Bass Michael B, Yao Bin, Hei Yong-Jiang, Ye Yining, Davis Gerard J, Davis Michael T, Kaesdorf Barbara A, Chan Sabrina S, Patterson Scott D
Molecular Sciences and Computational Biology, Amgen Inc., Thousand Oaks, CA, United States of America.
Global Biostatistical Science, Amgen Inc., Thousand Oaks, CA, United States of America.
PLoS One. 2014 Oct 14;9(10):e108048. doi: 10.1371/journal.pone.0108048. eCollection 2014.
We sought to develop placental growth factor as a predictive pharmacodynamic biomarker for motesanib efficacy as first-line therapy in patients with advanced nonsquamous non-small-cell lung cancer.
Placental growth factor was evaluated at baseline and study week 4 (after 3 weeks treatment) in an exploratory analysis of data from a randomized phase 2 study of motesanib 125 mg once daily plus carboplatin/paclitaxel and in a prespecified analysis of data from a randomized, double-blind phase 3 study of motesanib 125 mg once daily plus carboplatin/paclitaxel vs placebo plus carboplatin/paclitaxel (MONET1). Associations between fold-change from baseline in placental growth factor and overall survival were evaluated using Cox proportional hazards models.
In the phase 2 study, serum placental growth factor increased from baseline a mean 2.8-fold at study week 4. Patients with ≥2.2-fold change from baseline in placental growth factor (n = 18) had significantly longer overall survival than those with <2.2-fold change (n = 19; 22.9 vs 7.9 months; hazard ratio, 0.30; 95% CI, 0.12-0.74; P = 0.009). Consequently, placental growth factor was investigated as a pharmacodynamic biomarker in the phase 3 MONET1 study. There was no association between log-transformed placental growth factor fold-change from baseline to week 4 (continuous variable) and overall survival (hazard ratio, 0.98; 95% CI, 0.79-1.22; P = 0.868). MONET1 did not meet its primary endpoint of overall survival. Likewise, median overall survival was similar among patients with ≥2.0-fold change in placental growth factor (n = 229) compared with <2.0-fold change (n = 127; 14.8 vs 13.8 months; hazard ratio, 0.88; 95% CI, 0.67-1.15, P = 0.340).
Our results illustrate the challenges of successfully translating phase 2 biomarker results into phase 3 studies.
ClinicalTrials.gov NCT00460317, NCT00369070.
我们试图将胎盘生长因子开发为一种预测性药效学生物标志物，用于评估莫替沙尼作为晚期非鳞状非小细胞肺癌一线治疗药物的疗效。
在一项莫替沙尼125毫克每日一次联合卡铂/紫杉醇的随机2期研究的数据探索性分析中，以及在一项莫替沙尼125毫克每日一次联合卡铂/紫杉醇对比安慰剂联合卡铂/紫杉醇的随机、双盲3期研究（MONET1）的预设数据分析中，在基线和研究第4周（治疗3周后）评估胎盘生长因子。使用Cox比例风险模型评估胎盘生长因子相对于基线的变化倍数与总生存期之间的关联。
在2期研究中，血清胎盘生长因子在研究第4周时较基线平均升高了2.8倍。胎盘生长因子相对于基线变化≥2.2倍的患者（n = 18）的总生存期显著长于变化<2.2倍的患者（n = 19；22.9个月对7.9个月；风险比，0.30；95%置信区间，0.12 - 0.74；P = 0.009）。因此，在3期MONET1研究中对胎盘生长因子作为药效学生物标志物进行了研究。从基线到第4周经对数转换的胎盘生长因子变化倍数（连续变量）与总生存期之间无关联（风险比，0.98；95%置信区间，0.79 - 1.22；P = 0.868）。MONET1未达到其总生存期的主要终点。同样，胎盘生长因子变化≥2.0倍的患者（n = 229）与变化<2.0倍的患者（n = 127）的中位总生存期相似（14.8个月对13.8个月；风险比，0.88；95%置信区间，0.67 - 1.15，P = 0.340）。
我们的结果说明了将2期生物标志物结果成功转化为3期研究的挑战。
ClinicalTrials.gov NCT00460317，NCT00369070。
