Honselmann Kim C, Krauss Tobias, Geserick Sebastian, Wellner Ulrich F, Wittel Uwe, Hopt Ulrich T, Keck Tobias, Bausch Dirk
Department of Surgery, University Medical Center Schleswig-Holstein, Campus Luebeck, Ratzeburger Allee 160, 23538, Luebeck, Germany.
Department of Radiology, University Medical Center Freiburg, Freiburg im Breisgau, Germany.
Langenbecks Arch Surg. 2016 Jun;401(4):449-56. doi: 10.1007/s00423-016-1416-1. Epub 2016 Apr 5.
The purpose of this study was to retrospectively evaluate diagnostic accuracy of cystic lesions of the pancreas in order to determine if less aggressive surgical treatment might be safe and therefore warranted.
A retrospective cohort study was conducted in 232 patients with either observed or resected cystic lesions of the pancreas referred for evaluation and treatment to the University Medical Center Freiburg, Germany, between 2001 and 2011.
Most patients had MRI or CT for preoperative imaging (90.6 %). Preoperatively, benign pseudocysts (BPC) were diagnosed in 84 (36.2 %) patients and intraductal papillary mucinous neoplasm (IPMN) in 59 (25.2 %) patients, whereas serous cyst adenoma, mucinous cystic neoplasm (MCN), solid pseudopapillary tumors (SPPTs), and neuroendocrine tumors (NETs) were less common. In 43 % of patients, the preoperative diagnosis concurred with the postoperative diagnosis. The preoperative diagnosis was accurate in BPC, less so in IPMN, and inaccurate in MCN, NET, and SPPT. However, prediction of tumor biology was accurate; only 11 % of the lesions regarded as benign turned out to be malignant after resection, and no patient without resection developed malignancy at a median follow-up of 8 months. Subsequently, 89 % of diagnosed benign tumors had indeed benign pathology.
The prediction of biology is often correct, whereas specific diagnosis is often wrong. A considerable amount of benign lesions are treated more aggressively than warranted if malignancy is suspected prior to surgery. Parenchyma-sparing techniques might be an option, but prospective multicenter studies need to follow. Experienced pancreatic radiologists can improve accuracy of preoperative biology.
本研究旨在回顾性评估胰腺囊性病变的诊断准确性，以确定侵袭性较小的手术治疗是否安全且必要。
对2001年至2011年间转诊至德国弗莱堡大学医学中心进行评估和治疗的232例胰腺囊性病变患者进行回顾性队列研究，这些患者的病变已被观察或切除。
大多数患者术前行MRI或CT检查（90.6%）。术前，84例（36.2%）患者被诊断为良性假性囊肿（BPC），59例（25.2%）患者被诊断为导管内乳头状黏液性肿瘤（IPMN），而浆液性囊腺瘤、黏液性囊性肿瘤（MCN）、实性假乳头状肿瘤（SPPT）和神经内分泌肿瘤（NET）则较少见。43%的患者术前诊断与术后诊断一致。术前诊断在BPC中准确，在IPMN中准确性稍低，在MCN、NET和SPPT中不准确。然而，肿瘤生物学行为的预测是准确的；切除后，仅11%被视为良性的病变被证实为恶性，在中位随访8个月时，未接受切除的患者均未发生恶性病变。随后，89%被诊断为良性的肿瘤病理结果确实为良性。
生物学行为的预测通常是正确的，而具体诊断往往是错误的。如果术前怀疑为恶性，相当数量的良性病变会接受比必要程度更积极的治疗。保留实质的技术可能是一种选择，但需要进行前瞻性多中心研究。经验丰富的胰腺放射科医生可以提高术前生物学行为预测的准确性。
