功能性高危多发性骨髓瘤通过基因表达谱分析呈现统一的高增殖特征

Background: Multiple myeloma (MM) is a haematological malignancy with globally improving survival rates. Functional high-risk (FHR) MM, defined as progressive disease less than 18 months from diagnosis, has a median overall survival <3 years, but FHR patients are not identified by conventional prognostic scores nor by gene expression (GE) signatures, such as SKY92 (SkylineDx). Additionally, the biology of therapy resistance and aggressive behaviour in FHR myeloma is poorly understood. We sought to define the transcriptional features of FHR MM and identify molecular pathways that unify this aggressive phenotype. Methods: The MMProfiler microarray was performed on RNA extracted from bone-marrow aspirate samples of consecutive newly diagnosed MM (NDMM) patients. SKY92 risk classification (high- vs standard-risk) and genome-wide expression data (~20,000 genes) was obtained. Data was normalised across all samples prior to dimensionality reduction techniques, differential gene expression (DGE) analysis and pathway analysis. Clinical data and survival outcomes were collected. DGE analysis cutoffs were adj p<0.01 and |log2FC| >1.0). Results: Forty-eight patients were included, of whom 13 (27%) patients met criteria for FHR MM. FHR status was associated with inferior survival (3-year overall survival 58.7% versus 88%, p=0.018). Firth logistic regression demonstrated that R-ISS, R2-ISS and SKY92 classification were not significantly associated with FHR status (R-ISS OR 2.62; 95% CI 0.82-9.85; R2-ISS OR 2.08, 95% CI 0.94-5.64; SKY92 OR 3.55, 95% CI 0.96-16.03). DGE analysis identified a 16-gene proliferative module upregulated in FHR MM ( MCM2, FOXM1, CDCA5, KIF14, UBE2T, TYMS, KIF4A, NCAPH, SAC3D1, FAM72C, KIF15, ZWINT, CDT1, OIP5, TPX2, CDC20 ). These genes functionally converge on cell cycle progression and mitosis, and contain a significant FOXM1 -driven mitotic program. Formal pathway enrichment confirmed activation of E2F targets, G2/M checkpoint, MYC targets, and DNA repair pathways, consistent with a unified high-proliferation phenotype in FHR MM. A 6-gene proliferative signature derived from the DGE set ( FOXM1 and targets KIF14 , ZWINT , as well as MCM2 , UBE2T , and TYMS ) demonstrated strong discrimination of FHR status (AUC = 0.912; p = <0.001). Conclusion: FHR MM is poorly identified by current prognostic algorithms. Biologically, we show FHR MM is defined by a cell-cycle driving transcriptional program, at the gene expression level suggestive of a FOXM1 -driven program. This data suggests early progression in MM reflects a distinct biological state with increased cell-cycle and replication stress pathways. The 6-gene proliferative signature described here warrants validation as a diagnostic tool for FHR MM. The proliferative program seen in FHR MM may represent a novel therapeutic vulnerability, suggesting a role for cell-cycle targeting therapeutic approaches.