Epidermal Growth Factor Receptor Mutations in Non–Small Cell Lung Cancer: Predicting Clinical Response to Kinase Inhibitors

In this issue of Clinical Cancer Research, Taron et al., describe their experience with 68 gefitinib-treated, chemorefractory patients with advanced non–small cell lung cancer (NSCLC) from Asia, Europe, and North America (1). By sequencing exons 18, 19, and 21 of the epidermal growth factor receptor (EGFR), they found mutations in 25% of specimens analyzed, consisting of the characteristic in-frame deletions in exon 19, and missense point mutations in exons 18 and 21. A significant difference in response to gefitinib treatment was evident depending on mutation status, with EGFR mutation–positive cases showing a much greater response (94%) than mutation-negative cases (13%, P < 0.001). In addition, never-smokers, heavily pretreated patients, Asians, and younger patients experienced a higher response rate to gefitinib. Median survival was 10 months in the EGFR mutation–negative cases and was not reached at a median follow-up of 11 months in patients with mutation-positive tumors.

Somatic activating mutations in the EGFR gene correlating with profound clinical response to EGFR tyrosine kinase inhibitors (TKI) were first identified in NSCLC ∼1 year ago (2, 3). Compared with wild-type EGFR, mutant receptors exhibit increased activation following ligand binding, as well as enhanced inhibition by TKIs. These initial observations have recently been extended in a number of molecular studies: we have learned that the association of EGFR mutations with TKI responsiveness applies to erlotinib as well as gefitinib (4); that EGFR mutations lead to alteration of downstream signaling pathways making cells more susceptible to TKI-mediated apoptosis (5); and that secondary EGFR point mutations contribute to resistance to TKI therapy (6, 7). Clinical studies have shown that erlotinib prolongs survival as second- and third-line treatment for NSCLC, compared with placebo (8), and a number of studies are now testing the molecular determinants that drive this clinical response.

The results from Taron et al. are consistent with a growing database of EGFR mutation frequency and clinical outcomes after EGFR TKI treatment (Table 1; refs. 2–4, 9–13). Retrospective analyses have shown that NSCLC patients whose tumor harbors an EGFR mutation have a response rate to EGFR TKI treatment of ∼65% to 100%. The 94% response rate observed by Taron et al. is higher than most reports of sequentially tested patients, but all studies have involved small numbers of patients. Three analyses have shown that overall survival is significantly prolonged in these patients, compared with mutation-negative cases, with an observed median survival of ∼2 years in patients with mutation-positive tumors. This represents a substantial improvement in survival compared with traditional chemotherapy in unselected, previously treated patients with advanced NSCLC, where median survival has been ∼7 to 8 months in randomized clinical trials (14, 15). In addition, the clinical characteristics previously identified as predictive for response to EGFR TKIs (16–19), including gender (female), race (Asian), adenocarcinoma tumor histology, and a history of never smoking, have consistently been correlated with somatic EGFR mutations, whose increased prevalence in these clinical subsets likely explains their correlation with drug response.

As a result of these collective data, there is growing interest in moving EGFR TKI therapy earlier in the treatment algorithm for EGFR mutation–positive NSCLC, including first-line treatment of advanced disease and adjuvant therapy for early stage resected disease. Neither gefitinib nor erlotinib have benefited unselected patients with advanced NSCLC when used as first-line treatment in combination with chemotherapy (20–23), but either TKI alone may be effective in a patient population selected for EGFR mutations or clinical characteristics of responsive disease. A phase II study of gefitinib as first-line treatment in never-smoking patients with advanced adenocarcinoma was recently completed in South Korea (24), reporting a response rate of 69%, and an estimated 1-year overall survival of 73%, among 36 patients enriched for clinical characteristics of responsive NSCLC. Median survival has not yet been reached at a median follow-up of 11 months. A multicenter phase II study using gefitinib as first-line therapy in patients with advanced NSCLC known to harbor somatic EGFR mutations is now under way in the U.S. to estimate the response rate in a more specific population. Additional ongoing studies in the U.S. include first-line EGFR TKI treatment in a population defined by clinical criteria of TKI responsiveness, and phase II and III clinical trials of EGFR TKIs in combination with chemotherapy as adjuvant and neoadjuvant therapy for early stage disease. In all these studies, the magnitude of benefit from EGFR TKI therapy will need to be assessed in specific subpopulations of patients defined by molecular and clinical characteristics, and adequate tissue for confirmation of molecular findings will be prospectively collected. Detecting mutations in limited amounts of paraffin-embedded samples has proved challenging and may have biased retrospective studies. Until these trials are completed, it may be reasonable to advocate for early use of EGFR TKI therapy in patients with advanced NSCLC and proven EGFR mutations if mutation testing is available, or in patients with multiple clinical characteristics associated with mutations if mutation testing is not feasible. Predicting the role of adjuvant EGFR TKI therapy, with or without chemotherapy, following surgical resection of early stage disease in EGFR mutation–positive cases is more challenging and this will require further clinical investigation.

The search for additional molecular markers of prognosis and response to EGFR TKI therapy has yielded potentially interesting, but often conflicting results. In a subgroup of their study population with sufficient tumor material, Taron et al. examined the association of EGFR mutations with other molecular markers, such as the CA dinucleotide repeat in intron 1 of EGFR, EGFR mRNA levels and gene amplification, and expression of the regulator of receptor turnover, caveolin. Although their sample size was not sufficient to yield any statistically significant associations, they noted that tumors with somatic EGFR mutations had high numbers (>19) of CA dinucleotide repeats in intron 1. This association is of potential interest, because an increased number of CA repeats in EGFR has been observed in Asians, who are more likely to have EGFR-mutant tumors than Caucasians (25). However, in other studies, increased number of CA repeats has been correlated with decreased EGFR expression and decreased sensitivity to erlotinib in head and neck cancer cell lines (26, 27).

Taron et al., also noted a trend toward an association between EGFR mutations and EGFR gene amplification. The role of EGFR gene amplification in NSCLC is also of potential interest. In one clinical study, high gene copy numbers in NSCLC was correlated with worse survival (28), whereas another study reported EGFR gene amplification to be positively correlated with response to EGFR TKI therapy and survival (29). In the study by Taron et al. response to EGFR TKI in the subgroup of cases with sufficient tumor material for multiple analyses was much greater in those with EGFR mutations (89%) compared with those with gene amplification (45%, P = 0.02). Moreover, EGFR mutations and gene amplification were correlated with each other, making it difficult to determine the extent to which EGFR amplification in the absence of mutations is predictive of response. Nonetheless, it is possible that both EGFR mutations and gene amplification may represent markers of drug response, and even that the quality and duration of drug response might eventually be correlated with different types of mutations within EGFR or different levels of gene amplification. As shown by the work of Taron and coworkers, clinical studies of EGFR aberrations in NSCLC are rapidly accumulating and clinical practice patterns are likely to evolve along with these molecular discoveries. During the past year, progress in molecular therapeutics relating to lung cancer has been breathtaking, a most welcome development for the treatment of this usually devastating disease.