Ishido Keinosuke, Fujita Hiroaki, Kimura Norihisa, Nagase Hayato, Wakasa Yusuke, Tsuruta Satoru, Muroya Takahiro, Hakamada Kenichi
Department of Gastroenterological Surgery, Hirosaki University Graduate School of Medicine, Japan.
Pancreas. 2025 Jul 18;54(9):e765-75. doi: 10.1097/MPA.0000000000002508.
Despite advancements in pancreatic surgery, new-onset diabetes mellitus following pancreatectomy (NODMP), persists, affecting patients' quality of life. Predicting NODMP before surgery could significantly enhance postoperative care.
This study included 220 patients who underwent pancreatoduodenectomy or distal pancreatectomy at Hirosaki University Hospital between January 2008 and December 2020. Patients with preoperative diabetes or <6 months' follow-up were excluded. The anticipated remnant pancreatic-to-splenic parenchyma computed tomography value ratio (remP/S ratio) was used to assess pancreatic fat content, with its cutoff determined using the receiving operator characteristic curve. Time to diabetes onset was analyzed using the Kaplan-Meier method. The risk factors for NODMP were identified using the Cox proportional hazards model.
The mean diabetes-free period was 89.2 months over a median follow-up of 25.1 months. The incidence rates of NODMP at 1, 3, and 5 years after resection were 7.21%, 21.3%, and 28.0%, respectively. The significant risk factors for NODMP identified by univariate analysis were pancreatic cancer, preoperative HbA1c >5.7%, remP/S ratio <0.66, and remnant pancreatic volume <32.7 cm3. Multivariate analysis confirmed that a remP/S ratio <0.66 and preoperative HbA1c >5.7% were independent predictors of NODMP. The risk scoring system indicated that patients with both risk factors have a fivefold higher risk of developing NODMP within 2 years compared with those without either risk factor.
Preoperative remP/S ratio and HbA1c were significant predictors of NODMP, enabling the effective stratification of NODMP risk and facilitating the early treatment of the disease.
尽管胰腺手术取得了进展，但胰十二指肠切除术后新发糖尿病（NODMP）仍然存在，影响患者的生活质量。术前预测NODMP可显著改善术后护理。
本研究纳入了2008年1月至2020年12月期间在弘前大学医院接受胰十二指肠切除术或远端胰腺切除术的220例患者。排除术前患有糖尿病或随访时间<6个月的患者。采用预期残余胰腺与脾脏实质的计算机断层扫描值比（remP/S比）评估胰腺脂肪含量，并使用受试者工作特征曲线确定其临界值。采用Kaplan-Meier方法分析糖尿病发病时间。使用Cox比例风险模型确定NODMP的危险因素。
在中位随访25.1个月期间，平均无糖尿病期为89.2个月。切除术后1年、3年和5年的NODMP发病率分别为7.21%、21.3%和28.0%。单因素分析确定的NODMP的显著危险因素为胰腺癌、术前糖化血红蛋白>5.7%、remP/S比<0.66和残余胰腺体积<32.7 cm³。多因素分析证实，remP/S比<0.66和术前糖化血红蛋白>5.7%是NODMP的独立预测因素。风险评分系统表明，与没有任何一个危险因素的患者相比，同时具有这两个危险因素的患者在2年内发生NODMP的风险高五倍。
术前remP/S比和糖化血红蛋白是NODMP的重要预测因素，能够有效地对NODMP风险进行分层，并有助于疾病的早期治疗。
