A risk-scoring system for assessing prognosis in multiple myeloma patients based on a large real-world sample

Abstract

Objective Multiple myeloma (MM) is an incurable malignancy with a diversity of clinical characteristics and prognoses. The commonly used staging system has obvious shortcomings. Exploring accurate prognostic models is essential. Methods A total of 1,276 newly diagnosed MM patients were selected from Zhongshan Hospital Fudan University between January 2010 and April 2021. After excluding patients with amyloidosis or other tumors, a total of 802 patients receiving standard first-line therapy were included. 703 patients in the non-transplant group (527 patients in the training set and 176 patients in the validation set) and 109 patients in the transplant group. We enrolled 41 baseline parameters including clinical, laboratory, and pathological features. We used univariate and multivariate Cox analyses to screen for factors associated with overall survival and to develop prognostic models. Results The final risk-scoring system includes ECOG score, extramedullary lesion, thrombocyte, reticulocyte, anion gap, hypercalcemia, complement C3, β2-microglobulin, cytogenetics and interleukin-2 receptor. We identify the optimal cut-off for the risk score and divide the patients into high-risk and low-risk groups. Kaplan-Meier curves and Log-rank tests showed that the risk score was significant with overall survival in the training set (P < 0.001), validation set (P < 0.001) and transplant group (P = 0.02). The time-dependent receiver operator characteristic curve shows that the risk score has a better predictive value than the commonly used staging system. Conclusion A novel MM risk score system is developed based on a large real-world sample. We have performed a comprehensive assessment of baseline disease characteristics, which is of high application and practice value.