The ratio of T790M to EGFR-activating mutation predicts response of osimertinib in 1st or 2nd generation EGFR-TKI-refractory NSCLC

The most frequent mechanism of resistance after 1st/2nd-generation (G) epidermal growth factor receptor (EGFR)-tyrosine kinase inhibitors (TKIs) is secondary point mutation Thr790Met (T790M) in EGFR. Afatinib followed by osimertinib (Afa group) may provide better outcomes for T790M-positive non-small cell lung cancer (NSCLC) than 1st-G EGFR-TKI followed by osimertinib (1st-G group). We studied 111 consecutive NSCLC patients with T790M mutation treated with osimertinib after progression following 1st/2nd-G EGFR-TKI between March 28, 2016 and March 31, 2018. We analyzed the ratio of T790M to EGFR-activating mutation (T790M ratio) in post EGFR-TKI resistance re-biopsy tissue using droplet digital polymerase chain reaction. And investigated whether afatinib purified the T790M mutation more than 1st-G EGFR-TKI. Among 60 patients with preserved re-biopsy tissue, we analyzed 38 having adequate DNA content. The response rate in Afa group was 81.8% (n = 11) and 1st-G group was 85.2% (n = 27). The mean T790M ratio in total population was 0.3643. The ratio in those with response to osimertinib was significantly higher than in the non-responders (0.395, 0.202; p = 0.0231) and was similar in Afa and 1st-G group (0.371, 0.362; p = 0.9693). T790M ratio significantly correlated with osimertinib response and was similar between the 1st/2nd-G EGFR-TKIs in 1st/2nd-G EGFR-TKI-refractory tumors.

www.nature.com/scientificreports/ Moreover, osimertinib is associated with a longer PFS and overall survival (OS) than 1st-G EGFR-TKIs against advanced NSCLC harboring EGFR mutation (exon-19 deletion and L858R) as a first-line treatment 12 . However, in the Asian subset (especially in the Japanese subset) analysis of OS in the FLAURA study, osimertinib was not superior to 1st-G EGFR-TKIs 13 . There may be no molecular targets for therapy due to the heterogeneity of resistance mechanisms, which are not well understood 14,15 . As a result, during clinical care for most patients following cancer progression after osimertinib treatment, chemotherapy is the only remaining option for secondline treatment.
In contrast, afatinib demonstrated superior RR, PFS, and a trend of longer OS compared to gefitinib in the LUX-Lung 7 16 . A study showed that afatinib treatment followed by osimertinib indicated an extremely long treatment time to failure (TTF) and OS, even after removal of selection bias 17 . Furthermore, we reported that treatment with afatinib followed by osimertinib (Afa group) may provide better outcomes for T790M-positive NSCLC than that with 1st-G EGFR-TKIs (1st-G group) 18 . These results may be explained by the possibility that afatinib exhibits increased efficacy by targeting co-occurring EGFR mutations that enrich T790M cells.
Therefore, to investigate whether afatinib purifies T790M mutation more effectively than the 1st-G EGFR-TKIs, we analyzed the difference in the ratio of T790M mutation to EGFR-activating mutation ratio (T790M ratio) between the Afa and 1st-G group by using droplet digital polymerase chain reaction (ddPCR).

Materials and methods
We conducted a multicenter-retrospective study across three medical centers (Osaka International Cancer Institute, Osaka Habikino Medical Center, and National Hospital Organization Kinki-Chuo Chest Medical Center) in Japan. The study design and methodology were approved by the Institutional Review Board of each participating institution (Institutional Review Board of Osaka International Cancer Institute, Institutional Review Board of Osaka Habikino Medical Center, and Institutional Review Board of National Hospital Organization Kinki-Chuo Chest Medical Center), and it was conducted in accordance with the Declaration of Helsinki and the World Health Organization's Guidelines for Good Clinical Practice. We obtained the informed consent from participants and used an opt-out method so that patients and their families could refuse to participate in the study.
Patient selection and ddPCR measurement. Between March 28, 2016 (the date osimertinib was approved in Japan) and March 31, 2018, study participants were consecutively enrolled according to the following criteria: patients with T790M mutation who were treated with osimertinib after AR to EGFR-TKIs at any time for advanced NSCLC, had good Eastern Cooperative Oncology Group (ECOG) performance status (PS): 0-2, and had major EGFR mutation (Exon-19: deletion19 or Exon-21: L858R) before initial EGFR-TKI treatment to reduce bias towards patient conditions. Furthermore, we excluded cases where we detected T790M mutation in the pre-treatment tissue by EGFR test in general practice.
We used RIKEN GENESIS CO., LTD. (Tokyo, Japan) to measure exon-19 deletion, L858R, and T790M by droplet digital polymerase chain reaction (ddPCR). The ddPCR analysis was performed using QX200 AutoDG Droplet Digital PCR System (BioRad). In this study, we excluded samples that had less than 1000 total copies/20 uL well or less than 10 copies with EGFR-activating mutation/20 uL well. T790M ratio was calculated by the following method. First, the fractional abundance (mutant allele/mutant allele + wild type allele) of activating mutation and T790M were measured with droplet count by ddPCR. Next, T790M ratio defined the ratio of T790M to activating mutation.

