Predicting survival of persons with newly-diagnosed multiple myeloma.

BACKGROUND: Predicting survival in newly diagnosed multiple myeloma (NDMM) patients accurately remains challenging. Although mSMART3.0 is widely used it has not been rigorously tested for prediction accuracy. METHODS: We analysed clinical and laboratory variables in 1,008 NDMM patients, randomly split into a training set (n = 704) and a validation set (n = 304). Univariable and multivariable Cox regression analyses identified independent prognostic factors, forming a baseline survival model. For patients with available mSMART3.0 data, we integrated it with the baseline nomogram to enhance predictive accuracy. Time-dependent area under the receiver-operator characteristic curve (timeROC), calibration plots as well as comparisons with other survival prediction systems were used to estimated model performance. RESULTS: The baseline models, incorporating eight factors, achieved 5-year survival AUCs of 0.77 (95% CI: 0.72–0.81) in the training set and 0.73 (95% CI: 0.66–0.80) in the validation set. In the training set, 5-year survival was 34% (95% CI: 28–41%) in the high-risk group versus 72% (95% CI: 67–78%) in the low-risk group (p < 0.001), with similar differences observed in the validation set. The model exhibited excellent discrimination and calibration in both sets. In the training cohort the merged nomogram increased the 5-year survival AUCs of mSMART3.0 from 0.64 (0.57, 0.71) to 0.75 (0.67, 0.83; p = 0.003), better than the other staging systems of MM. CONCLUSIONS: The proposed nomograms increased the survival prediction accuracy in patients with NDMM.