Department of Gastroenterological Surgery, Graduate School of Medicine, Osaka University, Osaka, Japan.
Department of Surgery, Kansai Rosai Hospital, Osaka, Japan.
Ann Surg Oncol. 2017 Sep;24(9):2762-2769. doi: 10.1245/s10434-017-5914-3. Epub 2017 Jun 20.
Fatty pancreas (FP) was recently recognized as a risk factor for pancreatic ductal adenocarcinoma (PDAC). It is unclear whether computed tomography (CT) can be used to make a FP diagnosis. This study investigated whether CT could provide a predictive value for PDAC by diagnosing FP.
The study included 183 consecutive patients who underwent distal pancreatectomy from February 2007 to January 2017, including 75 cases of PDAC and 108 cases of other pancreatic disease. Pancreatic CT density (pancreatic index; PI) at the initial diagnosis was calculated by dividing the CT number in the pancreas by the number in the spleen. To assess whether CT could be used to detect FP, 43 cases were evaluated pathologically for FP. We investigated the correlation between FP and PI, and determined the optimal PI cutoff value for detecting FP using receiver operating characteristics analysis. We then investigated whether the PI value could be used as a predictor for PDAC.
Fourteen cases (32.6%) were pathologically diagnosed with FP. PI was significantly lower in the FP group versus the non-FP group (0.51 vs. 0.83; p = 0.0049). ROC analysis indicated that the PI had good diagnostic accuracy for FP diagnosis (cutoff value 0.70; sensitivity 0.79, specificity 0.79). Low PI (≤0.70) was identified in the multivariate analysis as an independent risk factor for PDAC (odds ratio 2.31; p = 0.023).
PI was strongly associated with pathological FP, which was independently associated with PDAC. PI shows promise as an imaging predictor for PDAC.
脂肪胰腺（FP）最近被认为是胰腺导管腺癌（PDAC）的一个危险因素。目前尚不清楚 CT 是否可用于 FP 的诊断。本研究通过诊断 FP 来探讨 CT 是否可以为 PDAC 提供预测价值。
本研究纳入了 2007 年 2 月至 2017 年 1 月期间连续行胰体尾切除术的 183 例患者，包括 75 例 PDAC 和 108 例其他胰腺疾病。在初始诊断时，通过将胰腺的 CT 数除以脾脏的 CT 数来计算胰腺 CT 密度（胰腺指数；PI）。为了评估 CT 是否可用于检测 FP，对 43 例病例进行了 FP 病理评估。我们研究了 FP 与 PI 之间的相关性，并通过接受者操作特征分析确定了用于检测 FP 的最佳 PI 截断值。然后，我们研究了 PI 值是否可作为 PDAC 的预测因子。
14 例（32.6%）经病理诊断为 FP。FP 组的 PI 明显低于非 FP 组（0.51 比 0.83；p=0.0049）。ROC 分析表明，PI 对 FP 诊断具有良好的诊断准确性（截断值 0.70；灵敏度 0.79，特异性 0.79）。多因素分析显示，低 PI（≤0.70）是 PDAC 的独立危险因素（比值比 2.31；p=0.023）。
PI 与病理 FP 密切相关，且与 PDAC 独立相关。PI 有望成为 PDAC 的影像学预测因子。
