The impact of rare EGFR mutations on the treatment response of patients with non-small cell lung cancer

Abstract

Mutational activation of the epidermal growth factor receptor (EGFR) gene is implicated in lung cancer; clinical and cancer genome sequencing studies have identified hundreds of mutations in the protein kinase domain. EGFR mutation testing usually focuses on common mutations like the exon 19 deletion and exon 21 point mutation (L858R). However, molecular screening methods have started to extend beyond identification of classic EGFR mutations to prevent exclusion of patients with rare or complex mutations who may benefit from anti-EGFR therapy. Rare EGFR-mutated non-small-cell lung cancers are heterogeneous: exon 20 insertions lack sensitivity to tyrosine kinase inhibitors while exon 18 or complex mutations are more sensitive and require individual assessment. Until testing for uncommon EGFR mutations evolves and studies with large number of patients are performed, knowledge of this field will remain limited.

Keywords:

• epidermal growth factor receptor mutations
• exon 18 mutations
• exon 20 mutations
• non-small-cell lung cancer
• tyrosine-kinase inhibitors

In the era of personalized therapeutics, the discovery of mutations in the EGFR in 9% of lung adenocarcinomas in European populations and the associated response to EGFR-targeting tyrosine kinase inhibitors (TKIs) have provided a successful avenue of attack in high-stage adenocarcinomas [1,2]. From the current body of available literature, three defined regions (exons 18, 19 and 21) in the EGFR gene are commonly mutated and predict sensitivity to EGFR TKI. Exon 19 deletions of 15–18 pb represent more than 50% of EGFR mutations, and the exon 21 point mutation at the residue L858R represents more than 30% [3]. However, there are a number of relatively rare EGFR mutations whose associations to EGFR TKIs are not well clarified. Herein, we summarize the current knowledge of multiple rare EGFR mutations and their sensitivity to EGFR TKIs.

From the initial studies in 2004, patients sensitive to gefitinib showed mutations in the EGFR ATP-dependent TK domain (‘catalytic kinase domain’), with most being point mutations (G719C, L858R and L861Q) or in-frame deletions (delE746-A750, delL747-T751inS and delL747-P753inS) [3]. Functional analysis of the L858R and delL747-P753insS in COS-7 cells demonstrated that cells harboring the mutations had higher phosphorylation activity of EGFR (Y1068) and were more susceptible to gefitinib compared with wild-type EGFR-expressing COS-7 cells [3]. In kinetic kinase experiments, proteins with L858R or exon 19 deletions exhibit significantly higher Michaelis constants for ATP (Km [ATP]; substrate concentration at half the maximum reaction rate), and have increased affinity to gefitinib or erlotinib compared with the wild-type receptor. The gefitinib- and erlotinib-resistant EGFR T790M mutation is thought to increase the affinity of EGFR to ATP, thus decreasing the drugs’ effectiveness [4].

Not all activating mutations confer the same degree of sensitivity to TKIs. Non-small-cell lung cancer (NSCLC) cells expressing the L858R mutation are significantly more sensitive to gefitinib than are those expressing the L861Q. Based on studies with small number of patients with the EGFR L861Q mutation, response has not been consistently reported with documentation of progressive disease and shorter response duration after EGFR TKI therapy. In a recent study, L861Q mutations were detected in 15 patients treated with gefitinib or erlotinib, with a response rate of 60%, median progression-free survival (PFS) of 6.0 months and median overall survival (OS) of 15.2 months [5]. In a post-hoc analysis of patients with the L861Q mutation from the NEJ002 trial, OS with gefitinib was significantly shorter compared to patients with the common mutations, but there were no differences in the chemotherapy group [6]. In contrast, L861Q shows better sensitivity to the irreversible EGFR TKI afatinib, and response and PFS of patients enrolled in the LUX-Lung 2, 3 and 6 trials have been seen to be notably improved [7].

It has been suggested that in-frame deletions of EGFR predict higher sensitivity to TKI compared with point mutations in exons 18 and 21. Indeed, lung adenocarcinomas harboring the EGFR exon 19 deletion (del746-750) may be more susceptible to gefitinib as compared with tumors with the exon 21 point mutation (L858R). Response rates to EGFR TKIs are higher in exon 19 mutated NSCLCs (70–100%) than exon 21 mutated (20–67%) [8]. These differential response rates translated into longer survival, where patients with an exon 19 deletion had median OS ranging from 26 to 34 months and patients with exon 21 (L858R) had median OS from 8 to 17 months [8]. A recent pooled analysis of the LUX-Lung 3 and 6 trials demonstrated a differential activity for afatinib according to the type of EGFR mutation (exon 19 deletions: HR, 0.59; p < 0.001; L858R mutations: HR, 1.25; p = 0.16) [9].

In addition to deletions, insertions in exon 19 have been described in a very small number of patients. Exon 19 insertions were recently found in 12 NSCLC samples from mainly female, never smoker patients with adenocarcinoma histology [10]. Sequencing of available samples showed an 18 base pair insertion that resulted in the addition of a XPVAIK sequence, where X can be I, V or T. Four patients received palliative EGFR-TKIs and three had a partial response by RECIST criteria. Accompanying in vitro data revealed the oncogenicity of exon 19 insertions in transfected cell lines as well as their sensitivity to TKI [10]. D761Y a ‘T790M-like secondary mutation’ on exon 19 has been found to co-occur with a pre-existing L858R mutation and confers decreased sensitivity of the L858R mutant to EGFR TKIs. However, the decrease in EGFR TKI sensitivity was markedly less than that of T790M and the real effect of D761Y on the EGFR kinase domain remains unknown [11]. For instance, Ba/F3 cells containing L858R plus D761Y showed intermediate sensitivity to irreversible EGFR inhibitors compared with L858R alone or L858R plus T790M [12].

