正电子发射计算机体层成像-CT预测表皮生长因子受体酪氨酸激酶抑制剂对非小细胞肺癌疗效及预后的价值

Early predictive and prognostic value of 18^F-fluorodeoxyglucose positron emission tomography-CT for response assessment in non-small cell lung cancer treated with epidermal growth factor receptor tyrosine kinase inhibitor

目的 探讨18氟-脱氧葡萄糖（18F-FDG）PET-CT在早期预测表皮生长因子受体酪氨酸激酶抑制剂（EGFR-TKI）对非小细胞肺癌（NSCLC）治疗有效性及预后中的应用价值.方法 2009年8月至2015年4月行EGFR-TKI治疗的NSCLC患者22例,分别于治疗前及治疗1个月后行PET-CT检查,以18F-FDG摄取最高病灶作为靶病灶（总数不超过5个,每个器官不超过2个）,分别测量治疗前后最大标准化摄取值（SUVmax）,并计算SUVmax变化率（ΔSUV%）,参照实体瘤治疗疗效PET评价标准（PERCIST）评估代谢改变.此后每2个月一次CT随访,以实体瘤疗效评价标准1.1版（RECIST 1.1）评价疗效.应用Kappa一致性检验分析基于PERCIST标准的PET-CT疗效评价与基于RECIST标准的疗效评价结果 的一致性.应用Fisher精确检验比较治疗早期代谢缓解组与无缓解组中出现疾病进展的概率.通过ROC曲线分析获得ΔSUV%预测治疗敏感或病变无进展的临界值和预测效能.应用Kaplan-Meier法进行生存分析,采用log-rank检验比较组间无进展生存期差异.结果 EGFR-TKI治疗1个月后,12例患者代谢部分缓解（PMR）,6例代谢稳定（SMD）,4例代谢进展（PMD）.治疗1个月后代谢评估结果 与治疗3个月后CT疗效评价结果 具有中等度的一致性（Kappa=0.506,P〈0.05）;治疗早期表现为代谢缓解（PMR）的患者出现疾病进展的概率显著低于代谢无缓解者（χ^2=11.941,P=0.005）;而治疗早期表现为代谢进展（PMD）的患者均出现了疾病进展;以ΔSUV%=40.36%为界值,预测治疗敏感的敏感度和特异度分别为88.9%和84.6%,曲线下面积为0.906（95%置信区间：0.766~1.000,P=0.002）.以ΔSUV%=25.84%为界值对患者分组进行生存分析,组间PFS差异有统计学意义（P=0.026）.结论 EGFR-TKI治疗1个月时行18F-FDG PET-CT检查可用于早期预测NSCLC治疗疗效及预后.

Objective To evaluate whether an early change in 18^F-fluorodeoxyglucose （18^F-FDG） uptake can predict tumor response to epidermal growth factor receptor tyrosine kinase inhibitor （EGFR-TKI） and prognosis in patients with non-small cell lung cancer （NSCLC）. Methods From August 2009 to April 2015, 22 patients with NSCLC who were eligible to EGFR-TKI treatment were enrolled. PET-CT scan was performed before （baseline） and 1 month after EGFR-TKI administration. Up to 5 hottest single tumor lesions （no more than 2 per organ） were considered to be target lesions. Maximum standardized uptake values （SUVmax） were measured, and post-treatment percentage changes in SUVmax （ΔSUV%） were calculated. PET responses were classified using PET response criteria in solid tumors （PERCIST）. Then conventional CT scan was performed every 2 months for follow-up. Kappa statistic was used to compare agreement between the RERCIST recommendations-based therapeutic response evaluation and those based on RECIST1.1 criteria. Fisher exact test was used to compare the probability of disease progression in the early metabolic response and non-response groups. Predictive accuracy of ΔSUV% with respect to response or non-progression at CT scan was evaluated by ROC analysis. Progression-free survival （PFS） was determined by Kaplan-Meier survival analysis, and between-group comparison was performed by log-rank test. Results After 1 month of EGFR-TKI treatment, 12 patients （55%） showed partial metabolic response （PMR）, 6 （27%） had stable metabolic disease （SMD）, and 4 （18%） had progressive metabolic disease （PMD）. There was a moderate agreement（Kappa=0.506,P〈0.05） between PET response at 1 month based on PERCIST recommendations and CT response at 3 months according to RECIST 1.1. Non-progression was significantly more frequent in patients with an early PMR （χ2=11.941, P=0.005）. Progression had been confirmed later during therapy in all patients with PMD . By using ROC analysis, the area under the curve for prediction of response was 0.906 （95% CI, 0.766—1.000; P=0.002）, corresponding to a sensitivity of 88.9% and specificity of 84.6% at a cut-off of 40.36% in ΔSUV%. Using a cut-off value of 25.84% in ΔSUV%, highΔSUV% group （ΔSUV% ≥ 25.84%） had significantly longer PFS than low ΔSUV% group （ΔSUV%〈25.84%）. Conclusion Early assessment of PET-CT at 1 month of EGFR-TKI treatment could be useful to predict tumor response and clinical outcome in patients with NSCLC.