1338 - Retrospective efficacy and safety analysis of erlotinib (E), pemetrexed (P) and docetaxel (D) in previously treated non-small-cell lung cancer patie...

Date 28 September 2012
Event ESMO Congress 2012
Session Publication Only
Topics Anticancer agents
Non-small-cell lung cancer
Biological therapy
Presenter jumpei Takeshita
Authors J. Takeshita1, N. Katakami1, A. Nishiyama2, H. Yoshioka3, M. Iwasaku3, A. Hata1, F. Imamura4, K. Nishino4, S. Yokota5, S. Morita6
  • 1Integrated Oncology, Institute of Biomedical Reseach and Innovation, 6500047 - Kobe/JP
  • 2Respiratory Medicine, Kurashiki Central Hospital, Kurashiki/JP
  • 3Respiratory Medicine Dept., Kurashiki Central Hospital, 710-8602 - Kurashiki/JP
  • 4Department Of Thoracic Oncology, Osaka Medical Center for Cancer and Cardiovascular Diseases, Osaka/JP
  • 5Respiratory Medicine, Toneyama National Hospital, 560-8552 - Toyonaka/JP
  • 6Biostatistics And Epidemiology, Yokohama City University Medical Center, Yokohama/JP



EGFR mutation is considered to be a strong predictive marker for EGFR tyrosine kinase inhibitor (TKI). Several phase III trials have shown that EGFR-TKI is superior to chemotherapy for NSCLC patients with EGFR mutation. On the other hand, there is no prospective trial which compares EGFR-TKI to chemotherapy for NSCLC patients without EGFR mutation. It remains unclear whether there is a difference between EGFR-TKI and chemotherapy for these types of patients.


We retrospectively investigated efficacy and safety of E, P and D in non-squamous NSCLC patients without EGFR mutation who had a history of one or two prior chemotherapy treatments, by using individual patient data. To examine the comparability in patient's clinical backgrounds between treatment groups, propensity score analysis was conducted. EGFR mutation analysis was performed with highly sensitive PCR methods.


Between December 2007 and September 2011, data of 370 patients who were treated either with E, P or D was collected in five institutions. There were distinct differences between the clinical characteristics of patients who received E and those of patients who received P or D.Propensity score analysis showed that clinical characteristics in a subgroup of patients with PS 0-1 and 2nd line chemotherapy were almost the same in each group.We analyzed this subgroup of patients (167pts, E 19 pts, P 53 pts, D 38 pts).The PFS was 2.80 months, 2.53 months, 1.90 months, respectively (p = 0.7745). The OS was 16.13 months, 7.40 months, 9.97 months, respectively (p = 0.4258). Grade 3 or worse hematologic toxicities were reported more frequently in P and D group. In all patients, three treatment-related deaths occurred in the E group (two deaths as a result of interstitial lung disease, one as a result of diarrhea).


This is the first report of retrospective analysis which compared E, P and D in pre-treated NSCLC without EGFR mutation. There was no efficacy difference between E, P and D in patients with PS 0-1 and 2nd line chemotherapy.Future prospective randomized trials need to be conducted in pre-treated NSCLC patients without EGFR mutation.


N. Katakami: I receive honoraria from chuugai Pharma Japan, Lilly Japan, and sanofi-aventis Japan. I receive research funding from chuugai Pharma Japan, Lilly Japan, and sanofi-aventis Japan.

H. Yoshioka: I receive honoraria from Chuugai Japan, Lilly japan,and Sanofi-aventis Japan.

All other authors have declared no conflicts of interest.