If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Exploring the resistance mechanisms of second-line osimertinib and their prognostic implications using next-generation sequencing in patients with non-small-cell lung cancer
1 Present address: Division of Hemato-Oncology, Department of Internal Medicine, Korea University College of Medicine, Korea University Guro Hospital, 148, Gurodong-ro, Guro-gu, Seoul 08308, Republic of Korea.
2 Kyoungmin Lee and Deokhoon Kim equally contributed this paper.
Kyoungmin Lee
Footnotes
1 Present address: Division of Hemato-Oncology, Department of Internal Medicine, Korea University College of Medicine, Korea University Guro Hospital, 148, Gurodong-ro, Guro-gu, Seoul 08308, Republic of Korea. 2 Kyoungmin Lee and Deokhoon Kim equally contributed this paper.
Affiliations
Department of Oncology, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Republic of Korea
2 Kyoungmin Lee and Deokhoon Kim equally contributed this paper.
Deokhoon Kim
Footnotes
2 Kyoungmin Lee and Deokhoon Kim equally contributed this paper.
Affiliations
Department of Pathology, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Republic of KoreaAsan Institute for Life Science, University of Ulsan College of Medicine, Seoul, Republic of Korea
Corresponding author: Department of Oncology, Asan Medical Center, University of Ulsan College of Medicine, 88, Olympic-ro 43-gil, Songpa-gu, Seoul, 05505, Republic of Korea. Fax: +82 2 3010 6961.
1 Present address: Division of Hemato-Oncology, Department of Internal Medicine, Korea University College of Medicine, Korea University Guro Hospital, 148, Gurodong-ro, Guro-gu, Seoul 08308, Republic of Korea. 2 Kyoungmin Lee and Deokhoon Kim equally contributed this paper.
Variety of resistance alterations develop with second-line osimertinib treatment.
•
Loss of T790M mutation is apparently associated with early resistance.
•
Acquiring MET amplification in addition to T790M loss further worsens the prognosis.
•
The patient's prognosis depends on which resistant subclone becomes the dominant.
•
Identifying dominant alteration would be helpful to determine subsequent treatment.
Abstract
Introduction
Although osimertinib overcomes the T790M mutation acquired after traditional epidermal growth factor receptor (EGFR) gene tyrosine kinase inhibitor (TKI) treatment, resistance to osimertinib eventually occurs. We explored resistance mechanisms of second-line osimertinib and their clinical implications by comparing next-generation sequencing (NGS) results before and after resistance acquisition.
Methods
We enrolled 34 patients with advanced EGFR-mutant adenocarcinoma whose biopsied tumour tissues were subjected to targeted NGS at the time of progression on osimertinib. For comparison, NGS was also performed on archived tumour tissues from each patient excised before osimertinib initiation.
Results
The tumours of three patients’ were observed to have transformed to small-cell carcinoma and those of two patients to squamous cell carcinoma. Among the remaining 29 patients, T790M mutations were maintained in seven patients (24.1%), including four patients (13.8%) acquiring C797S mutations and one with MET amplification. Among the 22 patients (75.9%) with T790M loss, a variety of novel mutations were identified, including KRAS mutations, PIK3CA mutations, and RET fusion, but MET amplifications (n = 4, 18.2%) were most frequently identified variations. Progression-free survival (PFS) on osimertinib was shorter among patients with T790M loss than among those who maintained T790M (5.36 versus 13.81 months, p = 0.009), and MET-amplified patients were found to have much worse PFS among patients with T790M loss (2.10 versus 6.35 months, p = 0.01).
Conclusions
Loss of the T790M mutation was associated with early resistance to osimertinib, and this was exacerbated by MET amplification. Further work is needed to fully understand the implications of each resistance mechanism.
With targeted therapy using tyrosine kinase inhibitors (TKIs), the prognosis of epidermal growth factor receptor (EGFR) gene-mutated non-small-cell lung cancer (NSCLC) has dramatically improved [
]. Osimertinib, the third-generation EGFR TKI that overcomes the T790M mutation–induced drug resistance associated with first- and second-generation TKIs, has further increased the progression-free survival (PFS) of patients with NSCLC acquiring that mutation [
]. However, resistance to osimertinib eventually occurs as well. Various molecular alterations have been identified in osimertinib-resistant patients [
High MET amplification level as a resistance mechanism to osimertinib (AZD9291) in a patient that symptomatically responded to crizotinib treatment post-osimertinib progression.
