Clinical Modality of Resistance and Subsequent Management of Patients with Advanced Non-small Cell Lung Cancer Failing Treatment with Osimertinib

Non-small cell lung cancer (NSCLC) accounts for approximately 80% of all lung cancer cases [1]. Approximately 50% of Asian patients with NSCLC harbor activating epidermal growth factor receptor (EGFR) mutations [2, 3]. In EGFR-mutant lung adenocarcinoma, targeted therapy with EGFR–tyrosine kinase inhibitors (TKIs) performs substantially better than standard cytotoxic chemotherapy in terms of efficacy and tolerability [4‐7].

Osimertinib is an oral irreversible third-generation EGFR-TKI, which targets both activating mutations and the T790M resistance mutation, while sparing wild-type EGFR [8]. Several clinical trials have shown higher response rates (60–70%) and longer progression-free survival (PFS) for osimertinib than platinum-based chemotherapy for advanced NSCLC patients with acquired T790M-mediated resistance to prior early-generation EGFR-TKIs [9]. However, despite initial benefits, progression inevitably occurs after about 10 months [9‐11].

Efficacy of continued EGFR inhibition, cytotoxic chemotherapy, and combination therapy were reported after progression on early-generation EGFR-TKIs in the pre-osimertinib era [12‐14]. Such clinical management options are all also used in patients who develop resistance to osimertinib, but data are lacking and clinical benefits are unclear. There is currently no established standard treatment for patients resistant to osimertinib. Hence, exploration of optimal therapy once osimertinib has failed is an area of highly unmet medical need. In this study, we assessed disease course and therapeutic options in patients with advanced NSCLC after progression on osimertinib in a real-world setting.

To improve the understanding of the clinical modality of resistance to osimertinib and to investigate survival outcomes of post-resistance treatments, we retrospectively assessed the medical records of 65 osimertinib-resistant NSCLC patients. The data revealed that subsequent chemotherapy was a key prognostic factor for longer survival and should be considered in patients developing PD to osimertinib.

Our data indicate that the diversity of the clinical modality of osimertinib failure could help inform strategies for subsequent treatment and predict prognosis. Median OS and pOS were significantly longer in patients with local or gradual progression than in those with dramatic progression. Switching to chemotherapy improved survival compared with other therapeutic regimens in the dramatic progression population. For the local or gradual progression population, OS and pOS data were not mature in our study, but similar data have been reported after resistance to early-generation EGFR-TKIs in the pre-osimertinib era [15]. Previous data demonstrated a phenomenon of rapid clinical progression after discontinuation of EGFR-TKIs [16, 17]; thus, inhibition of the EGFR pathway beyond progression was explored in early-generation EGFR-TKIs. In the phase II, single-arm ASPIRATION (Asian Pacific trial of Tarceva as first-line in EGFR mutation) study, the median OS in patients who continued erlotinib beyond PD was 33.6 (95% CI 27.3–34.3) months compared with 22.5 (95% CI 20.1–27.0) months for patients who did not continue erlotinib treatment [13]. In IMPRESS (Iressa Mutation-Positive Multicentre Treatment Beyond ProgRESsion Study), a randomized phase III study, patients who continued gefitinib plus chemotherapy presented an OS improvement compared with patients who received chemotherapy only after progression on first-line gefitinib. The LUX-Lung 7 study showed a median TD beyond progression of 2.7 months for afatinib and 2.0 months for gefitinib [18]. Other available evidence was mostly from small retrospective studies [19‐21]. Yu et al. [22] reported an impressive OS of 41 months and PFS of 10 months in 18 NSCLC patients who received local treatment with continued TKI after oligometastatic PD on EGFR-TKIs. Although the results were mostly from retrospective studies, and may be biased by investigators’ choice, continuation of EGFR-TKIs with or without local treatment has become an option for patients with local or slow progression in the pre-osimertinib era.

Continued osimertinib treatment beyond progression similarly resulted in a median post-progression TD of 4.1 months in our study, providing additional clinical benefit, but showed no OS benefit compared with patients who did not continue in the overall non-selected population.

