Abstract
BACKGROUND: This study aims to explore the prognostic value of myosteatosis in multiple myeloma (MM) and to analyze the factors influencing myosteatosis. METHODS: A retrospective analysis was conducted on 182 patients treated for MM at our institution from 2009 to 2020 who underwent MRI examinations. The fatty infiltration rate (FIR) of the erector spinae and multifidus muscles at the L3 level was measured to assess the degree of myosteatosis. Patients were grouped based on fracture presence and median FIR, and group differences were compared, with P < 0.05 considered statistically significant. Survival and fractures were used as prognostic indicators, and regression analysis was performed to determine the impact of FIR on these outcomes in MM patients. The factors influencing FIR were analyzed, and the relationship between myosteatosis and MM prognosis was further analyzed within its sensitive subgroups. Finally, a nomogram based on FIR was established and validated. RESULTS: Significant differences were observed between the fracture and non-fracture groups in lactate dehydrogenase, serum phosphorus, visual analogue scale, oswestry disability index and FIR (P < 0.05). When patients were grouped based on the median FIR (28.89 %), there were significant differences in age, sex, body mass index (BMI), red blood cell (RBC) count, hemoglobin, hematocrit, albumin, visual analogue scale, oswestry disability index, and fracture incidence (P < 0.05). Univariate COX regression analysis indicated that myosteatosis had no significant impact on survival prognosis in MM patients (HR = 0.999, P = 0.852), with a log-rank test P value of 0.11 when grouped by the cut-off FIR value of 33.67 %. Multivariate logistic regression indicated that FIR is an independent predictor of fractures (OR = 1.054, P = 0.000). Multivariate linear regression revealed that age, sex, RBC count, and BMI are independent factors influencing FIR (P < 0.05). When not grouped, FIR's prediction of fractures showed no significant interaction with age, sex, RBC count, or BMI (P for interaction > 0.05). In subgroups with BMI ≥ 25 kg/m(2) or RBC count > 3.68 × 10^12/L, FIR lost its predictive significance for fractures. The FIR nomogram model had a C-index of 0.777, and the calibration curve, decision curve analysis, and clinical impact curve all validated its effectiveness. CONCLUSIONS: Myosteatosis characterized by FIR is not a reliable predictor of survival in MM patients but is effective in predicting fractures and is closely related to back pain and functional impairment. FIR is significantly associated with age, sex, RBC count, and BMI.