Brief Report: Combination of Osimertinib and Dacomitinib to Mitigate Primary and Acquired Resistance in EGFR -Mutant Lung Adenocarcinomas

,


Introduction
Twenty percent of patients with metastatic lung adenocarcinomas have somatic activating mutations in the EGFR gene (1,2). Osimertinib, a third generation EGFR tyrosine kinase inhibitor (TKI), is approved in the first-line setting in patients with EGFR-mutant lung cancers given the progression-free survival (PFS) and overall survival (OS) benefit observed over other EGFR TKIs (3,4). Similar to earlier generation EGFR TKIs, responses to osimertinib are incomplete with eventual disease progression. In up to 35% of cases in the later-line setting and 15% of cases in the first-line treatment setting, acquired alterations in EGFR such as C797S or G724S mutations mediate resistance to osimertinib (5). Similar to development of EGFR T790M after treatment with earlier generation EGFR TKIs, osimertinib binds to EGFR at C797, and a mutation at this location impedes drug binding. Interestingly, in the presence of the original EGFR activating mutation (i.e., exon 19 deletion or L858R) and EGFR C797S without EGFR T790M, cells retain sensitivity to firstand second-generation EGFR TKIs (6). Other reports also demonstrated that cells with acquired resistance via G724S were sensitive to earlier generation EGFR TKI afatinib (7,8).
Dacomitinib is a pan-HER inhibitor with superiority over other earlier generation EGFR TKIs as initial treatment for EGFR-mutant lung cancers (9). With the aim of preventing on-target resistance to osimertinib, we designed a phase I study using evolutionary modeling-optimized dose levels of combination dacomitinib and osimertinib as first-line treatment for patients with EGFR-mutant lung cancer. The modeling and initial dose-finding experience from the phase II study were previously published (10). For patients with tumors resistant to osimertinib with acquired on-target EGFR alterations, we designed a phase I study of dacomitinib with or without osimertinib after progression on osimertinib. Dacomitinib monotherapy for unselected patients without acquired EGFR alterations after initial osimertinib treatment was also assessed and deemed ineffective in our prior report (11). Herein, we present a combined final report from these two phase I studies of combination osimertinib and dacomitinib in patients with EGFRmutant lung cancers as a means to address on-target EGFRmediated resistance.

Patients and Methods
Both trials were conducted after approval of the institutional review board at Memorial Sloan Kettering Cancer Center according to the Declarations of Helsinki. All patients provided written informed consent prior to participation

Study designs
Dacomitinib with osimertinib as initial therapy The initial therapy (first-line trial) was a phase II study in patients with EGFR-mutant lung cancers without prior EGFR TKI treatment (NCT03810807). The primary objective was to determine the maximum tolerated dose of the combination of osimertinib and dacomitinib. Secondary objectives were to measure best overall response, PFS and OS.

Dacomitinib with and without osimertinib after disease progression on osimertinib
The second trial was a prospective phase I/II trial in patients with EGFR-mutant lung cancers with disease progression on initial osimertinib (NCT03755102). We report on patients that received combination osimertinib and dacomitinib or single-agent dacomitinib with acquired second-site alterations in EGFR prior to study enrollment. The primary objective was to determine the overall response rate of osimertinib and dacomitinib after progression on osimertinib. Secondary objectives were to measure PFS and OS. This study was initially designed as evaluation of single-agent dacomitinib in all patients after progression on osimertinib. Given the lack of efficacy observed for dacomitinib alone in unselected patients after an interim analysis, the protocol was amended to evaluate the role of combination osimertinib and dacomitinib after progression on osimertinib in only patients with acquired second-site EGFR alterations. The first 10 patients enrolled in this study who received dacomitinib alone were published previously (11). The dose-finding experience for the initial study was also published previously (10).

Patients and inclusion criteria
Patients treated on study had stage IV or recurrent lung cancers with a somatic activating mutation in EGFR. For the initial therapy trial, no prior EGFR TKI therapy was permitted. For the acquired resistance trial, patients demonstrated radiologic progression during treatment with initial osimertinib and were required to have had a repeat biopsy after osimertinib progression that identified an acquired EGFR mutation in addition to the original EGFR sensitizing mutation.

