Machine learning-based cuproptosis prognostic signature identifies LDHA as a critical regulator of progression in multiple myeloma

Multiple myeloma (MM) is a cancer of plasma cells and is characterized by a poor prognosis. However, the role of cuproptosis in MM remains uncertain. From expression data of multicenter cohorts, 107 prognostic genes were identified. Ninety-five unique algorithmic combinations were evaluated through machine learning optimization to establish an optimal cuproptosis prognostic signature (CPPS). The CPPS model holds significant clinical relevance for MM, as individuals with low CPPS levels exhibit a better prognosis, heightened immune cell infiltration, increased expression of specific immune checkpoints, and enhanced sensitivity to immunotherapy. Furthermore, CPPS remains predictive of patient prognosis in other tumors. RT-qPCR analysis revealed elevated levels of SNRPE , ILF2 , LDHA , and PDHA1 in MM patient samples compared to those in healthy individuals, whereas FLNA , TAGLN2 , and ANXA2 expression was lower in the MM group than in the control cohort. The CCK-8 assay and flow cytometry confirmed that LDHA knockdown inhibits MM cell proliferation and promotes apoptosis. Immunofluorescence experiments have indicated that LDHA may promote MM progression and development via the NF-κB pathway. This study provides a valuable tool for guiding future clinical and personalized treatment approaches for MM.