Acquired resistance of EGFR-mutant lung cancer to a T790M-specific EGFR inhibitor: emergence of a third mutation (C797S) in the EGFR tyrosine kinase domain.
], yet it is still unclear which plays the cardinal role in developing resistance. Furthermore, data on the clinical impact of each resistance-associated mutation are limited.
Because of the invasiveness of tissue biopsies, recent studies have attempted to elucidate the resistance mechanisms using liquid biopsies [
Sequential liquid biopsies reveal dynamic alterations of EGFR driver mutations and indicate EGFR amplification as a new mechanism of resistance to osimertinib in NSCLC.
]. However, liquid biopsies have not been able to identify histologic transformations, which comprise approximately 5%–7% of EGFR TKI resistance cases [
]. Therefore, tumour tissue is still the gold-standard source for clinical molecular analyses and sufficient tissue analyses on osimertinib-resistant patients and is needed to clarify the mechanisms for developing resistance. Next-generation sequencing (NGS) allows the simultaneous sequencing of multiple genomic alterations within a relatively short time. Coupling NGS with the sequential biopsies along the disease course of each patient would help identify the mechanisms for tumour resistance and provide a basis for the development of fourth- or fifth-generation EGFR TKIs.
We explored resistance mechanisms of second-line osimertinib and their clinical implications by comparing histologic changes and NGS results of paired tumour tissues before and after resistance acquisition in patients with EGFR-mutant adenocarcinoma (ADC) who progressed on osimertinib.
2. Material and methods
2.1 Study population
We enrolled patients with advanced EGFR-mutant ADC who underwent tumour biopsies and targeted NGS at the time of progression on osimertinib. All patients had a history of previous treatment with first- or second-generation EGFR TKIs and, at the time of osimertinib initiation, had EGFR T790M mutations detected by quantitative polymerase chain reaction assays using either tumour tissue or plasma. For comparison – except when resistance mechanisms (such as acquired C797S mutation or histologic transformation) could be clearly described after osimertinib treatment was completed or stopped – NGS was also performed on archived tumour tissues excised from each patient before osimertinib initiation.
Participants provided informed consent and allowed access to their cancer treatment records. Patient information, including sex, age, types of baseline EGFR mutations, histologic diagnoses, treatment histories, and treatment outcomes, were retrieved from electronic medical records. This study was approved by the institutional review board of Asan Medical Center (AMC), Seoul, Korea (approval number: 2015-1328).
2.2 NGS and genomic analysis
Genomic DNA was extracted from previously collected formalin-fixed, paraffin-embedded tissue specimens. Multiple sequencing assays were used in this study, including an in-house panel developed at the AMC (OncoPanel AMC, versions 3 and 4). The OncoPanel AMC, version 3 (OP AMC v3) and version 4 (OP AMC v4) were run using the MiSeq and NextSeq platforms (Illumina; San Diego, CA, USA) and captured 383 and 322 cancer-related genes, respectively (OP AMC v3, 199 genes for entire exons, eight genes for partial introns, and 184 genes for hotspots; OP AMC v4, 225 genes for entire exons, six genes for partial introns, and 99 for hotspots).
The sequence mapping steps for OP AMC v3 and v4 were performed in accordance with a method described elsewhere [
Therapeutic relevance of targeted sequencing in management of patients with advanced biliary tract cancer: DNA damage repair gene mutations as a predictive biomarker.
]. Germline variants of candidates for somatic variants (found in ≥1% of samples) were filtered out with a common germline variants database [dbSNP (build 141), gnomAD, common germline variants from 1100 healthy Koreans] [
Qualitative or categorical variables are presented as frequencies and proportions. Continuous variables are presented as medians with interquartile ranges (IQRs) or 95% confidence intervals (CIs). PFS on osimertinib was calculated from the date of osimertinib treatment initiation to the date of progressive disease or death. Overall survival (OS) on osimertinib was also calculated from the date of osimertinib treatment initiation to the date of death. OS data were censored if the patients were still alive at the time of analysis (July 31, 2019). Survival curves were estimated using the Kaplan-Meier method and compared using the log-rank test. All statistical analyses were performed using IBM SPSS Statistics for Windows, version 24.0 (IBM Corp., Armonk, NY, USA), with p values less than 0.05 considered statistically significant.
3. Results
The analysis included 34 patients with NSCLC ADC treated at AMC. Nine patients (26.5%) were men, and the median age at osimertinib initiation was 60.5 years. Regarding the founder EGFR mutation, the deletion at exon 19 was found in 20 patients, and the L858R point mutation was seen in 14 patients. The patients’ characteristics, including previous anti-EGFR TKI information, are summarized in Table 1. Details for individual patients, along with the biopsy site and time point of the tissue evaluated, are also presented in Supplemental Table 1.
