XALKORI® (crizotinib): Final Overall Survival (OS) Analysis from PROFILE 1014

Background

The first PROFILE 1014 results were published in 2014 after a median follow-up of 17 months. The study met its primary endpoint showing the superiority of XALKORI to pemetrexed-plus-platinum chemotherapy for progression-free survival, as assessed by independent radiologic review. Median overall survival (OS) was not reached in either group at the time of the first report.1 For the original study design and data, please see the clinical efficacy in ALK+ NSCLC page.

Study Objectives

To investigate the secondary endpoints of OS and safety in XALKORI versus pemetrexed-plus-platinum chemotherapy over an additional 3 years of follow-up to the initial PROFILE 1014 results.

Results

Adapted from Soloman et al 20182.  a Estimated by Cox proportional hazards regression analysis with adjustment for ECOG PS, race, brain metastases;  b 2-sided p-value from the log-rank test stratified by ECOG PS, race, brain metastases.

  • 4-year OS rates were 56.6% for XALKORI and 49.1% for pemetrexed-plus-platinum chemotherapy (HR: 0.760; 95% CI: 0.548, 1.053; p=0.0978). This difference did not reach statistical significance.2

Although statistical significance was not reached, 84.2% of patients in the chemotherapy arm crossed over to XALKORI as their first subsequent follow-up treatment and 19.2% of patients in the XALKORI arm crossed over to platinum-based chemotherapy.2 The rank-preserving structural failure time model (RPSFTM) was used to analyse OS adjusted for crossover.

Analysis: OS Adjusted for crossover using RPSFTM3

RPSFTM applies a parameter to relate the observed OS to its hypothetical OS, i.e. what would have been observed without crossover to other treatments.

  • Due to randomization, the hypothetical OS without XALKORI treatment was expected to be the same between two arms, taking into account both the XALKORI randomized treatment phase and the crossover phase
  • Based on the final OS data, the estimated parameter demonstrated survival advantage for XALKORI: survival time was shorter for patients initially randomized to chemotherapy and who crossed over to XALKORI, and survival time was prolonged for patients initially randomized to XALKORI and who crossed over to chemotherapy
  • Hence, the observed survival times for patients who crossed over were corrected to the survival time that would have been observed if there was no crossover

Use of RPFSTM allows an unbiased estimation of treatment effect on OS as if there was no crossover

Assumptions:

  • The treatment effect of an experimental treatment received by patients who crossover must be the same as the treatment effect in patients initially randomized to that treatment (“Independent of line of treatment”)
  • The effect of other subsequent systemic therapies on OS is the same between the 2 arms

Final OS adjusted for crossover by RPSFTM

 Adapted from Soloman et al 20182. a Calculated using the stratified Cox proportional hazards regression approach applied to crossover adjusted data by a parameter derived using the stratification log-rank test and its bootstrap CI.

  • After adjusting for crossover using the RPFSTM , OS in the XALKORI arm was longer than in the chemotherapy arm (HR, 0.346 [95% bootstrap CI, 0.081–0.718])2

Impact of subsequent therapy on OS: ALK TKI vs treatment other than ALK TKI

Adapted from Soloman et al 20182. Patients who did not receive subsequent therapy after XALKORI or chemotherapy, and patients still on XALKORI treatment at the end of the study are not shown.

  • Patients who received XALKORI followed by another ALK TKI were found to have a longer OS than those receiving XALKORI followed by any follow-up therapy other than an ALK TKI, as well as those patients who received chemotherapy first-line.2

Conclusions

  • The safety profile for XALKORI was consistent with previously reported PROFILE 1014 data.  The most frequent adverse events reported for XALKORI were vision disorder (73.1%), diarrhoea (65.5%), nausea (59.1%) and oedema (52.6%)2
  • The difference in OS between the XALKORI and chemotherapy arm (HR= 0.76; two-sided p= 0.978) did not reach statistical significance, however, after adjusting for crossover using the RPFSTM , OS in the XALKORI arm was longer than in the chemotherapy arm (HR, 0.346 [95% bootstrap CI, 0.081–0.718])2
  • Patients who received XALKORI followed by another ALK TKI were found to have a longer OS than those receiving XALKORI followed by any follow-up therapy other than  an ALK TKI, as well as those patients who received chemotherapy first-line.2
  • The magnitude of OS seen in both treatment arms of this trial is noteworthy for patients with advanced, ALK-positive NSCLC2

 

CI, confidence interval; HR, hazard ratio; ITT, intention to treat; NSCLC, non-small cell lung cancer; TKI, tyrosine kinase inhibitor.

References:
1. Solomon BJ, et al. First-line crizotinib  versus chemotherapy in ALK-positive lung cancer. N Engl J Med. 2014:371(23):2167-2177
2. Solomon BJ et al. Final overall survival analysis from a study comparing first-line crizotinib with chemotherapy: results from PROFILE 1014, J Clin Oncol 2018. JCO2017774794
3. Robins JM, et al. Commun Stat Theory Methods 1991;20(8):2609–31

PP-XLK-GBR-1006. March 2019