138例Ⅳ期胃癌者预后因素分析

Prognostic Factors for Survival of 138 Patients with Stage IV Gastric Cancer

目的:探讨接受以铂类为基础的联合化疗的Ⅳ期胃癌患者的预后因素。方法:回顾性分析伴有远处转移并接受一周期或者一周期以上的以铂类为基础联合化疗的Ⅳ期胃癌患者共138例,生存率的计算采用Kaplan-Meier法,各亚组生存率差别的比较采用log-rank时序检验,采用Cox比例分析模型作其对生存率影响的多因素分析,并根据其结果结合有意义的预后因素推导出预后指数的计算方法。结果:1)单因素及Cox比例分析模型多因素分析发现ECOG评分(P=0.001)、体重减轻(P=0.001)、浸润深度(P=0.000)、腹膜转移(P=0.005)、转移灶数目(P=0.029)和血清总胆红素升高(P=0.018)为影响生存期的独立预后因素(P<0.05)。2)根据Cox分析的结果,初步建立了Ⅳ期胃癌预后的预测公式。依据上述公式计算出患者的PI指数,有16例患者被分入低危(PI≤9.817),有28例患者被分入中危组(9.817<PI<12.256),还有8例患者被分入高危组(PI≥12.256)。三组6个月、12个月和24个月总生存率作比较。结论:本文筛选出ECOG评分、体重减轻、浸润深度、腹膜转移、转移灶的数目和血清胆红素水平升高是影响Ⅳ期胃癌患者预后的独立因素,可作为反映胃癌生物学行为和判断预后的有效指标。依据Cox分析结果初步建立Ⅳ期胃癌预后的预测公式,对临床医生判定预后及制定相应的治疗方案有一定的指导意义。

Objective： To evaluate the prognostic factors of patients with stage Ⅳ gastric cancer. Methods： A total of 138 patients with stage Ⅳ gastric cancer treated with platinum-based chemotherapy were analyzed. Survival rate was estimated by Kaplan-Meier method. The prognostic factors were analyzed by univariate （Log rank） and multivariate （Cox model） analysis methods. Results： Univariate analysis and multivariate analysis showed that poor performance status （P=0.001）, weight loss （P=0.001）, depth of invasion （P=0.000）, presence of peritoneal metastasis （≤=-0.005）, more than 1 metastatic site （P=0.029） and elevated total bilirubin （P=0.018） were independent prognostic factors. According to the outcomes of the Cox model analysis, a formula of the prognostic index was developed. According to the values of PI, 16 patients were categorized as the high-risk group （PI≤9.817）, 28 patients were categorized as the moderate-risk group （9.817≤PI≤12.256） and 8 patients were categorized as the poor-risk group （PI≥ 12.256）, respectively. The overall survival ratios of the 3 groups were compared. Conclusion： Poor performance status, depth of invasion, elevated total bilirubin, more than 1 metastatic site, presence of peritoneal metastasis and weight loss were independent prog- nostic factors. A formula of the prognostic index was developed and it could help clinicians in decision-making and treatment tailoring based on the prognosis evaluation.