Amivantamab Plus Lazertinib in Atypical -Mutated Advanced Non-Small Cell Lung Cancer: Results From CHRYSALIS-2.

INTRODUCTION
Non–small cell lung cancer (NSCLC) is a complex and heterogeneous disease that accounts for approximately 80%-85% of all lung cancer cases.1,2 Activating mutations in the epidermal growth factor receptor (EGFR) gene are commonly observed in NSCLC, with exon 19 deletions (Ex19del) and exon 21 L858R (L858R) substitutions being the most common (85%-90%).2-4 Exon 20 insertions (Ex20ins) and atypical EGFR mutations each account for approximately 5%-10% of all EGFR mutations in NSCLC.3-8 The most common atypical mutations are G719X in exon 18, S768I in exon 20, and L861Q in exon 21, but can include others.9,10 These can present in patients as single or compound mutations.11

CONTEXT

Key Objective

Does amivantamab plus lazertinib (amivantamab-lazertinib) exert clinically meaningful and durable antitumor activity in participants with advanced non–small cell lung cancer (NSCLC) harboring atypical epidermal growth factor receptor (EGFR) mutations?

Knowledge Generated

Amivantamab-lazertinib demonstrated durable antitumor activity in participants with atypical EGFR-mutated advanced NSCLC (objective response rate [ORR], 52%), including in those who were treatment-naïve (ORR, 57%) and those with previous treatment (ORR, 48%). No new safety signals were identified. At a median follow-up of 17.3 months, participants who received first-line amivantamab-lazertinib had a median progression-free survival of 19.5 months and a median response duration of 20.7 months, and median overall survival was not estimable.

Relevance (T.E. Stinchcombe)

This combination is another treatment option for patients with NSCLC harboring these rare EGFR mutations.*

*Relevance section written by JCO Associate Editor Thomas E. Stinchcombe, MD.

Patients with atypical EGFR-mutated advanced NSCLC have significantly worse outcomes compared with those with common EGFR mutations when receiving EGFR-targeted therapies.12 Consequently, overall survival (OS) is substantially shorter among patients whose tumors harbor atypical versus common mutations.12 Patients with NSCLC harboring compound atypical EGFR mutations exhibit even poorer survival compared with those with single atypical mutations.13,14
In a post hoc analysis of prospectively collected data from three trials including 38 patients with tumors harboring atypical EGFR mutations, the second-generation EGFR-tyrosine kinase inhibitor (TKI) afatinib demonstrated a median progression-free survival (PFS) of 10.7 months,15 resulting in afatinib's US Food and Drug Administration (FDA) approval for first-line treatment of patients with nonresistant EGFR mutations (S768I, L861Q, and G719X) other than Ex19del or exon 21 L858R substitutions and European Medicines Agency (EMA) approval for EGFR-TKI-naïve patients with advanced/metastatic NSCLC harboring uncommon EGFR mutations.16,17 However, this study included participants with tumors harboring atypical mutations with or without compound common EGFR mutations (Ex19del or L858R substitutions), which could influence the benefit observed. Patients with tumors harboring atypical mutations alongside common mutations experience more favorable outcomes with EGFR-targeted therapies compared with those with only atypical compound mutations.11,18-21 While not currently indicated for this population, the third-generation EGFR-TKI osimertinib is used globally for first-line treatment of patients with atypical EGFR-mutated advanced NSCLC.22-26
Amivantamab is an EGFR-MET bispecific antibody that exerts antitumor activity through multiple mechanisms of action, including inhibition of ligand binding, endocytosis and degradation of receptors, and immune cell–directing activity.27-30 It is approved for use in several indications for patients with both common mutations and Ex20ins, alone or in combination with other agents. Lazertinib, a highly selective, CNS-penetrant, third-generation EGFR-TKI, has demonstrated activity in patients with atypical EGFR-mutated advanced NSCLC.31,32 In the LASER301 study, among patients harboring common EGFR mutations, alone or in combination with other EGFR mutations, lazertinib markedly improved efficacy over gefitinib for first-line treatment of EGFR-mutated advanced NSCLC and is approved as a monotherapy in the Republic of Korea.33,34 Combining these two agents results in synergistic EGFR inhibition, with amivantamab targeting the extracellular domains of EGFR and MET, thereby being unaffected by intracellular resistance mechanisms, whereas lazertinib binds to the EGFR receptor intracellularly.30,35 Amivantamab-lazertinib has shown potent clinical activity across a wide range of EGFR alterations.36-39 In the phase III MARIPOSA trial, amivantamab-lazertinib improved the median PFS by 7.1 months versus osimertinib in patients with treatment-naïve, common EGFR-mutated advanced NSCLC (hazard ratio [HR], 0.70 [95% CI, 0.58 to 0.85]; P < .001). In addition, in the protocol-specified final OS analysis of MARIPOSA (median follow-up of 37.8 months), amivantamab-lazertinib showed a statistically significant and clinically meaningful improvement in OS versus osimertinib (HR, 0.75; P < .005), with the median OS benefit projected to exceed 1 year.40 Amivantamab-lazertinib is FDA-approved as first-line treatment for patients with advanced NSCLC with EGFR Ex19del or L858R substitutions.41
In Cohort C of the CHRYSALIS-2 study (ClinicalTrials.gov Identifier: NCT04077463), we prospectively evaluated the efficacy and safety of amivantamab-lazertinib in participants with advanced NSCLC harboring atypical EGFR mutations, excluding those with compound common mutations.