Statistical analysis.
We evaluated the systemic response to osimeritinib using the Response Evaluation Criteria in Solid Tumors (RECIST) ver1.1 19 . We used Fisher's exact tests for categorical comparisons of data and compared differences in continuous data using the Wilcoxon test. Kaplan-Meier curves were used to evaluate PFS, which were compared using the log-rank test. Median values and 95.0% confidence intervals (CIs) were also reported. All statistical analyses were conducted using R software, version 2.8.1 (http://R-proje ct. org) (The R Foundation for Statistical Computing, Vienna, Austria). p < 0.05 was considered a statistically significant difference.

Results
Of the 111 enrolled patients, the re-biopsy tissue obtained from each patient before osimertinib treatment was preserved for 60 patients. Among them, 38 samples yielded enough DNA for ddPCR analysis, and the T790M ratio was assessed (Fig. 1). The patient characteristics are shown in Table 1. The median age was 68 years. Of the 38 patients, 21.1% were men, 23.7% had a history of smoking, 44.7% had L858R mutation, and 55.3% had exon 19 deletion. Eleven (28.9%) were in the Afa group and 27 (71.1%) were in the 1st-G group. The RR in all patients was 84.2%, 81.8% in the Afa group, and 85.2% in the 1st-G group ( Table 2).
Furthermore, we divided two groups based on a threshold T790M ratio of 0.35 from the mean T790M ratio in total population. The response rate in patients with T790M ratio ≥ 0. 35  2), however this was not statistically significant (Fig. 3c). Pearson correlation analysis was performed to examine the relationship between T790M ratio and PFS. The analysis showed a correlation coefficient of r = 0.06123 (p = 0.715) (Fig. 4).

Discussion
This is the first analysis to reveal a potential implication that afatinib purifies the T790M mutation more than the 1st-G EGFR-TKI. This study suggested that the T790M ratio in 1st/2nd-G EGFR-TKI-refractory tumors significantly correlated with response to osimertinib. The T790M ratio in CR or PR subgroup of the osimertinib group (mean: 0.395) was significantly higher than in the SD or PD subgroup (mean: 0.202); p = 0.0231. However, Figure 1. CONSORT diagram. 1st-and 2nd-G EGFR-TKIs: first and second generation epidermal growth factor receptor-tyrosine kinase inhibitors, DNA deoxyribonucleic acid.  Table 2. The efficacy of osimertinib following between afatinib and 1st-G group. 1st-G group: patients treated with 1st-generation EGFR-TKIs followed by osimertinib; afatinib group: the patients treated with afatinib followed by osimertinib. RR response rate, DCR disease control rate, CR complete response, PR partial response, SD stable diseases, PD progression diseases. www.nature.com/scientificreports/ we did not observe a difference in the T790M ratio between the 1st-and 2nd-G EGFR-TKIs. The T790M ratio in the Afa group (mean: 0.3711) was similar to the 1st-G group (mean: 03616); p = 0.9693. Osimertinib was developed as a drug against T790M that works by covalently binding to T790M-muted EGFR 10 . Therefore, although osimertinib is also effective against sensitive EGFR mutations, the most effective targets are T790M-positive EGFR mutations. In contrast, afatinib irreversibly binds to EGFR, ERBB2, and ERBB4   www.nature.com/scientificreports/ and blocks transphosphorylation of ERBB3 to inhibit all ERBB family signaling 20 . Two hypotheses were proposed: the first is that the broader inhibition of the EGFR family may improve the efficacy of EGFR-TKIs for NSCLC with EGFR mutation, and the second is that the broader inhibition of the EGFR family may attenuate T790M in cancer cells during relapse after EGFR-TKI treatment. Osimertinib could also be more effective against the AR of T790M after EGFR-TKI. The results of LUX-Lung7 and ARCHER1050 supported this theory in the first-line treatment for NSCLC with EGFR mutation 16,21 . Furthermore, afatinib provided the efficacy for uncommon EGFR mutations in preclinical and clinical studies 22,23 . In the analysis of GIO-TAG trials, the duration of response to osimertinib in patients receiving afatinib and subsequent osimertinib was highly encouraging 17,24,25 . In our study, the concentration of T790M correlated with the response to osimertinib, and our results suggested that the degree of attenuation of T790M is important in eliciting the effects of osimertinib. Some prior studies have also shown similar results: the ratio of T790M to EGFR-activating mutation in cell samples or plasma samples may predict the response to osimertinib or rociletinib [26][27][28] .

