Jacek Rutkowski1, Everardo D Saad2, Tomasz Burzykowski3, Marc Buyse4, Jacek Jassem5. 1. Medical University of Gdańsk, Gdańsk, Poland. 2. Dendrix Research, Sao Paulo, Brazil; International Drug Development Institute, Louvain-la-Neuve, Belgium. 3. International Drug Development Institute, Louvain-la-Neuve, Belgium; Interuniversity Institute for Biostatistics and statistical Bioinformatics (I-BioStat), Hasselt University, Diepenbeek, Belgium. 4. Interuniversity Institute for Biostatistics and statistical Bioinformatics (I-BioStat), Hasselt University, Diepenbeek, Belgium; International Drug Development Institute, San Francisco, California. 5. Medical University of Gdańsk, Gdańsk, Poland. Electronic address: jjassem@gumed.edu.pl.
Abstract
BACKGROUND: There is a debate about the merits of progression-free survival (PFS) versus overall survival (OS) as primary endpoints in NSCLC. It has been postulated that post-progression therapy may influence OS in both arms. To investigate this issue, we analyzed chronological trends in PFS and OS in advanced NSCLC using restricted mean survival times (RMSTs). METHODS: We digitized survival curves from first-line phase III trials published between 1998 and 2015 in 13 leading journals to compute RMSTs for PFS and OS at three truncation landmarks (5, 12, and 18 months). RESULTS: Among the 161 trials identified, RMSTs could be computed for both endpoints in 102, 97, and 82 trials for the 5-, 12-, and 18-month truncation landmarks, respectively. Post-progression survival in the control arm, quantified as mean OS minus mean PFS truncated at 18 months, was on average 3.3 months between 1998 and 2003, 4.4 months between 2004 and 2009, and 5.4 months between 2010 and 2015. This increase was due to increasing RMST for OS over time, with no increase in RMST for PFS. The average within-trial difference in RMSTs between experimental and control arm was close to 0 for OS and less than 1 month for PFS. CONCLUSIONS: There is a progressive increase in post-progression survival in NSCLC trials, likely from salvage therapy. These results question both PFS and OS as sensitive endpoints in first-line trials, but suggest that the outlook for patients is improving regardless of within-trial gains.
BACKGROUND: There is a debate about the merits of progression-free survival (PFS) versus overall survival (OS) as primary endpoints in NSCLC. It has been postulated that post-progression therapy may influence OS in both arms. To investigate this issue, we analyzed chronological trends in PFS and OS in advanced NSCLC using restricted mean survival times (RMSTs). METHODS: We digitized survival curves from first-line phase III trials published between 1998 and 2015 in 13 leading journals to compute RMSTs for PFS and OS at three truncation landmarks (5, 12, and 18 months). RESULTS: Among the 161 trials identified, RMSTs could be computed for both endpoints in 102, 97, and 82 trials for the 5-, 12-, and 18-month truncation landmarks, respectively. Post-progression survival in the control arm, quantified as mean OS minus mean PFS truncated at 18 months, was on average 3.3 months between 1998 and 2003, 4.4 months between 2004 and 2009, and 5.4 months between 2010 and 2015. This increase was due to increasing RMST for OS over time, with no increase in RMST for PFS. The average within-trial difference in RMSTs between experimental and control arm was close to 0 for OS and less than 1 month for PFS. CONCLUSIONS: There is a progressive increase in post-progression survival in NSCLC trials, likely from salvage therapy. These results question both PFS and OS as sensitive endpoints in first-line trials, but suggest that the outlook for patients is improving regardless of within-trial gains.
Authors: Richard M Goldberg; Richard Adams; Marc Buyse; Cathy Eng; Axel Grothey; Thierry André; Alberto F Sobrero; Stuart M Lichtman; Al B Benson; Cornelis J A Punt; Tim Maughan; Tomasz Burzykowski; Dirkje Sommeijer; Everardo D Saad; Qian Shi; Elisabeth Coart; Benoist Chibaudel; Miriam Koopman; Hans-Joachim Schmoll; Takayuki Yoshino; Julien Taieb; Niall C Tebbutt; John Zalcberg; Josep Tabernero; Eric Van Cutsem; Alastair Matheson; Aimery de Gramont Journal: J Natl Cancer Inst Date: 2022-06-13 Impact factor: 11.816