Aysal Anil, Agalar Cihan, Egeli Tufan, Unek Tarkan, Oztop Ilhan, Obuz Funda, Sagol Ozgul
Department of Pathology, Dokuz Eylul University School of Medicine, Izmir, Turkey.
Department of General Surgery, Dokuz Eylul University School of Medicine, Izmir, Turkey.
Endocr Pathol. 2021 Dec;32(4):461-472. doi: 10.1007/s12022-021-09687-w. Epub 2021 Jul 20.
The question of how successful we are in predicting pancreatic neuroendocrine tumors (panNET) with poor prognosis has not been fully answered yet. The aim of this study was to investigate the effects of clinicopathological features on prognosis and to determine their validity in prediction of prognosis and whether a better prognostic classification can be made. Fifty-six patients who underwent pancreatic resection for pancreatic neuroendocrine tumor were included. The associations between clinicopathological parameters and prognosis were evaluated statistically. Efficiencies of different thresholds for tumor size, mitotic count, and Ki67 proliferation index for prognosis prediction were compared. Vascular invasion was statistically associated with high tumor grade, advanced pT stage, and mortality rate. The presence of non-functional tumor, lymphatic invasion, and > 10 cm tumor size were significantly related to shorter overall survival. Advanced pT stage (pT3-4), > 5 cm tumor size, and high tumor grade (grades 2-3) were significantly associated with shorter disease-free survival. The mortality rate showed the strongest statistical significance with mitotic count when grouped as 1: < 2, 2: 2-10, and 3: > 10 mitosis/ 2 mm. The 10% threshold value for Ki67 index was more successful in predicting adverse prognosis. Among the morphologic variants, the ductulo-insular variant was the most promising to have positive prognostic value in our series, although no statistical significance was detected. In conclusion, threshold values of 5 cm and 10 cm for tumor size, 10% for Ki67 proliferation index, and 10/2 mm for mitotic count and vascular and lymphatic invasion assessed separately are potential prognostic candidates for better stratification of panNETs.
我们在预测预后不良的胰腺神经内分泌肿瘤（panNET）方面的成功程度问题尚未得到充分解答。本研究的目的是调查临床病理特征对预后的影响，并确定其在预后预测中的有效性，以及是否可以进行更好的预后分类。纳入了56例因胰腺神经内分泌肿瘤接受胰腺切除术的患者。对临床病理参数与预后之间的关联进行了统计学评估。比较了肿瘤大小、有丝分裂计数和Ki67增殖指数的不同阈值对预后预测的效率。血管侵犯与高肿瘤分级、晚期pT分期和死亡率在统计学上相关。无功能肿瘤的存在、淋巴管侵犯和肿瘤大小>10 cm与较短的总生存期显著相关。晚期pT分期（pT3-4）、肿瘤大小>5 cm和高肿瘤分级（2-3级）与较短的无病生存期显著相关。当按1：<2、2：2-10和3：>10个有丝分裂/2 mm分组时，死亡率与有丝分裂计数显示出最强的统计学意义。Ki67指数的10%阈值在预测不良预后方面更成功。在形态学变异中，导管岛状变异在我们的系列中最有希望具有阳性预后价值，尽管未检测到统计学意义。总之，肿瘤大小的5 cm和10 cm阈值、Ki67增殖指数的10%、有丝分裂计数的10/2 mm以及分别评估的血管和淋巴管侵犯是panNETs更好分层的潜在预后指标。
