吉非替尼治疗非小细胞肺癌的临床疗效预测模型的初步建立

目的 尝试建立预测吉非替尼治疗晚期非小细胞肺癌（NSCLC）临床疗效的模型。方法 对中国医学科学院肿瘤医院过去4年间（2002年7月至2007年3月）应用吉非替尼治疗的262例晚期NSCLC患者的临床结果进行回顾性分析,并对其中有足够肿瘤组织的55例进行了EGFR基因突变的检测（DNA测序）。在综合分析影响疗效主要因素的基础上,建立模型。结果 吉非替尼治疗晚期NSCLC有效率为30．1％,疾病控制率为78．6％;中位无疾病进展时间（PFS）和中位生存期分别为6个月和16个月;1、2、3年生存率分别为60．8％、35．6％和18．3％。24例（47.3％）发生EGFR基因突变患者的治疗有效率（50．0％）明显高于无突变者（16．1％）,差异有统计学意义（χ^2=7．27,P=0.009）。吉非替尼耐受性好,主要不良反应为皮疹,腹泻。单因素分析显示：女性、年龄〈65岁、腺癌、不吸烟、紫杉类未耐药和有基因突变患者对吉非替尼治疗更敏感。但COX多因素回归分析显示,只有年龄、病理类型、吸烟状况和基因突变影响吉非替尼的疗效。以这4个因素为基本元素建立疗效预测模型,以2分为界值,〉2分的患者有效率（34．2％）明显高于≤2分患者（9．3％）,差异有统计学意义（χ^2=10．619,P=0．001）;每增加1分,疗效增长1倍。结论初步建立了一个能有效预测吉非替尼治疗NSCLC临床疗效的模型,对选择合适患者具有一定参考价值。

Establishment of a model to assess the clinical efficacy of Gefitinib in treatment of non-small cell lung cancer

OBJECTIVE: To establish a model to predict the clinical response of Gefitinib in non-small cell lung cancer (NSCLC). METHODS: The clinical outcomes of 262 consecutive advanced NSCLC patients to oral treatment of gefitinib 250 mg daily in the past 4 years were reviewed. DNA sequencing was used to detect the mutations in the exons 18, 19, 20, and 21 of the epidermal growth factor receptor (EGFR) tyrosine kinase domain in 55 patients who had enough tumor tissues. RESULTS: The response rate and disease control rate of gefitinib in the advanced NSCLC patients were 30.1% and 78.6% respectively. The median progression free survival and median overall survival were 6.0 and 16.0 months respectively, while the 1, 2 and 3-year survival rates of the NSCLC patients were 60.8%, 35.6%, and 18.3% respectively. The clinical response rate of the patients with EGFR mutation was 50.0%, significantly higher than that of those patients without mutation (16.1%; P = 0.009). In the majority of patients Gefitinib was well tolerated, and the common adverse effects were skin rash and diarrhea. Based on the results of our patients, we try to establish a model which may predict the response of patients by logistic multivariate regression analysis. Patients being female aged under 65, with adenocarcinoma, not smoking, taxone-unresistant and with EGFR mutation were more sensitive to gifitinib. However, COX multivariate regression analysis showed that only the age, histology, smoke status and EGFR mutation were valuable in selection of sensitive patients. So, we set up the cut-off value at 2 in the model based on these 4 factors. The response rate of the patients with 2 or more scores was 34.2%, significantly higher than that of the patients with the scores < 2 (9.3%, P = 0.001). The sensitivity would be doubled when the score increases by one point. CONCLUSION: A simple clinical model based on the patients' age, histology, smoking status and EGFR mutation has been established and is useful to determine the efficacy of gifitinib in NSCLC patients.