Despite initial progress with tyrosine kinase inhibitors (TKIs) for treating EGFR-mutated non-small-cell lung cancer (NSCLC), treatment resistance has become a significant obstacle. Dr Caroline McCoach (Helen Diller Family Comprehensive Cancer Center, CA, USA) and Dr Howard (Jack) West (Swedish Cancer Institute, WA, USA) survey insights gained into tumor biology and resistance mechanisms, as well as lessons learned from clinical trials, which may illuminate the way forward for treatment of EGFR-mutated NSCLC.
Introduction
The approval of EGFR TKIs 15 years ago allowed a dramatic leap in the clinical outcome of patients with EGFR-mutated NSCLC. The hope that this treatment generated for a cure was slightly diminished by the realization that resistance to TKIs is inevitable. Nevertheless, important insights into the mechanisms of resistance were gained by characterization of the drug-resistant tumor cells. These studies provided critical information about on- and off-target mechanisms of resistance and ultimately led to the development of treatments such as osimertinib. This drug was first approved in patients with activating EGFR mutations with concurrent T790M mutations. In this setting, the objective response rate (ORR) was 71% versus 31% for second-line chemotherapy (p<0.001), and progression free survival (PFS) was 8.5 months compared with 4.2 months for chemotherapy [1]. Other targetable bypass pathways, such as activation of the MET signaling pathway, have also been identified [2–4]. However, the magnitude of the challenge of resistance has become abundantly clear, particularly the fact that, as the number of lines of targeted therapy patients receive increases, the length of time on subsequent treatments decreases, while the molecular complexity of the tumors increase [5]. This process is known to occur in other molecular subtypes of lung cancer as well as other cancers [6–12].
Unraveling tumor biology and treatment resistance
In parallel with the development of multiple generations of EGFR-directed treatment, there have been numerous advances in understanding tumor biology. Among the most clinically relevant include the fact that there is substantial existing molecular heterogeneity of tumors prior to any systemic treatment, a finding that has been demonstrated in early and advanced-stage disease. In the lung TRACERx program, Jamal-Hanjani et al found that, in 100 early stage NSCLC patients, a median of 30% of somatic mutations and 48% of copy number alterations are subclonal, indicating that mutational processes are ongoing during tumor development [7]. Interestingly, these investigators also examined a population of patients with stage IV disease and demonstrated that as cancer metastasizes, it becomes more molecularly complex [13]. Thus, even in treatment-naïve patients the seeds of resistance are present and single-targeted therapies can provide selective pressure resulting in the growth of subclones with resistance mechanisms [14]. This stresses the need for diagnosing and characterizing tumors early, and highlights the need to identify treatments that can maximally control clonal “driver” mutations and anticipate or even prevent likely resistance pathways. These are critical points to consider as we move into the next phase of TKI treatment and the practical application of this information to treatment of NSCLC.
We have begun to see this evolutionary process play out in the first-line treatment of patients with EGFR-activating mutations. In the FLAURA trial, we saw a dramatic shift in PFS when comparing a third- with first-generation TKI and noted a PFS of 18.9 months compared with 10.2 months (hazard ratio [HR] 0.46, p<0.001) [15]. The data on overall survival (OS) are not mature, but the authors did report that at 18 months there was a marked survival difference between the two treatment arms (83% vs 71%; HR 0.63, p=0.007, not considered statistically significant due to O’Brien-Fleming rule). Critically important to this treatment is the fact that osimertinib crosses the blood brain barrier and produces a central nervous system (CNS) ORR of 66% versus 43% with first-generation EGFR TKI therapy (odds ratio 2.5, p=0.011). Given its superior efficacy and good tolerability, osimertinib was approved in the US as a first-line EGFR TKI and is currently the preferred systemic therapy in this setting for patients with EGFR-activating mutations. Notably, however, the profile of resistance differs between first and third generation TKIs due to their specific features. We are only beginning to determine the patterns of acquired resistance to osimertinib, though studies thus far indicate that there is no dominant pattern of resistance with osimertinib unlike, for example, the 50-60% frequency of the T790M mutation observed in cases of resistance to first- and second-generation EGFR TKIs. Instead, there are several different mechanisms, some of which are in complex signaling pathways or involve histologic changes, which indicate that it will be challenging to target resistance with molecularly directed therapies [16,17].
The way forward: revisiting combinations?
