Biomarker analysis of the BR.21 study showed survival among patients with high EGFR expression was longer in the erlotinib arm versus the placebo arm, whereas a limited advantage U0126 order of erlotinib treatment was seen in patients with EGFR IHC-negative tumors [22]. These results were the basis for the inclusion of PFS in patients with EGFR IHC-positive disease as a co-primary endpoint in SATURN. However,

Pérez-Soler et al. reported no correlation between survival and EGFR expression (p = 0.90) in NSCLC patients treated with erlotinib in the second-/third-line setting [23]. Additionally, Murray et al. demonstrated no correlation between EGFR protein expression and disease control rate in erlotinib-treated patients

when staining for total EGFR or phosphorylated EGFR [24]. For the SATURN study, using a positive threshold of ≥10% membrane staining failed to identify any correlation between EGFR expression and patient outcomes. Using a different IHC analysis method (H-score with application of the magnification rule) in the present analysis did not change the correlation MLN0128 ic50 between EGFR expression levels and PFS or OS in SATURN. The different results between these studies suggest that the value of EGFR IHC to predict clinical outcomes may vary between different EGFR inhibitors and across different patient populations and treatment settings. The BioLOGUE advisors recently concluded that EGFR IHC status was weakly prognostic Loperamide but not predictive of outcomes with erlotinib, and noted that inconsistency across trials meant EGFR IHC was not a suitable biomarker [25]. Assessment of total receptor expression may not be the most accurate indicator of response to EGFR TKIs, as EGFR activating mutations are considered to be more important than EGFR protein expression levels. It has been suggested that a combination of IHC and fluorescence in situ hybridization may provide more suitable analysis [24], but this method has not yet been investigated in clinical trials. One reason that previous EGFR IHC studies might not have shown correlations with treatment response may be that the majority of diagnostic

antibodies target the external domain of the receptor, while it is mutations in the internal tyrosine-kinase domain that result in the increased response to erlotinib. The use of a diagnostic antibody that targets the internal EGFR domain (such as 5B7) [26] might result in better prediction of response with erlotinib using IHC. The results of this re-analysis suggest that EGFR IHC does not accurately predict erlotinib benefit for the overall population or the EGFR WT population in the first-line maintenance setting for advanced NSCLC. Dr Mazieres has received honoraria from Roche, Pfizer, Eli Lilly and Boehringer Ingelheim. Dr Bara and Dr Klingelschmitt are employees of Roche. Dr Klughammer is an employee of Roche and owns stocks in F.