), we performed a differential expression analysis to identify genes consistently associated with proliferative activity. These genes were subsequently incorporated into the development of a prognostic model using baseline RNA sequencing (RNA-seq) profiles in purified CD138⁺ cells from 656 patients in the CoMMpass cohort, in which an elastic net-based Cox model yielded an 8-gene-based proliferation score (GPS). Differential expression analysis identified 108 proliferation-associated genes, and the resulting 8-gene GPS (CHEK1, EXO1, MCM4, TK1, KIF14, SEPTIN11, PHF19, and HMGB3) formed a highly connected biological network enriched for DNA damage response and cell cycle regulation. CHEK1, TK1, EXO1, and MCM4 were mapped in a large network with CHEK1 as the hub gene. In the CoMMpass data set, on dividing the GPS score by the median, the high-GPS group (n = 328) had a median overall survival (OS) of 63.6 months (95% confidence interval, 56.5-76.7), whereas the low-GPS group (n = 328) had a median OS of 106.1 months (P < .001). Similarly, in another Mayo Clinic RNA-seq cohort, high-GPS group (n = 12) demonstrated worse survival (median OS, 39.2 vs 157.1 months for the low-GPS group; P = .03). GPS emerged as an independent prognostic factor alongside high-risk markers and also predicted S-phase content in an independent data set.