Additional mutations in the EGFR gene have been identified in exons 20 and 18, but their role in predicting sensitivity or resistance to EGFR TKI remains to be fully elucidated in clinical studies [13]. Mutations which may predict different EGFR TKI response include exon 20 point mutations, a variety of exon 20 insertions and combinations of different exon 20 mutations with the more frequent exons 21 and 19 mutations [14]. The single S768I mutation is associated with poorer EGFR TKI response [15]. There have been reports of response of double mutations of S768I with L858R or exon 19 deletion but with very few patient cases [15]. The most common EGFR exon 20 insertions reported in NSCLC samples are post-C helix insertions of one to four amino acids, which, in aggregate, account for 80–90% of all exon 20 insertions. Yasuda et al. identified an EGFR exon 20 insertion mutation (A763_Y764insFQEA) that was inhibited by gefitinib and erlotinib at submicromolar concentrations. This sequence alteration accounts for 10–20% of all reported EGFR exon 20 insertions [16]. Exon 20 insertions like D770_N771_(ins NPG), D770_(ins SVQ), D770_(ins G) N771T and H773_V774 (ins H) were initially reported to cause resistance to EGFR TKIs, something which has been supported by subsequent studies, with few patients involved. The true frequency of exon 20 insertions is unknown, with reports describing from 1 to 10% contribution to the total number of EGFR mutations identified. However, 4% seems to be the most widely reported in compiled cohorts of published EGFR mutations [4]. In a recent study, exon 20 insertions were found in 33 (2.2%) of 1500 lung adenocarcinoma cases screened; using computer-assisted 3D modeling, the authors postulated that exon 20 insertions alter drug affinity [17]. Only five of these patients received TKI therapy, and the two who received single-agent TKI therapy showed no response [17].

The activity of irreversible EGFR TKIs has been examined in exon 20 mutant proteins. In preclinical models, the IC₅₀ of afatinib and dacomitinib for A767_V769_(dupASV), D770_N771_(ins NPG), delN771_(insGY) and H773_V774 (ins H) were in similar dose ranges as T790M, 100 times less sensitive than the classic L858R and exon 19 deletions [18]. In a Phase I trial of dacomitinib, six patients with EGFR exon 20 insertions were included and only one (with delA770insGY) had a response [19]. The efficacy of afatinib in patients with uncommon EGFR mutations enrolled in the LUX-Lung 2, 3 and 6 was recently presented. In the group of patients harboring exon 20 insertions, response, PFS and OS were extremely poor. However, afatinib was active in lung tumors harboring S768I in combination with G719X mutation on exon 18 [7].

The exon 18 point mutation accounts for 4% of EGFR mutations and includes a Gly719 change to Ser, Ala or Cys (G719S/A/C). Although the G719S is associated with a 10-fold EGFR activation compared to wild-type, NSCLC cells expressing the L858R mutant are significantly more sensitive to gefitinib than those expressing the G719S mutant [20]. G719 mutants do not appear to be resistant to EGFR TKIs and mutations to Ala, Ser or Cys promote EGFR kinase activation. Other exon 18 mutations reported in responders to EGFR-TKIs include: V689M, S720P/F, P699S, N700D, E709Q, G721A and L718P [21]. However, there is limited published experience regarding exon 18 mutations. In the study from Wu et al., of the 62 patients who had uncommon EGFR mutations of unknown clinical significance and received EGFR TKI treatment (gefitinib or erlotinib), G719 mutations were noted in 15 patients with response rate of 53.3%, median PFS of 8.1 months and median OS of 16.4 months [5]. In the same study, besides G719 and L861, five patients had the E709 mutation on exon 18 and four patients had the S768 exon 20 mutation. All were complex mutations and those not in combination with G719, L861, L858R or deletions in exon 19 had a lower response to EGFR TKIs (20%), median PFS 1.6 months and median OS 11.1 months, all lower than in patients with mutations on G719 or L861 [5]. In the Watanabe et al. post-hoc analysis, patients with the G719 mutation had significantly shorter OS with gefitinb compared to patients with the common mutations. OS was similar with chemotherapy for patients with common and uncommon mutations [6], but response rate and PFS was notably improved for the G719X mutants in the LUX-Lung 2, 3 and 6 trials with afatinib [7].

Finally, a novel EGFR kinase domain somatic mutation, E884K (Glu884Lys) in exon 22 in combination with L858R has been described as having an apparently differential response to erlotinib and gefitinib. Compared with L858R, the dual mutant L858R + E884K is less sensitive to erlotinib and more sensitive to gefitinib in the inhibition of tyrosine phosphorylation of EGFR. Indeed E884K in-cis with L858R decreases erlotinib inhibition of AKT and STAT3 but increases inhibition by gefitinib [22].

Because data are scarce and results controversial, the question of how to treat patients with uncommon EGFR mutations remains unanswered. Some investigators consider that in the absence of robust evidence, chemotherapy is a reasonable therapeutic approach. Our opinion is that uncommon EGFR mutations are heterogeneous and should be considered individually and not as a group. Targeted therapies may still be useful for these patients. Compiling databases with the largest possible amount of data on uncommon EGFR mutations, conducting in vitro experiments and using bioinformatics tools for testing sensitivity to EGFR TKIs are the most reasonable approaches that should be followed for lung cancer patients with uncommon EGFR mutations.

Financial & competing interests disclosure

Work in R Rosell’s laboratory was partially supported by grants from the La Caixa Foundation and Red Tematica de Investigacion Cooperativa en Cancer (RTICC; grant RD12/0036/0072). The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.