探究初诊多发性骨髓瘤肾损伤可逆性的 预测模型

Exploring a Predictive Model for the Reversibility of Renal Impairment in Newly Diagnosed Multiple Myeloma

摘要: 目的：探究初诊多发性骨髓瘤(newly diagnosed multiple myeloma, NDMM)患者肾功能损伤可逆性的预测指标及初步构建预测模型。方法：收集并回顾性分析2013年3月至2025年3月我院收治的101例NDMM且伴有肾功能损伤患者的临床资料，根据国际骨髓瘤工作组(international myeloma working group, IMWG)定义的抗骨髓瘤治疗肾反应的标准，分为肾损伤恢复和肾损伤未恢复组，分析两组患者的一般临床资料和实验室指标，采用单因素和多因素Logistic回归分析，筛选出肾功能损伤恢复的独立影响因素并构建预测模型。结果：多因素Logistic分析结果得出白蛋白是多发性骨髓瘤患者肾功能损伤恢复的独立保护因素，而胱抑素C是影响肾功能恢复的独立危险因素(P < 0.05)。构建模型：P(z) = 1/(1 + e^−z)，z = −3.888 + 0.123 × 白蛋白 − 0.827 × 胱抑素C，联合预测概率的最佳临界值为0.73。在此截断值下敏感度为80.6%，特异度为82.8%。结论：血清白蛋白与胱抑素C是初诊多发性骨髓瘤患者肾功能损伤恢复的独立预测因子，基于两者联合建立的预测模型具有较高的预测价值。

Abstract: Objective: To explore the predictive indicators for the reversibility of renal impairment in patients with newly diagnosed multiple myeloma (NDMM) and to preliminarily construct a prediction model. Methods: The clinical data of 101 NDMM patients with renal impairment admitted to our hospital from March 2013 to March 2025 were collected and retrospectively analyzed. According to the criteria for renal response to anti-myeloma therapy defined by the International Myeloma Working Group (IMWG), the patients were divided into a renal recovery group and a renal non-recovery group. The general clinical characteristics and laboratory examinations of the two groups were compared. Univariate and multivariate logistic regression analyses were performed to identify the independent influencing factors for renal recovery and to construct a prediction model. Results: Multivariate logistic regression analysis demonstrated that albumin was a protective factor for the recovery of renal impairment in patients with multiple myeloma, whereas cystatin C was a risk factor (P < 0.05). The constructed prediction model was formulated as P(z) = 1/(1 + e^−z), where z = −3.888 + 0.123 × albumin − 0.827 × cystatin C. The optimal cut-off value for the combined predicted probability was 0.73. At this threshold, the sensitivity was 80.6% and the specificity was 82.8%. Conclusions: Serum albumin and cystatin C are independent predictors for the recovery of renal impairment in patients with NDMM, and the prediction model constructed by combining these two indicators demonstrates favorable predictive efficacy.