In this study, we first analyzed the frequency and character of EGFR mutations in patients with unresectable locally advanced adenocarcinoma. In our study, rare mutations accounted for 9.6% of all EGFR mutations, and G719X (2.2%, 11/511) and 20INS (1.7%, 9/511) were the most common single rare mutation sites. We also observed 3.3% (17/511) of EGFR compound mutations, including G719X, L858R and S768I compound mutations. The above mutations were comparable to the prevalence reported in other studies9,22–25. The two groups in our study were similar in terms of sex, age, smoking and stage.
Compared to the standard chemotherapy, the NEJ002 study showed that the PFS and OS tended to be shorter among patients treated with EGFR-TKIs in the advanced lung cancer with uncommon EGFR mutation11. Similarly, our results showed that the PFS of standard CRT as first-line treatment was significantly longer than in patients who received upfront TKI. Furthermore, there was no apparent difference in survival between the rare and common mutations (median PFS 11.9 vs. 12.6 months, P = 0.393; median OS 43.6 vs. 53.9 months, P = 0.479) in the CRT group. This result was similar with another retrospective study of limited size on CRT in patients with uncommon mutations16.Th These results suggest that CRT remains the standard of care for this subgroup.
The previous real-world data and clinical trials suggest that patients with the rare seem inferior response to treatment with EGFR TKIs compared to the two common mutations, especially for first-generation EGFR TKIs. 10–12,22. Our data were like previous research reported, the PFS and OS in the population received upfront EGFR TKIs shows significantly shorter to common mutations (median PFS 5.0 vs. 16.9 months, P < 0.001; median OS 30.9 vs. 49.3 months, P = 0.143). Nevertheless, afatinib showed a favorable response in patients harboring certain types of rare mutations, particularly G719X, L861Q and S768I mutations, from the data of the LUX-Lung 2, 3, 6 studies.10. Similarly, a phase II clinical trial of osimertinib showed favorable outcomes and tolerable toxicity in patients harboring G719X (n = 19; 53%), L861Q (n = 9; 25%), S768I (n = 8; 22%) mutations12. Importantly, the relevant models showed that the most common EGFR exon 20 insertion mutation failed to respond to both first-generation (gefitinib, erlotinib) and second-generation (neratinib, afatinib) EGFR TKIs.26. Few patients with EGFR exon 20 insertions showed clinical response to EGFR TKIs from some clinical evidence. As a result, these patients were excluded from the trial of the third-generation EGFR TKI (osimertinib)12,27. According to our findings, the upfront EGFR TKIs should not be recommended as first choice for locally advanced NSCLC harboring uncommon mutations, however, the optimal treatment strategy for these patients remained a challenge, the combination of RT and next-generation EGFR TKIs may be a promising choice according to the genotype.
Few studies have looked specifically at the prevalence and survival of compound mutations. The frequency of EGFR compound mutations in this cohort was 3.3% (17/511), which is slightly lower than previous data. In this subgroup analysis, we found that patients in the EGFR compound mutation group had similar PFS (median 11.9 vs. 10.0 months, P = 0.304). The previous study also showed that the PFS of the EGFR compound mutation group was significantly higher than the individual uncommon sub-cohorts (median, 11.9 vs 6.5 months, P = 0.010) among patients receiving EGFR TKIs alone28. Another study also showed different results for PFS (median 9.3 vs. 7.0 months, P = 0.495), but there was a significant difference for OS (median 31.4 vs. 12.4 months, P = 0.001).29. Due to the limited sample size of the study, these results need to be confirmed by further studies.
Several limitations cannot be ignored. The most important limitation is the small size of the rare mutation cohort, which requires further data collection. In addition, the detection of the rare mutation is limited by the sequencing methods and relevant samples. The polymerase chain reaction (PCR)-based or direct sequencing methods used may have led to an underestimation of patients with EGFR rare mutations compared to next-generation sequencing (NGS). Extraction of plasma samples for genetic testing is feasible in clinical practice, but accuracy and sensitivity are lower than tissue biopsy. More importantly, clinical decision making for LA-NSCLC with EGFR mutations differs between institutions and physicians. Finally, the pulmonary toxicity of concurrent RT and EGFR TKIs is of particular concern. We retrospectively analyzed 45 patients treated with concurrent radiotherapy and EGFR TKI. 37.7% of these patients developed symptomatic radiation pneumonitis (grade 2+) and 6.7% had severe (grade 3) radiation pneumonitis. However, no fatal adverse events and limited late lung toxicity were observed.30
In conclusion, this is the first study that systemically summarized the prevalence and outcomes of EGFR uncommon mutations in patients with unresectable LA-NSCLC. the uncommon mutations associated with worse survival for patients receiving EGFR TKIs, the CRT therapy and RT combined next-generation EGFR TKIs may be a promising choice and the EGFR TKIs alone should not be recommended as a primary choice for patients with uncommon EGFR mutations. Future studies should validate these results.