Progression-free survival and one-year milestone survival as surrogates for overall survival in previously treated advanced non-small cell lung cancer

Shen Zhao; Zhonghan Zhang; Yaxiong Zhang; Shaodong Hong; Ting Zhou; Yunpeng Yang; Wenfeng Fang; Hongyun Zhao; Li Zhang

doi:10.1002/ijc.31995

Progression-free survival and one-year milestone survival as surrogates for overall survival in previously treated advanced non-small cell lung cancer

Int J Cancer. 2019 Jun 1;144(11):2854-2866. doi: 10.1002/ijc.31995. Epub 2019 Jan 7.

Authors

Shen Zhao^{1

2

3

4}, Zhonghan Zhang^{1

2

3

4}, Yaxiong Zhang^{1

2

3}, Shaodong Hong^{1

2

3}, Ting Zhou^{1

2

3}, Yunpeng Yang^{1

2

3}, Wenfeng Fang^{1

2

3}, Hongyun Zhao^{1

2

3}, Li Zhang^{1

2

3}

Affiliations

¹ Department of Medical Oncology, Sun Yat-sen University Cancer Center, Guangzhou, China.
² State Key Laboratory of Oncology in South China, Guangzhou, China.
³ Collaborative Innovation Center for Cancer Medicine, Guangzhou, China.
⁴ Zhongshan School of Medicine, Sun Yat-Sen University, Guangzhou, China.

PMID: 30430561
DOI: 10.1002/ijc.31995

Abstract

The advent of immunotherapy leads to greater availability of effective subsequent treatments and extended survival in previously treated advanced non-small cell lung cancer (NSCLC), complicating the evaluation of overall survival (OS) in second-line NSCLC trials. Here, we aimed to assess the surrogacy of progression-free survival (PFS) and milestone survival for OS in second-line NSCLC trials investigating chemotherapy, targeted therapy and immunotherapy, respectively. We systemically searched for active-controlled, second-line NSCLC trials. The milestone time point was set at one-year based on pre-analysis. A two-stage meta-analytic validation model was adopted to assess associations between surrogate endpoint (SE) and OS and associations between treatment effects on SE and OS. Treatment effects on SE and OS were expressed as PFS hazard ratios (HR_PFS ), 1 yr-milestone ratio (Ratio_1y-SUR ) and HR_OS . Subgroup analyses stratified by treatment types and trial publication years evaluated the surrogacy in different clinical contexts. The study included 50 trials with 22,804 patients. One-year survival strongly correlated with OS (R² [95% confidence interval]: one-year survival -median OS = 0.707 [0.704-0.708]; Ratio_1y-SUR -HR_OS = 0.829 [0.828-0.831]). No correlation was established between PFS and OS (median PFS-median OS = 0.100 [0.098-0.101]; HR_PFS -HR_OS = 0.064 [0.059-0.069]), except in immunotherapy subgroup (HR_PFS -HR_OS = 0.835 [0.791-0.918]). In subgroup analyses, surrogacy of one-year survival persisted in different clinical contexts, and the disassociation between PFS and OS persisted in recent trials. One-year milestone survival showed strong surrogacy for OS in second-line NSCLC trials. Although no association was identified between PFS and OS, the strong HR_PFS -HR_OS correlation in immunotherapy trials indicates the potential of PFS as a SE in NSCLC trials involving immunotherapies.

Keywords: advanced non-small cell lung cancer; milestone survival; progression-free survival; surrogate endpoint.

Publication types

Evaluation Study
Meta-Analysis
Research Support, Non-U.S. Gov't

MeSH terms

Antineoplastic Agents / therapeutic use*
Carcinoma, Non-Small-Cell Lung / mortality
Carcinoma, Non-Small-Cell Lung / therapy*
Humans
Immunotherapy / methods*
Lung Neoplasms / mortality
Lung Neoplasms / therapy*
Progression-Free Survival*
Randomized Controlled Trials as Topic
Survival Analysis
Time Factors

Substances

Antineoplastic Agents