Prognostic Analysis of Socioeconomic Factors in Multiple Myeloma Patients and Construction of A Myeloma-specific Survival Prediction Model / 肿瘤防治研究

ABSTRACT

Objective To investigate the effects of socioeconomic factors on the prognosis of multiple myeloma (MM) patients and construct a prediction model for evaluating myeloma-specific survival (MSS) rates. Methods A total of 32625 patients diagnosed with MM between January 2007 and December 2016 were included through the SEER database. Cox regression model was used to analyze the predictive indicators of MSS. The results of the multivariate subgroup analysis were presented as forest plots. The significant factors identified in the multivariate Cox analysis were used to construct a nomogram. The predictive performance of the nomogram was assessed using the AUC and calibration plots. A nomogram score-based risk stratification system was constructed using a restricted cubic spline. Results Patients were divided into five groups according to their socioeconomic status (SES). Groups with higher SES had relatively higher proportions of those part of the White, insured, married, and urban populations. Age, gender, race, marital status, insurance status, and SES were independent prognostic factors of MSS (all P < 0.001). The linear trend of increased MSS risk with decreasing SES was most pronounced among the White, married, insured, and urban patients (all P < 0.001). The nomogram exhibited good discrimination and accuracy in both training and validation sets, showing AUC values of 0.701, 0.709, and 0.722 for predicting 3-, 5-, and 8-year MSS, respectively. A risk stratification model was established based on the nomogram total points and the HR, which then divided patients into three different risk levels with substantial survival disparities (all P < 0.001). Conclusion Socioeconomic factors, such as marital status, insurance status, and SES, have a significant impact on the survival outcomes of MM patients. The nomogram and the risk stratification model based on these factors can accurately and reliably predict MSS.