N RR (%) DCR (%) CR (%) PR (%) SD (%) PD (%)
However, we did not observe differences in the clonality of T790M after AR between afatinib and 1st-G EGFR-TKI. One reason was that the response of osimertinib after AR was the same in Afa group and 1st-G group when we collected the available tissue, although our previous study revealed that the objective RR and disease control rate were significantly higher in the Afa group than in the 1st-G group 18 . We intended to collect all the samples in our previous study; however, this was not possible. Another reason is that later lines of therapy may have advanced clonal evolution and may not reflect the effect of osimertinib after 1st-G EGFR-TKI and afatinib treatment.
Next, Pearson correlation analysis showed no significant relationship between T790M ratio and PFS of osimertinib after AR of EGFR-TKIs (r = 0.06123, p = 0.715). We believed that the PFS of osimertinib after AR in NSCLC with T790M mutation was influenced by co-mutation.
Based on clinal evolution, which was reported for the first time as early as 1976 29 , lung adenocarcinoma seems to follow a multistep progression from atypical adenomatous hyperplasia to adenocarcinoma in situ, and finally invasive adenocarcinoma 30 . EGFR driver alterations are acquired in the early step of cancer progression and can be identified in most neoplastic cancer cells 31 . Furthermore, by layering exacerbations of disease after treatment, especially EGFR-TKI use, cell progression will show different genomic patterns of selection through different lines of therapy 32 . In this setting, EGFR-sensitive mutant tumor cells may coexist with sub-clonal tumor cells harboring other gene mutations, and co-mutations such as TP53 affect the efficacy of osimertinib 33,34 . Furthermore, histologic transformation and other off-target molecular alterations are frequent early emerging resistance mechanisms to osimertinib and are associated with poor clinical outcomes 35 .
We must consider some limitations of the present study. First, this was a retrospective study, and the participants were selected based on obtainment of adequate samples. Second, the line of treatment for EGFR-TKIs was not determined. Since the lines of EGFR-TKIs varied from case to case, the effectiveness of osimertinib may also vary accordingly. Although osimertinib response was not different between the two groups, we focused on the attenuation of tumor cells by EGFR-TKIs immediately before the administration of osimertinib to clarify the correlation between T790M ratio and osimertinib response. Third, in this study, we examined only exon-19 deletion, L858R and T790M by ddPCR and did not perform next generation sequencing. Therefore, we cannot know about the presence of other EGFR mutations, amplifications, alternative bypass pathways or co-mutations that may provide a mechanism of resistance to osimerutinib.

Conclusion
In our analysis, the T790M ratio in 1st/2nd-G EGFR-TKI-refractory tumors significantly correlated with osimertinib response. Furthermore, the T790M ratio between the 1st-and 2nd-G EGFR-TKIs was almost similar in patients with EGFR-TKI-refractory NSCLC who showed response to osimertinib. Further analysis in a larger number of prospective studies is warranted to confirm our results.  www.nature.com/scientificreports/ Author contributions M.T. designed the study. M.T. and A.T. wrote the initial draft of the manuscript. M.T. and H.S. contributed to analysis the results of ddPCR from RIKEN GENESIS CO., LTD. All other authors have contributed in data collection and in review of the manuscript. All authors approved the final version of the manuscript and agree to be accountable for all aspects of the work.

Funding
The study was funded by Boehringer Ingelheim.