胃肠道间质瘤患者生存和预后因素综合分析

Comprehensive analysis of factors that involve in survival and prognosis in patients with gastrointestinal stromal tumors

目的探讨影响胃肠道间质瘤（GIST）患者生存和预后的因素。方法复阅153例患者切片，以免疫组化法检测L、D117、CD34、血小板衍生生长因子受体-α和Ki-67蛋白表达，结合l临床病理特征和GIST生物学行为分级，分析影响GIST患者生存和预后的相关因素。采用卡普兰-迈耶（Kaplan-Meier）法和Cox比例风险模型比较不同因素对生存的影响。结果患者1、3、5年生存率分别为94．1％、76，3％和65．9％。单因素分析显示，患者预后与肿瘤直径（X^2=40.565，P〈0.01）、肿瘤部位（X^2=3.245，P〈0.01）、核分裂象数目（x^2=22.626，P〈0．01）、危险度分级（X^2=19.186，P〈0.01）、肿瘤有无坏死（x^2=28.665，P〈0.01）、手术方式（X^2=29.110，P〈0.01）和Ki-67指数（X^2=15.953,P〈0．01）有关。多因素分析表明，肿瘤直径〉10cm、位于肠道、核分裂象数目〉10／50HPF、危险度分级属高度危险性、肿瘤有坏死及Ki-67指数〉5％与预后呈明显负相关。且Ki-67指数、肿瘤大小及核分裂象数目是GIST预后的强预告因子。结论GIST生物学行为分级法对评价GIST、患者预后具有较好的临床价值；判断GIST患者预后需结合Ki-67指数和肿瘤部位等因素，为治疗提供依据。

Objective To investigate the survival and prognostic factors in patients with gastrointestinal stromal tumors（GIST）. Methods The histopathological slides from 153 patients with GIST were reviewed. The expression of CD117, CD34, platelet-derived growth factor receptor alpha （PDGFR- α） and Ki-67 proteins were measured by immunohistochemical staining. The factors that involved in the survival and prognosis were analyzed based on the clinical features and GIST biological behavior ranking. The Kaplan-Meier and Cox model were used to evaluate the effect of variant factors 0n survival and prognosis. Results The survival rateuf 135 patients was 94.1%, 76.3% and 65.9% at 1, 3 and 5 years, resprctivcly, On univariate analysis survival was predicted by tumor size （X^2=40.565, P〈0.01）. primary tumor location （X^2=13.245, P〈0.01）, mitotic count （x^2=22.626, P〈0.01 ）, risk ranking （x^2= 19.186, P〈0.01）, necrosis （X^2=28.665, P〈0.01）, incomplete resection （x^2 =29.110, P〈0.01） and Ki-67 index （X^2=15.953, P〈0.01）. Multivariate analysis demonstrated that the tumor size 〉 10 cm, primary tumor location, mitotic count〉 10/50 HPF, high risk subgroup, tumor necrosis and Ki-67 index〉5% were poor predictors of survival. Ki-67, tumor size and mitotic count were strong poor predictors of survival. Conclusions Fletcher＇s biological behavior ranking is a good approach to predict prognosis of GIST patients and has significant clinical value. It＇s better to combine it with other factors such as Ki-67 index and primary tumor location in order to provide evidence for therapy.