Sawaki Akira, Kanemitsu Yukihide, Mizuno Nobumasa, Takahashi Kuniyuki, Nakamura Tsuneya, Ioka Tatsuya, Tanaka Sachiko, Nakaizumi Akihiko, Salem Ahmed As, Ueda Ryuzo, Yamao Kenji
Department of Gastroenterology, Aichi Cancer Center Hospital, Chikusa-ku, Nagoya, Japan.
J Gastroenterol Hepatol. 2008 Aug;23(8 Pt 1):1292-7. doi: 10.1111/j.1440-1746.2006.04734.x.
The aim of this study was to identify factors that predict treatment outcome in patients with metastatic pancreatic cancer treated with gemcitabine, and then to use these factors to develop a practical prognostic index.
A retrospective study was performed on 66 consecutive patients with histologically confirmed pancreatic adenocarcinoma who were treated with gemcitabine. Factors that predicted treatment outcome were identified by univariate and multivariate analyses using the Cox proportional hazards model.
Multivariate analysis identified Eastern Cooperative Oncology Group performance status, primary tumor location, and C-reactive protein as important independent predictive factors. Prognostic score was calculated using the following formula: score = (1 if performance status is 0 or 1; 2 if performance status is 2; and 5 if performance status is 3) + (1 if primary site is body or tail, 3 if primary site is head) + (1 if C-reactive protein is <1 mg/dL, 3 if C-reactive protein is 1-3 mg/dL, 6 if C-reactive protein is >3 mg/dL). Patients were accordingly divided into three groups: good (prognostic index = 3 or 4), fair (prognostic index = 5-7), and poor (prognostic index = 8). Median survival was 265, 155, and 65 days for each group, respectively (P < 0.0001). The internally validated c-index (receiver operating characteristics area under the curve) of this model was 0.711. Applied to another data set, the externally validated c-index was 0.692. Prognosis was favorable in the good prognosis group, patients in the fair prognosis group were likely to benefit from gemcitabine, and those in the poor prognosis group were unlikely to benefit.
This index improved predictive ability in patients with metastatic pancreatic cancer treated with gemcitabine, which may be helpful in counseling patients and making first treatment decisions.
本研究旨在确定预测接受吉西他滨治疗的转移性胰腺癌患者治疗结局的因素，然后利用这些因素制定一个实用的预后指数。
对66例经组织学确诊为胰腺腺癌并接受吉西他滨治疗的连续患者进行回顾性研究。使用Cox比例风险模型通过单因素和多因素分析确定预测治疗结局的因素。
多因素分析确定东部肿瘤协作组（Eastern Cooperative Oncology Group）体能状态、原发肿瘤位置和C反应蛋白是重要的独立预测因素。使用以下公式计算预后评分：评分 = （如果体能状态为0或1则为1；如果体能状态为2则为2；如果体能状态为3则为5）+（如果原发部位为体部或尾部则为1，如果原发部位为头部则为3）+（如果C反应蛋白<1mg/dL则为1，如果C反应蛋白为1 - 3mg/dL则为3，如果C反应蛋白>3mg/dL则为6）。患者据此分为三组：良好（预后指数 = 3或4）、中等（预后指数 = 5 - 7）和差（预后指数 = 8）。每组的中位生存期分别为265天、155天和65天（P < 0.0001）。该模型的内部验证c指数（曲线下面积）为0.711。应用于另一个数据集时，外部验证c指数为0.692。预后良好组的预后较好，中等预后组的患者可能从吉西他滨治疗中获益，而预后差组的患者不太可能获益其。
该指数提高了接受吉西他滨治疗的转移性胰腺癌患者的预测能力，这可能有助于为患者提供咨询并做出初始治疗决策。