We must move away from considering first-line therapy in isolation, instead favoring an effort to minimize overall tumor burden, maximize performance status, and preserve the most targeted therapy options over several lines of therapy
Several phase III studies provide a window into the next wave of treatments for patients with EGFR-mutated NSCLC and demonstrate that ultimately the path forward may be to revisit previously explored combinations but to test them in a more patient-specific or timing-specific setting. The importance of focusing on the right patient population is historically illustrated in the Iressa Pan-Asia Study (IPASS) comparing gefitinib with carboplatin plus paclitaxel in the first-line treatment of advanced NSCLC, in which the benefit from gefitinib was only seen in the subgroup of patients with an activating EGFR mutation [18]. The potential value of revisiting prior combinations in timing-specific clinical scenarios was observed in the FASTACT-2 study of intercalated erlotinib with chemotherapy in Asian patients. In the subgroup with an EGFR mutation, the PFS was 16.8 month with chemotherapy and erlotinib therapy, compared with 6.9 months in the chemotherapy plus placebo group (HR 0.25 95% CI 0.16-0.39; p<0.001). Further, the median OS was 31.4 months in the combination group versus 20.6 months in the chemotherapy with placebo group, despite 85% of the placebo group going on to receive an EGFR TKI as subsequent treatment [19]. In retrospect, the benefit of the combination regimen is likely due to its ability to eradicate tumor cells with a baseline resistance mechanism to gefitinib, and thus provide a more durable response to treatment. Demonstrating the benefit in an updated study on a larger scale, NEJ009 evaluated combined carboplatin/pemetrexed with gefitinib or gefitinib alone in EGFR mutation-positive patients. This study reported that the three-drug combination demonstrated a superior OS compared with gefitinib monotherapy (52.2 months vs 38.8 months) [20]. Notably, patients assigned to gefitinib alone were significantly more likely to develop a more diffuse pattern of progression and decline in performance status than patients assigned to the combination of chemotherapy and gefitinib, which was associated with an attrition rate of approximately 25% of the patients progressing on first-line gefitinib monotherapy.
Though these results are impressive, we caution against applying these results to clinical practice unilaterally for several reasons. First, the study did not test the current standard-of-care EGFR TKI, osimertinib, so we do not know how this combination would work with chemotherapy or the safety profile of this regimen. Second, we do not know what the OS advantage of osimertinib is and it may surpass that of the combination regimen from NEJ009. Third, the OS data in the monotherapy arm far surpasses that of gefitinib in other contemporary studies, suggesting that these results may not be extrapolated to broader healthcare systems and patient populations.
Patients with advanced NSCLC harboring an activating EGFR mutation have an array of potentially effective therapies available that can be administered in sequence or in combinations, including one or more EGFR TKIs, platinum-based chemotherapy, (potentially) other targeted therapies, later salvage chemotherapy, and immune checkpoint inhibitors. In some cases, early advantages in PFS will be eroded to no difference in OS due to subsequent effective therapies over time, as we have seen with the erlotinib/bevacizumab combination studied in Japanese randomized trials [21,22]. However, we have also seen the promise from the NEJ009 trial and the concept of improved time to distant metastatic disease or death with local consolidation therapy for oligometastatic disease [23]. In addition, more aggressive management of lower-volume disease may be associated with a change in disease trajectory over a more prolonged period than is captured in a single line of therapy. We have also seen that the sequencing of even the same planned treatments over time can significantly alter survival in oncology settings ranging from molecularly unselected NSCLC [24] to metastatic prostate cancer [25,26]. Similarly, in the setting of ALK fusion-positive NSCLC replete with a wealth of ALK inhibitors, sequencing of the different ALK inhibitors is an area of active investigation [27].
Conclusions and future steps
Our treatment options for patients with advanced EGFR mutation-positive NSCLC offer varying degrees of efficacy, tolerability, and activity within the CNS. Osimertinib arguably provides the most compelling option as first-line therapy when considering all of these factors. Nevertheless, we must move away from considering first-line therapy in isolation, instead favoring an effort to minimize overall tumor burden, maximize performance status, and preserve the most targeted therapy options over several lines of therapy. To do so, it is time to focus more on OS, as it may be influenced by several lines of therapy that can have a sequence-sensitive impact. We should also avail ourselves of every opportunity to capture, identify, and eventually target nascent molecular mechanisms of resistance, from tissue and/or blood, over time. We now need to turn to a more longitudinal approach and view our treatment options together, rather than independently, in order to determine the best care to our patients with advanced EGFR mutation-positive NSCLC.
Caroline McCoach and Howard West are members of the Academic Thoracic Oncology Medical Investigator’s Consortium’ (ATOMIC)