1261P - Comparison of outcomes between responders and nonresponders to first-line paclitaxel/carboplatin (P/C) doublet chemotherapy in patients (Pts) with...

Date 27 September 2014
Event ESMO 2014
Session Poster Display session
Topics Anticancer agents
Non-small-cell lung cancer
Biological therapy
Presenter Vera Hirsh
Citation Annals of Oncology (2014) 25 (suppl_4): iv426-iv470. 10.1093/annonc/mdu349
Authors V. Hirsh1, C. Langer2, F. Ju-Lin3, Y. Wan3, I. Okamoto4, S. Whiting5, T.J. Ong6, M. Botteman7
  • 1Royal Victoria Hospital, McGill University Health Centre, H3A 1A1 - Montreal/CA
  • 2Abramson Cancer Center, University of Pennsylvania, Philadelphia/US
  • 3Heor, Pharmerit International, Bethesda/US
  • 4Medical Oncology, Kinki University School of Medicine, Osakasayama/JP
  • 5Health Economics, Celgene, Summit/US
  • 6Oncology, Celgene Corporation, Summit/US
  • 7Outcomes Research, Pharmerit International, Bethesda/US



A randomized phase III trial showed that nab-P/C significantly improved the objective overall response rate (ORR) vs solvent-based P/C as first-line therapy in advanced NSCLC. This analysis aimed to understand the impact of achieving response, regardless of treatment, on survival and quality-of-life outcomes in the context of this clinical trial.


Pts with and without at least an objective confirmed partial response were compared using the Q-TWiST (quality-adjusted time without symptoms or toxicity) method, which estimates the quality-adjusted survival by incorporating toxicity, progression-free survival (PFS) and overall survival (OS) into a single measurement. In this approach, OS was partitioned into time: with grade ≥ 3 toxicity (TOX), without symptoms of progression or toxicity (TWiST), and after progression (REL). Mean Q-TWiST was calculated by taking the sum of the product of the time spent in each state by its respective utility. In the base case, the utilities were: TWiST = 1, TOX = 0.5, REL = 0.5. In sensitivity analyses, TOX and REL utilities varied from 0 to 1.


In total, 1052 pts with stage IIIB/IV NSCLC were included in the analysis (median age 60 y). While responder status was not significantly associated with demographics, smoking status, or baseline performance status, responders (n = 302) were more likely than nonresponders (n = 750) to have squamous cell carcinoma (P = 0.043) or stage IIIB NSCLC at baseline (P = 0.006) and to receive more chemotherapy cycles/doses (P < 0.001). Responders vs nonresponders experienced significantly longer median OS (19.8 vs 9.2 months, P < 0.001) and PFS (9.6 vs 4.4 months, P < 0.001). In the base case, responders gained 5.3 months of mean Q-TWiST (13.8 vs 8.5 months; 95% CI for the difference, 4.4 - 6.1). The Q-TWiST difference ranged from 4.1 to 6.5 months in sensitivity analysis and from 4.8 to 5.6 months when stratified by histology and disease stage.


Tumor response is significantly associated with better outcomes, including quality-adjusted survival. Therefore, tumor response can be an important surrogate for assessment in treatment outcomes in NSCLC.


V. Hirsh: Honoraria for Celgene Advisory Board; C. Langer: Research support and advisory boards with Celgene; F. Ju-Lin: I am an employee of Pharmerit International which received research funding from Celgene for this project; Y. Wan: Yin Wan is an employee of Pharmerit International who received research funding from Celgene for this project; S. Whiting: Employment or leadership position and stock ownership, Celgene Corporation; T.J. Ong: TJ Ong is an employee of Celgene and owns stock; M. Botteman: Shareholder and employee of Pharmerit International which received research funding from Celgene Corporation for this project. All other authors have declared no conflicts of interest.