Table 1Baseline characteristics of all included patients.
3.1 Identification of acquired osimertinib resistance alterations
When the pathological evaluation was performed on post-osimertinib tissues upon progression, histological changes were noted in five patients (14.7%): three with small-cell lung cancer (SCLC) and two with squamous cell carcinoma (SqCC).
Among the 29 patients who retained ADC histology, NGS analyses revealed that T790M mutations were maintained in seven patients (24.1%) upon disease progression with osimertinib. Of these patients, four had acquired EGFR C797S mutations (all of which were identified in cis with T790M) (Fig. 1), one developed MET amplification and one developed RET fusion. EGFR amplification was seen in the remaining one patient, but unlike with other cases, it was not known whether this developed after osimertinib treatment because the comparative pre-osimertinib tissue of that patient was obtained at the time lung cancer diagnosis and not at the time of osimertinib initiation.
Fig. 1The configuration of the EGFR T790M and C797S mutations shown in four patients. (A) Two patients for C797S (T to A); (B) one patient for C797S (GC to CT); (C) one patient for C797S (T to A and G to C).
Among the 22 patients (75.9%) with loss of the T790M mutation, a variety of novel mutations, such as KRAS, PIK3CA, and EGFR G724S mutations, and copy number variations, including MET, EGFR, and ERBB2 amplifications, were also identified. The most frequently observed accompanying variation with T790M loss was MET amplification (n = 4, 18.2% of those with loss of T790M). The overall mutation spectrum and suggested associated resistance mechanisms are described in Fig. 2.
Fig. 2Mutation spectrum and landscape of potential osimertinib resistance mechanisms.
Of note, the frequency of histologic transformation and the loss of T790M mutation or acquired MET amplifications (with a retained ADC tissue pattern) were not significantly different between the patients harbouring EGFR exon 19 deletion and L858R mutation (4/20 [20.0%] versus 1/14 [7.1%], p = 0.379; 13/16 [81.2%] versus 9/13 [69.2%], p = 0.667; 2/16 [12.5%] versus 3/13 [23.1%], p = 0.632).
3.2 Survival analyses according to the different resistance mechanisms
With a median follow-up of 25.58 months (95% CI: 18.15–33.02) after starting osimertinib treatment, the median PFS and OS durations on osimertinib of all 34 patients were 7.07 months (95% CI: 1.80–12.34) and 20.65 months (95% CI: 16.77–24.53), respectively. As shown in Fig. 3A and B, survival outcomes on osimertinib among patients who experienced histologic transformations (n = 5) were not significantly different from those who retained the ADC histologic type (n = 29) (SCLC and SqCC versus ADC) in terms of PFS (median 5.26 versus 9.11 months, p = 0.699) and OS (median 17.69 versus 20.65 months, p = 0.749).
Fig. 3The Kaplan-Meier survival analysis for osimertinib treatment. (A–B) PFS and OS of all patients based on the transformed histology; (C–D) PFS and OS according to the loss or retention of EGFR T790M mutation in patients who retained an adenocarcinoma histologic pattern; (E–F) PFS and OS according to MET amplification status in patients with loss of EGFR T790M mutation. PFS, Progression-free survival; OS, Overall survival.
When the outcomes were compared among the patients who retained the ADC pattern, PFS on osimertinib was significantly shorter among patients with T790M loss (n = 22) than among patients maintaining the T790M mutation (n = 7) (median 5.36 versus 13.81 months, p = 0.009), and patients who acquired MET amplification in addition to T790M loss (n = 4) were found to have much worse PFS than other patients who lost the T790M mutation (n = 18) (median 2.10 versus 6.35 months, p = 0.01) (Fig. 3C,E). Moreover, of the five patients who acquired MET amplification at the time of resistance, the one patient who maintained the T790M mutation had the longest PFS on osimertinib (9.11 months). These differences were also observed in the OS analysis – patients with loss of the EGFR T790M mutation had significantly shorter OS (median 17.13 versus 36 months, p = 0.02), and accompanying MET amplifications further shortened OS (median 5.98 versus 19.13 months, p = 0.083) (Fig. 3D,F).