Study assessment
Treatment cycles were 4 weeks in duration. Patients were assessed after 2 weeks for the first cycle and every 4 weeks subsequently. Toxicity was graded according to the NCI's Common Terminology Criteria for Adverse Events (CTCAE). Response to therapy was assessed by imaging every 8 weeks with response evaluated per RECIST 1.1.

Translational Relevance
Primary and acquired resistance to osimertinib remain significant challenges for patients with EGFR-mutant lung cancers. We conducted two prospective, phase I/II trials assessing combination osimertinib and dacomitinib to address on-target resistance in the primary and acquired resistance settings. Combination osimertinib and dacomitinib was effective at preventing acquired secondary alterations in EGFR, but this treatment was associated with greater toxicity. Our study highlights the need to develop better strategies to address on-target resistance in patients with EGFR-mutant lung cancers.

Statistical analysis
Safety and tolerability were summarized using descriptive statistics and CTCAE version 5 grading. Response rates were calculated using binomial proportions and exact 95% confidence intervals (CI). PFS was estimated using the Kaplan-Meier method and defined as the time from start of study treatment until progression or death. OS was estimated using the Kaplan-Meier method and defined as the time from start of study treatment until death.

Next-generation sequencing
Patients with available pre-and post-treatment tumor specimens underwent next-generation sequencing with MSK-IMPACT; cell-free DNA from peripheral blood was interrogated with MSK-ACCESS as described previously (12).

Data availability statement
Data will be made available upon reasonable request to the corresponding author.

Patients
Patient characteristics are presented in Table 1. From August 2019 to May 2020, 22 patients were enrolled on the initial therapy study and treated with the combination of osimertinib and dacomitinib. From February 2019 to October 2021, 7 patients were enrolled on the acquired resistance study and treated with either the combination osimertinib and dacomitinib (n ¼ 5) or dacomitinib alone (n ¼ 2). For the initial therapy study, 14 patients discontinued study treatment due to progression, 2 due to toxicity, 2 withdrew consent, and 5 patients remain on study. For the acquired resistance study, 6 patients discontinued study treatment due to progression, and 1 patient continues on study.

Safety and tolerability
All patients were evaluable for toxicity. Findings from the phase I dose escalation to identify the recommended phase II dose (RP2D) for combination therapy were published previously (10). Data regarding dose reductions are included in Supplementary Figs. S1 and S2. The dose escalation identified osimertinib 80 mg daily and dacomitinib 30 mg daily to be the RP2D. Among 16 patients in the first-line trial treated at the RP2D, 10 patients (63%) required dose reductions ( Supplementary Fig. S1). Among 5 patients in the second-line trial treated at the RP2D, 1 patient required a dose reduction (20%; Supplementary Fig. S2).

Efficacy
For the initial therapy study, all patients had baseline and ontreatment radiologic assessments except one (withdrew before first assessment due to toxicity). Fifteen patients had a confirmed partial response, and one patient had a confirmed complete response, for an ORR rate of 73% (95% CI, 50%-88%; Fig. 1). Median PFS is 20.3 months (Supplementary Fig. S3A) and median OS is 36.1 months (Supplementary Fig. S3B). For the acquired resistance study, all patients had baseline and on-treatment radiologic assessment except one (died before first assessment). One patient had a confirmed partial response, for an ORR rate of 14% (95% CI, 1%-58%; Fig. 2). Median PFS is 5.6 months (Supplementary Fig. S4A) and median OS is 26.1 months (Supplementary Fig. S4B).

