Borges M, Sculier J P, Paesmans M, Richez M, Bureau G, Dabouis G, Lecomte J, Michel J, Van Cutsem O, Schmerber J, Giner V, Berchier M C, Sergysels R, Mommen P, Klastersky J
Service de Médecine, Institut Jules Bordet, Bruxelles, Belgium.
Lung Cancer. 1996 Dec;16(1):21-33. doi: 10.1016/s0169-5002(96)00609-5.
To identify pretreatment variables predicting response to platinum derivatives containing chemotherapy in patients with unresectable non-small cell lung cancer (NSCLC).
Eligible patients included in one of the 7 consecutive clinical trials conducted by the European Lung Cancer Working Party between December 1980 and August 1991. All patients received a cisplatin or carboplatin containing chemotherapy. We analyzed 22 potential prognostic factors including sex, age, histology, performance status, weight loss, type of lesions, extent of disease, main metastatic sites and several biological parameters, namely white blood cell count (WBC), neutrophil count, platelet count, hemoglobinemia, creatininemia, serum alkaline phosphatases and LDH.
On 1052 eligible patients. 107 were not assessable for response. The objective response rate was 26% (95% C.I.: 23, 29%). Univariate analysis identified as statistically significantly associated with a higher objective antitumoral response rate the following characteristics: a normal platelet count, the absence of skin metastasis, the absence of adrenal metastasis, a higher creatininemia, a normal hemoglobinemia, an older age and a normal WBC count. On a restricted set of variables including data from 777 patients, a multivariate logistic regression model disclosed age and platelet count as significantly and independently related to response rate.
Clinical and demographic characteristics of patients with unresectable NSCLC, as well as routine laboratory parameters, could not accurately predict response to chemotherapy in a population of patients selected for a clinical trial. Future studies on this subject should include more sophisticated variables as new biomolecular makers.
确定不可切除的非小细胞肺癌（NSCLC）患者对含铂类衍生物化疗反应的预处理变量。
符合条件的患者纳入欧洲肺癌工作组在1980年12月至1991年8月期间进行的7项连续临床试验之一。所有患者均接受含顺铂或卡铂的化疗。我们分析了22个潜在的预后因素，包括性别、年龄、组织学类型、体能状态、体重减轻、病变类型、疾病范围、主要转移部位以及几个生物学参数，即白细胞计数（WBC）、中性粒细胞计数、血小板计数、血红蛋白血症、肌酐血症、血清碱性磷酸酶和乳酸脱氢酶。
在1052例符合条件的患者中，107例无法评估反应。客观缓解率为26%（95%置信区间：23，29%）。单因素分析确定以下特征与较高的客观抗肿瘤反应率在统计学上显著相关：血小板计数正常、无皮肤转移、无肾上腺转移、肌酐血症较高、血红蛋白血症正常、年龄较大和白细胞计数正常。在一组受限变量（包括来自777例患者的数据）上，多因素逻辑回归模型显示年龄和血小板计数与缓解率显著且独立相关。
不可切除NSCLC患者的临床和人口统计学特征以及常规实验室参数，无法准确预测入选临床试验患者群体对化疗的反应。关于该主题的未来研究应纳入更复杂的变量作为新的生物分子标志物。
