Department of Genetics, Assistance Publique-Hôpitaux de Paris, Pitié -Salpétrière Hospital, University Pierre et Marie Curie-Paris 6, 75013 Paris, France.
J Clin Endocrinol Metab. 2011 Aug;96(8):E1346-51. doi: 10.1210/jc.2011-0268. Epub 2011 Jun 15.
The diagnosis of maturity-onset diabetes of the young type 3 (MODY3), associated with HNF1A molecular abnormalities, is often missed.
The objective of the study was to describe the phenotypes of a large series of MODY3 patients and to reassess parameters that may improve its diagnosis.
DESIGN, SETTING, AND PATIENTS: This retrospective multicenter study included 487 unrelated patients referred because of suspicion of MODY3. Genetic analysis identified 196 MODY3 and 283 non-MODY3 cases. Criteria associated with MODY3 were assessed by multivariate analysis. The capacity of the model to predict MODY3 diagnosis was assessed by the area under the receiver-operating characteristic curve and was further validated in an independent sample of 851 patients (165 MODY3 and 686 non-MODY3).
In the MODY3 patients, diabetes was revealed by clinical symptoms in 25% of the cases and was diagnosed by screening in the others. Age at diagnosis of diabetes was more than 25 yr in 40% of the MODY3 patients. There was considerable variability and overlap of all assessed parameters in MODY3 and non-MODY3 patients. The best predictive model was based on criteria available at diagnosis of diabetes, including age, body mass index, number of affected generations, presence of diabetes symptoms, and geographical origin. The area under the curve of the receiver-operating characteristic analysis was 0.81. When sensitivity was set to 90%, specificity was 49%.
Differential diagnosis between MODY3 and early-onset type 2 diabetes remains difficult. Whether the proposed model will improve the pick-up rate of MODY3 diagnosis needs to be confirmed in independent populations.
与 HNF1A 分子异常相关的年轻起病的成年型糖尿病 3 型（MODY3）的诊断常常被忽视。
本研究的目的是描述一大系列 MODY3 患者的表型，并重新评估可能改善其诊断的参数。
设计、地点和患者：这项回顾性多中心研究纳入了因疑似 MODY3 而就诊的 487 名无血缘关系的患者。基因分析确定了 196 例 MODY3 和 283 例非 MODY3 病例。采用多变量分析评估与 MODY3 相关的标准。通过受试者工作特征曲线下面积评估该模型预测 MODY3 诊断的能力，并在另一个 851 例患者（165 例 MODY3 和 686 例非 MODY3）的独立样本中进行验证。
在 MODY3 患者中，25%的病例是因临床症状而发现糖尿病，其余病例是通过筛查而诊断的。40%的 MODY3 患者糖尿病的诊断年龄超过 25 岁。MODY3 和非 MODY3 患者的所有评估参数均存在相当大的差异和重叠。最佳预测模型基于糖尿病诊断时可用的标准，包括年龄、体重指数、受影响的代际数、糖尿病症状的存在和地理来源。受试者工作特征分析的曲线下面积为 0.81。当灵敏度设定为 90%时，特异性为 49%。
MODY3 与早发 2 型糖尿病之间的鉴别诊断仍然困难。所提出的模型是否能提高 MODY3 诊断的检出率，还需要在独立人群中得到证实。
