Indiana University School of Medicine, Department of Surgery, 545 Barnhill Dr., Indianapolis, IN 46202, USA.
Indiana University School of Medicine, Department of Surgery, 545 Barnhill Dr., Indianapolis, IN 46202, USA; Indiana University Health Pancreatic Cyst and Cancer Early Detection Center, 550 University Blvd., Indianapolis, IN 46202, USA.
HPB (Oxford). 2019 May;21(5):574-581. doi: 10.1016/j.hpb.2018.09.004. Epub 2018 Oct 5.
For pancreatic cysts with negative cytology, Integrated Molecular Pathology (IMP) is a malignancy risk score integrating clinical criteria with pancreatic cyst fluid DNA profiling. Aside from main pancreatic duct (MPD) diameter, integrated clinical criteria are not International Consensus Guidelines High-Risk Stigmata. We predicted exclusion of clinical criteria except MPD diameter could simplify the IMP and better distinguish invasive/malignant disease.
Records of >1100 patients with IPMN were reviewed retrospectively. Sensitivity, specificity, and accuracy of conventional IMP for invasive/malignant disease was compared to DNA profile including only MPD ≥10mm (IMP-10.) Invasive outcomes were invasive-IPMN/adenocarcinoma on surgical pathology, pathologic or radiographic evidence of invasive/metastatic disease during surveillance. Malignant outcomes included high grade dysplastic IPMN (HGD-IPMN).
225 patients who met study criteria underwent 283 IMP evaluations: 98 followed by surgery, 185 followed by ≥ 23 months surveillance. IMP-10 had greater specificity (90.1% vs. 73.7%) and accuracy (89.8% vs. 74.2%) for invasive disease compared to IMP in surgery + surveillance patients, but lower sensitivity (77.8% vs. 88.9%). Trends were similar in surgery patients alone and malignant outcome analyses.
IMP-10 excludes less-reliable clinical factors resulting in greater accuracy in predicting invasive/malignant disease and fewer patients with benign disease being recommended for surgery.
对于细胞学阴性的胰腺囊肿,综合分子病理学(IMP)是一种恶性风险评分,它将临床标准与胰腺囊液 DNA 分析相结合。除主胰管(MPD)直径外,综合临床标准不是国际共识指南的高危标志。我们预测排除除 MPD 直径以外的临床标准可以简化 IMP,并更好地区分侵袭性/恶性疾病。
回顾性分析了 >1100 例 IPMN 患者的记录。比较了传统 IMP 对侵袭性/恶性疾病的敏感性、特异性和准确性,与仅包括 MPD≥10mm 的 DNA 图谱(IMP-10)进行比较。侵袭性结局是手术病理上的侵袭性 IPMN/腺癌,在监测期间有侵袭性/转移性疾病的病理或影像学证据。恶性结局包括高级别异型增生 IPMN(HGD-IPMN)。
符合研究标准的 225 例患者接受了 283 次 IMP 评估:98 例接受了手术,185 例接受了≥23 个月的监测。与手术+监测患者的 IMP 相比,IMP-10 对侵袭性疾病具有更高的特异性(90.1%比 73.7%)和准确性(89.8%比 74.2%),但敏感性较低(77.8%比 88.9%)。在仅手术患者和恶性结局分析中也存在类似的趋势。
IMP-10 排除了不太可靠的临床因素,从而提高了预测侵袭性/恶性疾病的准确性,并减少了建议手术的良性疾病患者数量。