Abstract:

Objective To explore the early warning model of acute graft-versus-host disease（aGVHD）after haplotype hematopoietic stem cell transplantation in adult myelodysplastic syndrome（MDS）. Methods The clinical data of 50 adult MDS patients who received haplotype hematopoietic stem cell transplantation at the Second Affiliated Hospital of Air Force Medical University from January 2021 to January 2023 were included. Patients with acute graft-versus-host disease after treatment were included in the occurrence group，and patients withoutacute graft-versus-host disease were included in the non-occurrence group. The clinical data of patients were collected and analyzed，the influencing factors of acute graft-versus-host disease after haplotype hematopoietic stem cell transplantation were determined by Logistic regression analysis，and the nomogram early warning model was constructed by using the determined independent influencing factors，and the effectiveness of the nomogramprediction model for acute graft-versus-host disease after haplotype hematopoietic stem cell transplantation wasevaluated. Results Fifty adult MDS patients received haplotype hematopoietic stem cell transplantation in our hospital. The incidence of patients with acute graft-versus-host disease was 30%，while the proportion of patientswithout acute graft-versus-host disease was 70%. Logistic regression analysis showed that donor sex，race and C- reactive protein were the risk factors for acute graft-versus-host disease（OR ＞ 1，P ＜ 0.05），while uric acid was the protective factor for acute graft-versus-host disease（OR ＜ 1，P ＜ 0.05）in adult MDS patients after haplotype hematopoietic stem cell transplantation. ROC curve was drawn with certain independent influencing factors，and the results showed that the AUC values of donor's sex，race，uric acid and C- reactive protein were all above 0.60，which indicated that the above indexes provided value for early warning of acute graft-versus-host disease afterhaplotype hematopoietic stem cell transplantation. Using the above independent influencing factors，a nomogram earlywarning model was constructed，and the value obtained by adding the scores of risk factors was the probability of acute graft-versus-host disease after haplotype hematopoietic stem cell transplantation. The verification results showed that the C-index value of the calibration curve was 0.846，which indicated that the nomogram prediction model had good prediction energy efficiency. Conclusion The nomogram constructed by influencing factors of acute graft-versus-host disease after haplotype hematopoietic stem cell transplantation can predict the incidence of this disease，specific treatment and nursing should be carried out according to gender，race and C-reactive protein to reduce the risk of acute graft-versus-host disease. 

Key words:

Acute graft-versus-host disease , Myelodysplasia , Stem cell transplantation , Modelconstruction, Nurse

摘要：

目的 探讨成人骨髓增生异常综合征（myelodysplastic syndrome，MDS）单倍型造血干细胞移植后急性移植物抗宿主病（acute graft-versus-host disease，aGVHD）的预警模型构建。 方法 纳入 2021 年1 月至 2023 年 1 月空军军医大学第二附属医院 50 例接受单倍型造血干细胞移植成人 MDS 患者资料，将治疗后发生 aGVHD 的患者纳入发生组，未发生 aGVHD 的患者纳入未发生组。收集、分析患者临床资料，通过 Logistic 回归分析法确定单倍型造血干细胞移植后发生 aGVHD 的影响因素，利用确定的独立影响因素构建列线图预警模型并评判单倍型造血干细胞移植后发生 aGVHD 的列线图预测模型效能。 结果 共 50 例成人 MDS 患者在本院接受单倍型造血干细胞移植治疗，发生 aGVHD 的患者比例为 30%（15/50），未发生aGVHD 的患者比例为 70.00%（35/50）。Logistic 回归分析结果显示，供者性别、种族、C- 反应蛋白是成人MDS 患者单倍型造血干细胞移植后发生 aGVHD 的危险因素（OR ＞ 1，P ＜ 0.05），尿酸是成人 MDS 患者单倍型造血干细胞移植治疗后发生 aGVHD 的保护因素（OR ＜ 1，P ＜ 0.05）。采用确定的独立影响因素绘制ROC 曲线，曲线结果显示供者性别、种族、尿酸、C- 反应蛋白的 AUC 值均＞ 0.60，说明以上指标为预警单倍型造血干细胞移植后发生 aGVHD 提供价值。利用以上独立影响因素构建列线图预警模型，危险因素分值相加所得数值即为单倍型造血干细胞移植后发生 aGVHD 的概率，验证结果显示，校准曲线的 C-index 值为 0.846，表明该列线图预测模型预测能效良好。 结论 采用单倍型造血干细胞移植后 aGVHD 的影响因素构建的列线图可预测该病的发生概率，应针对性别、种族、C- 反应蛋白等进行针对性治疗及护理，以降低 aGVHD 发生风险。

关键词:

 , 急性移植物抗宿主病 , 骨髓增生 , 干细胞移植 , 模型构建 , 护理