目的: 探讨术前血小板参数对局部进展期肾癌预后的预测价值,为此类患者的危险分层提供参考。方法: 选择北京大学第三医院2015年1月—2017年12月局部进展期肾癌患者进行回顾性分析,依据随访过程中肿瘤是否复发或转移分成进展组和无进展组,比较两组术前血小板参数和临床资料的差异。通过受试者特征工作曲线(receiver operating characteristic curve, ROC)确定血小板参数的最佳临界值,Kaplan-Meier生存曲线分析不同血小板参数与疾病无进展生存时间(progression-free survival, PFS)的关系。通过Cox比例风险模型进行多因素分析确定PFS的独立危险因素。采用时间依赖ROC、净重新分类指数(net reclassification index, NRI)和综合判别改善指数(integrated discrimination improvement, IDI)评估纳入血小板参数后对SSIGN模型改良情况。结果: 共有215例患者入选本研究,其中192例(89.3%)患者获得随访,中位随访时间为36个月。64例(29.8%)患者随访过程中出现疾病进展,中位PFS为46个月。进展组患者的血小板数量(platelet count, PLT)相较无进展组高[(250.72±88.59)×10⁹/L vs. (227.27±66.94)×10⁹/L, P=0.042],血小板分布宽度(platelet distribution width, PDW)相较无进展组低[(12.01±2.27)% vs. (13.31±2.74)%, P=0.001]。将285×10⁹/L及12.65%作为PLT及PDW的最佳临界值,PLT≤285×10⁹/L组患者中位PFS显著长于PLT>285×10⁹/L组(53个月vs. 41个月,P=0.033);PDW>12.65%组患者中位PFS也显著长于PDW≤12.65%组(56个月vs. 41个月,P<0.001)。多因素分析显示术前PDW(HR=0.735, P<0.001)、细胞核分级Ⅲ~Ⅳ级(HR=2.425, P=0.001)、合并肉瘤样分化(HR=3.101, P=0.008)为PFS的独立危险因素。术前PDW联合SSIGN预后评分模型曲线下面积大于原有SSIGN模型[0.748 (95%CI:0.662~0.833) vs. 0.678 (95%CI: 0.583~0.773), P=0.193],NRI为0.262(P=0.04),IDI为 0.085(P=0.01),表明PDW纳入SSIGN模型后其预测能力提高。结论: 术前高PLT和低PDW与局部进展期肾癌不良预后相关,其中PDW是患者预后的独立危险因素,因此,术前PDW有助于对局部进展期肾癌进行危险分层。

Objective: To explore the prognostic value of preoperative platelet parameters in locally advanced renal cell carcinoma for the risk stratification of such patients. Methods: Clinical data of patients with locally advanced renal cell carcinoma in the Third Hospital of Peking University from January 2015 to December 2017 were collected. The patients were divided into progression group and progression-free group according to follow-up data, and preoperative platelet parameters and clinical data between the two groups were compared. The optimal cut-off value of platelet parameters was determined by receiver operating characteristic curve (ROC) and analyzed by Kaplan-Meier survival curve. Cox proportional hazards model was used to analyze the independent risk factors of PFS. Time dependent ROC curve, net reclassification index (NRI), and integrated discrimination improvement (IDI) were used to evaluate the improvement of SSIGN model by incorporating platelet parameters. Results: Of the 215 patients, 192 (89.3%) were followed up for a median of 36 months. Sixty-four patients (29.8%) had disease progression during the follow-up, and the median PFS was 46 months. In progression group, the platelet count (PLT) was higher [(250.72 ± 88.59)×10⁹/L vs. (227.27 ± 66.94)×10⁹/L, P=0.042] and the platelet distribution width (PDW) was lower [(12.01 ± 2.27)% vs. (13.31 ± 2.74)%, P = 0.001] than that of progression-free groups. 285×10⁹ /L and 12.65% as the best cut-off values of PLT and PDW, the median PFS of PLT≤285×10⁹ /L group was significantly longer than that of PLT>285×10⁹ /L group (53 months vs. 41 months, P=0.033), and the median PFS of PDW>12.65% group was also significantly longer than that of PDW≤12.65% group (56 months vs. 41 months, P<0.001). Multivariate analysis showed that preoperative PDW (HR=0.735, P<0.001), nuclear grade Ⅲ to Ⅳ (HR=2.425, P=0.001) and sarcomatoid differentiation (HR=3.101, P=0.008) were independent risk factors for PFS. The area under the curve of PDW combined with SSIGN model was larger than that with the original SSIGN model [0.748 (95%CI:0.662-0.833) vs. 0.678 (95%CI: 0.583-0.773), P=0.193], NRI was 0.262 (P=0.04), and IDI was 0.085 (P=0.01), indicating that the predictive ability of PDW combined with SSIGN model was improved. Conclusion: Preoperative high PLT and low PDW are associated with adverse prognosis of locally advanced renal cell carcinoma, and PDW is an independent risk factor. Therefore, preoperative PDW could serve as biomarker for risk stratification of locally advanced renal cell carcinoma.