4. Discussion
The substitution of threonine with methionine at amino acid position 790 (T790M) of exon 20, which affects the adenosine triphosphate (ATP) binding pocket of the EGFR kinase domain, increases the affinity for ATP and reduces the potency of any ATP-competitive kinase inhibitor [
]. Osimertinib is a novel irreversible, covalent third-generation EGFR TKI that overcomes this mutation, but acquired resistance to osimertinib can also develop after about a year of treatment. The patients in this study achieved 9.11 months of PFS on osimertinib after failure of first- or second-generation EGFR TKIs, similar to previous reports [
]. Numerous osimertinib resistance mechanisms have been suggested, and the reported results of mutation sites and mutation rates have varied widely between studies. Histologic transformation, acquired tertiary EGFR resistance via new EGFR point mutations, activation of bypass receptor tyrosine kinase signalling, aberrations in downstream signalling pathways, and target loss are the major categories to have been proposed to explain acquired resistance to osimertinib [
Histologic transformation is a common phenomenon for all generations of EGFR-TKIs, but the frequency of this phenomenon differs across the different EGFR TKIs. A recent study [
] reported osimertinib to be associated with the highest rate (14.5%) of histologic transformation, and this was in line with our results. This highlights the importance of tissue in the recent era of liquid biopsy, and negative liquid biopsy should lead to tissue biopsy when possible. Although we did not perform NGS comparisons for all of the patients with histologic transformation in this study, the molecular mechanisms involved in this transformation have been actively studied. Retinoblastoma 1 (RB1) mutation (especially the loss of RB1), TP53 mutation, lack of EGFR expression, and MYC amplification have been identified as representative mutations for transformation from NSCLC to SCLC [
Squamous cell transformation as a mechanism of acquired resistance to tyrosine kinase inhibitor in EGFR-mutated lung adenocarcinoma: a report of two cases.
The exon-20 EGFR C797S mutation is another well-known acquired resistance mutation to osimertinib. As the C797 residue is required by osimertinib to covalently bind to EGFR [
Landscape of acquired resistance to osimertinib in EGFR-mutant NSCLC and clinical validation of combined EGFR and RET inhibition with osimertinib and BLU-667 for acquired RET fusion.
], and in our cohort, we identified that 11.4% of patients developed C797S mutations. The combination of first- and third-generation EGFR TKIs has been found to restore EGFR inhibition among patients with both T790M and C797S on different alleles. Furthermore, as the EGFR C797S mutation is not associated with traditional EGFR TKIs, re-challenging the prior EGFR TKIs to T790M-negative tumours would be an attractive therapeutic option [
The allelic context of the C797S mutation acquired upon treatment with third-generation EGFR inhibitors impacts sensitivity to subsequent treatment strategies.
]. However, all of the patients acquiring C797S mutations in this study maintained the T790M mutation in the cis location, and this was in line with a previous study showing that almost all T790M mutations were present in cis with activating mutations [
]. There is still no ideal treatment option for these patients, and fourth-generation TKIs are yet to be developed.
In this study, the loss of T790M was seen much more frequently than the maintenance of T790M, and this was associated with poorer survival, apparently associated with early resistance. Our findings concurred with previous reports [
Heterogeneity underlies the emergence of EGFRT790 wild-type clones following treatment of T790M-positive cancers with a third-generation EGFR inhibitor.
], most of which suggested intratumoral genomic heterogeneity as a plausible explanation of the diminution of the T790M mutation when resistance developed. That is, when osimertinib targets potently against EGFR T790M mutation, selective pressure may occur wherein coexisting subclones with alternative competing resistance features can become the dominant population, which leads to early resistance to osimertinib [
]. Our findings supported this suggestion, given that various alterations were newly detected in most patients who lost the T790M mutation. Furthermore, we showed that among those with T790M loss, the prognosis for patients with MET amplification was significantly worse than for those with other mutations, indicating that prognosis depends not only on the loss of T790M itself but also on the nature or type of accompanying resistance alterations.
The MET signalling pathway has previously been suggested to correlate with carcinogenesis in NSCLC [
The tumour-stromal interaction between intratumoral c-Met and stromal hepatocyte growth factor associated with tumour growth and prognosis in non-small-cell lung cancer patients.