Genomic analyses
All patients had pretreatment biopsy samples. The main correlative objective of the first-line study was to determine the frequency of acquired on-target second-site EGFR mutations with the combination of osimertinib and dacomitinib. Among 16 patients with postosimertinib biopsy samples, none developed an acquired secondsite mutations in EGFR (0% observed rate; 95% CI, 50%-88%). As for other acquired mechanisms of resistance, 1 patient developed an EML4/ALK fusion, 1 developed a MET amplification, 1 acquired mutations in PIK3CA E545K and ATM E2272 , and 1 developed mutations in AKT2 E17K and ARID2. One patient underwent small cell transformation and another patient large cell neuroendocrine transformation (Fig. 1). In the acquired resistance study, the only patient with a confirmed partial response to the combination of osimertinib and dacomitinib had a baseline EGFR G724S, with a variant allele frequency (VAF) ratio (acquired EGFR/initial sensitizing EGFR mutation) of 61%. Of the 3 patients that were on therapy for ≥ 6 months, 2 had baseline G724S mutations. The individual that remained on osimertinib and dacomitinib for the longest (20 months) had an acquired EGFR alteration in L718V with a VAF of 88%. Four patients had post-study treatment samples. The patient with a confirmed PR acquired an EGFR C797S alteration. One patient developed a MET D1228N activating mutation on treatment (Fig. 2).

Discussion
We report a combined analysis of two distinct clinical trials combining dacomitinib with osimertinib in patients with EGFRmutant lung adenocarcinomas. As an initial therapy, we sought to prevent the development of a second-site acquired EGFR alteration(s) and delay progression on combination treatment. In patients with acquired second-site EGFR mutations and resistance to osimertinib, we sought to reverse on-target resistance. As initial combined treatment, we detected no acquired second-site EGFR alterations, demonstrating that dual inhibition of EGFR may be effective at preventing second-site acquired alterations to EGFR. However, 4 patients in the first-line study developed acquired resistance to EGFR including in secondary pathways such as MET and EML4/ALK, further underpinning the need to prevent the development of these acquired bypass mechanisms. The median PFS of 20.1 months for the combination was not substantially different from the 18.9 months observed with osimertinib alone (3) and there was substantial added toxicity attributed to dual inhibition of EGFR.
Given the complexity of combination drug dosing, employing an evolutionary modeling approach has the potential to incorporate tumor heterogeneity and intersubject variability of plasma drug concentrations to predict tumor evolution using different combination dosing regimens. Despite this strategy to inform dosing, dual inhibition of EGFR proved too toxic in this setting. As a single-agent, dacomitinib has significant toxicity (66% of patients require dose reduction with dacomitinib monotherapy) and therefore the combination of osimertinib and dacomitinib was limited by the toxicities of single-agent dacomitinib (13). There have been efforts to combine osimertinib with other earlier generation EGFR inhibitors such as afatinib, however this combination was also toxic and demonstrated limited efficacy (14). A different study combined osimertinib with gefitinib as first-line treatment, with 30% requiring ultimate discontinuation of one or both study drugs (15). These different studies highlight the challenge of dual EGFR inhibition therapies, especially in the first-line setting where the duration of treatment is long. Furthermore, the combination of osimertinib and dacomitinib did not successfully reverse resistance with acquired EGFR second-site mutations post-osimertinib, reinforcing the challenge of trying to reverse on-target resistance with combination therapy. Given the small number of patients enrolled in the second-line trial, no firm conclusions can be drawn about the efficacy of combination osimertinib and dacomitinib in the acquired resistance setting; however, the limited response and high toxicity did not justify studying this combination further especially in light of 4th generation EGFR inhibitors that target acquired EGFR alterations currently in clinical development (NCT05256290).
Significant heterogeneity exists among EGFR-mutant lung cancers (2), and risk factors such as concurrent alterations (TP53, RB1) or lack of ctDNA clearance might identify patients who may particularly benefit from treatment escalation with combination therapies up front. The patients that appeared to benefit most from therapy in the later-line setting largely had more clonal resistance alterations with higher variant allele frequencies. Our two studies highlight the need to develop better initial treatments to better eradicate persister cells from which resistance ultimately emerges, for example using antibody-drug conjugates in combination with EGFR inhibition, or up-front combinations of EGFR inhibitors with chemotherapy (16,17).