Subsequent chemotherapy after PD, however, resulted in a survival benefit in non-selected patients in our study. Longer previous exposure to EGFR inhibitors, the presence of more heterogenous tumors, and more complex molecular mechanisms of acquired resistance to osimertinib compared with early-generation EGFR-TKIs may help explain why survival was more favorable with chemotherapy in our study, especially for the subgroup of patients with dramatic progression.

Intrathoracic failure, rather than isolated CNS progression, was more frequently seen on osimertinib, which may further explain the significance of chemotherapy after osimertinib progression. CNS metastases affect 30–50% of NSCLC patients throughout the course of the disease [23], which has been partly blamed on the inferior blood–brain barrier (BBB) penetration by early-generation EGFR-TKIs. Locoregional CNS failure would not have been considered to be due to acquired systemic resistance to early-generation EGFR-TKI therapy, and continuous TKI administration beyond the occurrence of CNS PD was indicated. As BBB penetration of osimertinib is improved over early-generation EGFR-TKIs, in our study only ten (15.4%) cases presented with intracranial failure alone. Chemotherapy may perform better once osimertinib failure has occurred, considering the reasons mentioned earlier in this section. Le et al. [24] recently reported a study on osimertinib continuation beyond progression in EGFR-mutant NSCLC. The study showed that continuation of osimertinib beyond PD was associated with a longer OS than was discontinuation (11.2 vs. 6.1 months, p = 0.02), which was different to our study. However, for patients receiving additional treatment without osimertinib, only eight of 24 (33.3%) patients received chemotherapy in this study compared with 17 of 26 (65.4%) in our study. The difference in post-progression treatments between the studies may have resulted in different outcomes.

Various resistance mechanisms to osimertinib and subsequent treatments have been reported. T790M loss and T790M retention were two basic modes, which may inform strategies for subsequent treatment and predict survival. Our study showed a longer osimertinib TD post-progression in patients who retained the T790M mutation than in those who presented T790M loss, which means T790M-retained tumors may benefit more from osimertinib continuation treatment. Unfortunately, reassessment of molecular information in our study was performed in limited cases, and median OS was not reached. Further studies exploring molecular resistance information-based treatment selection are needed.

Additional EGFR mutations, such as the C797S mutation, were the most common mechanisms of resistance to osimertinib, and combining EAI045 or afatinib with cetuximab could be used to treat such patients [25, 26]. The combination of chemotherapy, bevacizumab, and atezolizumab in the IMpower (Immunology power) 150 study showed a median PFS of 9.7 months in an EGFR-mutated population who have failed to respond to a prior EGFR-TKI [27]. Other treatment strategies such as combining chemotherapy and pembrolizumab in the Keynote 789 trial are also ongoing. Although additional new treatment strategies are emerging, new medications or inclusion in clinical trials may not be possible or available for all patients. Chemotherapy could still be an option, especially in chemotherapy-naïve patients and when the mechanism of resistance is unknown. A study assessed the efficacy of platinum-doublet chemotherapy in NSCLC patients harboring EGFR mutations as second-line treatment and demonstrated that PFS and ORR were not different between T790M-positive and T790M-negative populations [28]. The study indicated that chemotherapy might be a feasible option after osimertinib failure, regardless of the T790M status at the time of resistance.

Our study was limited by its single-center, retrospective design, and relatively small sample size of patients with post-osimertinib molecular tests. The heterogeneity of treatment lines and subsequent treatments and the overlap of patients who continued osimertinib and who received chemotherapy may have biased the survival results of the different post-osimertinib treatment approaches in our study. In addition, few patients in our study were re-biopsied; thus, potential histologic transformation may have influenced survival outcomes.

Once osimertinib resistance has occurred in pretreated NSCLC patients, re-biopsy and reassessment of molecular information is recommended to guide subsequent treatment options. In cases where the mechanism of resistance is unknown or new medications or inclusion in clinical trials are not available, chemotherapy is a feasible option and can lead to prolonged survival in the non-selected population. Continued osimertinib beyond PD in the local or gradual progression population and T790M-retained patients may provide additional clinical benefit, and further study is needed regarding whether combination of continued osimertinib with chemotherapy provides further survival benefit.