], and a recent meta-analysis further asserted this association between high MET CNGs and worse prognosis, which was more pronounced in Asian patients with ADC [
]. Furthermore, the longest PFS on osimertinib in a patient with MET amplification maintaining the T790M mutation compared with the rest of MET-amplified patients losing T790M in this study might provide further evidence as to which dominant subclone would most affect individual prognosis. Therefore, finding out which alterations became the main mechanism facilitating the development of resistance to osimertinib is important for predicting prognosis, and subsequent treatment strategies should be established to target those dominant resistant variations. Currently, however, there is no proven way to determine whether the T790M-positive population is dominant or secondary [
]. Moreover, given that such various alterations were not detected during pre-osimertinib tissue analysis, it is difficult to know which subclones other than T790M existed at the time of osimertinib treatment initiation and which should be treated as targets to prevent or delay the emergence of resistance.
A number of mutations proposed in previous studies as the causes of resistance were also found in our patients. However, their clinical impact could not be assessed because of the low prevalence of each case. It was suggested that acquired resistance due to the activation of these bypass tracks could be overcome through dual pathway suppression [
]. Of note, one of our patients who acquired MET amplification and maintained the T790M mutation participated in a clinical trial of the MET inhibitor, savolitinib, which was administered concurrently with osimertinib. Nevertheless, she had a poor prognosis, with just 4.04 months of PFS with concurrent savolitinib (Fig. 4).
Fig. 4Clinical course and outcomes of the patient who acquired MET amplification with maintained the T790M mutation. This patient treated with savolitinib in combination with osimertinib after progression on osimertinib. Although she tolerated well, her disease had progressed after 4.04 months of concurrent savolitinib treatment, so she started to receive conventional chemotherapy with gemcitabine and carboplatin.
There were several limitations in this study that need to be acknowledged. There was a lack of information regarding the tissue samples for NGS analysis, in terms of tumour cellularity and quality of extracted DNA. Above all, we could not cover intertumor heterogeneity, which refers to differences between a primary tumour and its matched metastases in this study. Moreover, this heterogeneity issue may be further considered in cases where the tissue biopsy sites before and after osimertinib treatment were different. Besides, the acquired variations were not clearly identified in five patients, as their pre-osimertinib tissues were obtained at the time of their initial cancer diagnosis rather than before osimertinib treatment initiation. Furthermore, numerous novel variants were detected in post-osimertinib tissue analysis, yet it was unclear whether they contributed to the development of resistance to osimertinib. Further studies are needed to determine the true significance of each alteration and reveal which one plays the cardinal role in establishing resistance to osimertinib.
We demonstrated that a variety of resistance-associated alterations developed with osimertinib treatment and that prognosis varied according to the type of mutation. The dominant alteration in each patient should be considered before planning subsequent treatment after osimertinib, and future research is warranted to determine which patients are susceptible to developing early resistance to osimertinib and would benefit from early intervention, such as with combination targeted therapies, to facilitate a more sustained response to osimertinib.
This study was supported by a grant from the National R&D Program for Cancer Control, Ministry of Health & Welfare, Republic of Korea [grant number: 1520220].
Conflict of interest statement
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this article.
Appendix A. Supplementary data
The following is/are the Supplementary data to this article:
High MET amplification level as a resistance mechanism to osimertinib (AZD9291) in a patient that symptomatically responded to crizotinib treatment post-osimertinib progression.
Acquired resistance of EGFR-mutant lung cancer to a T790M-specific EGFR inhibitor: emergence of a third mutation (C797S) in the EGFR tyrosine kinase domain.
Sequential liquid biopsies reveal dynamic alterations of EGFR driver mutations and indicate EGFR amplification as a new mechanism of resistance to osimertinib in NSCLC.
Therapeutic relevance of targeted sequencing in management of patients with advanced biliary tract cancer: DNA damage repair gene mutations as a predictive biomarker.
Squamous cell transformation as a mechanism of acquired resistance to tyrosine kinase inhibitor in EGFR-mutated lung adenocarcinoma: a report of two cases.
Landscape of acquired resistance to osimertinib in EGFR-mutant NSCLC and clinical validation of combined EGFR and RET inhibition with osimertinib and BLU-667 for acquired RET fusion.
The allelic context of the C797S mutation acquired upon treatment with third-generation EGFR inhibitors impacts sensitivity to subsequent treatment strategies.
Heterogeneity underlies the emergence of EGFRT790 wild-type clones following treatment of T790M-positive cancers with a third-generation EGFR inhibitor.
The tumour-stromal interaction between intratumoral c-Met and stromal hepatocyte growth factor associated with tumour growth and prognosis in non-small-cell lung cancer patients.
00088-5&id=gr1.jpg){kind=link}
00088-5&id=gr2.jpg){kind=link}
00088-5&id=gr3.jpg){kind=link}
00088-5&id=gr4.jpg){kind=link}