PATIENTS AND METHODS

Participants
Eligible participants for CHRYSALIS-2 Cohort C were 18 years and older and had advanced or metastatic NSCLC. Participants with tumors harboring any atypical activating EGFR mutations, including but not limited to S768I, L861Q, and G719X, were eligible. Those with solitary Ex20ins or coexisting Ex20ins or Ex19del/L858R EGFR mutations were excluded. Participants had an Eastern Cooperative Oncology Group performance status (ECOG PS) score of 0-1. Participants were treatment-naïve or had received ≤2 previous lines of treatment with chemotherapy and/or first- or second-generation EGFR-TKIs as the most recent line of therapy; the receipt of third-generation EGFR-TKIs was exclusionary. Participants with stable, treated CNS metastases were eligible.

Study Design and Treatment
CHRYSALIS-2 is an open-label, 2-part, phase I/Ib, multicenter study of lazertinib, alone or in combination with amivantamab, in participants with advanced EGFR-mutated NSCLC (Data Supplement, Fig S1, online only). This analysis presents the results from Cohort C, which evaluated amivantamab-lazertinib in participants with tumors harboring atypical EGFR mutations. Subgroup analyses based on the line of treatment (first-line v later-line treatment with amivantamab-lazertinib) were performed. Participants in Cohort C received amivantamab-lazertinib in 28-day cycles until disease progression, unacceptable toxicity, noncompliance, withdrawal of consent, or discontinuation at the investigator discretion. Lazertinib was dosed at 240 mg orally once daily. Amivantamab was administered intravenously once every week during Cycle (C) 1 at a dose of 1,050 mg (1,400 mg for ≥80 kg weight), with the first dose split between 2 days (350 mg once daily on C1 Day [D] 1, and the remainder on C1D2) and then every 2 weeks in subsequent cycles.

End Points and Assessments
The primary end point was objective response rate (ORR) per RECIST v1.1, as determined by the investigator. Secondary end points included duration of response (DoR), clinical benefit rate (CBR; defined as a complete response [CR], partial response [PR], or stable disease [SD] for ≥11 weeks), PFS, OS, time to treatment discontinuation, and adverse events (AEs). Biomarkers were analyzed as exploratory end points.
Disease was assessed at baseline and every 6 weeks (±1 week) via contrast-enhanced computed tomography (CT; noncontrast CT was acceptable if contrast administration was contraindicated), magnetic resonance imaging, and other imaging/examination scans per RECIST v1.1.
AEs were recorded from consent until 30 days after the last treatment dose or start of subsequent anticancer therapy and graded according to the National Cancer Institute-Common Terminology Criteria for Adverse Events v5.0. Testing for atypical mutations, including solitary versus compound EGFR analyses, was based on local testing. Blood samples were collected at baseline for circulating tumor DNA (ctDNA) analyses and tested centrally. Next-generation sequencing of plasma ctDNA was performed and analyzed using Guardant360 (Guardant Health, Redwood City, CA), excluding participants from China.

Trial Oversight
The trial was conducted in accordance with the provisions of the Declaration of Helsinki, Good Clinical Practice guidelines (as defined by the International Council for Harmonisation), and applicable regulatory and country-/territory-specific requirements. The protocol was approved by the local institutional review boards and independent ethics committees of the participating centers. Participants provided written informed consent.

Statistical Analysis
The sample size for Cohort C was based on the null hypothesis that the ORR is ≤50% and the alternative hypothesis that the ORR is ≥68%. With a one-sided alpha of 2.5% and a power of 90%, 92 response-evaluable participants were needed for expansion Cohort C. Assuming a nonevaluable rate of 15%, approximately 108 participants were planned to be enrolled in Cohort C.
Data were summarized using descriptive statistics. Time-to-event end points were summarized using Kaplan-Meier estimates. The medians and 95% CIs were provided. No data imputation was applied for missing safety and efficacy evaluations. All data used for this interim analysis were reported before January 12, 2024.

RESULTS

Participants
As of January 12, 2024, 105 participants received amivantamab-lazertinib in Cohort C. Among the 105 participants, 49 participants were treatment-naïve, whereas 56 participants had received a previous treatment. The most common treatments among those who received previous therapy were EGFR-TKIs (88%).
Baseline demographic and disease characteristics are shown in Table 1. The median age was 64 years (range, 30-85); 50% was male, and 68% was Asian. The most common types of atypical EGFR mutations, including both single and compound mutations, were G719X (56%), L861X (26%), and S768X (23%); 29 (28%) participants had compound mutations. A total of 37 (35%) participants had brain/CNS metastasis confirmed by imaging at baseline.

Efficacy

Overall Population
At a median follow-up of 16.1 months (range, 0.1-31.5), the median duration of treatment in the overall population was 11.1 months (range, 0.03-31.5). The investigator-assessed ORR was 52% (95% CI, 42 to 62). All responses observed were PRs (52%; Fig 1 and Table 2). The median DoR was 14.1 months (95% CI, 9.5 to 26.2), and the CBR was 79% (95% CI, 70 to 86). The median PFS was 11.1 months (95% CI, 7.8 to 17.8; Fig 2A), and the median OS was not estimable (NE; 95% CI, 22.8 to NE; Fig 2B).

Treatment-Naïve Subgroup
At a median follow-up of 17.3 months (range, 0.1-31.5), the median duration of treatment with first-line amivantamab-lazertinib was 12.7 months (range, 0.03-31.5). The investigator-assessed ORR was 57% (95% CI, 42 to 71; Fig 1 and Table 2). The median DoR was 20.7 months (95% CI, 9.9 to NE), and the CBR was 84% (95% CI, 70 to 93). The median PFS was 19.5 months (95% CI, 11.2 to NE), and median OS was NE (95% CI, 26.3 to NE), with 77% of participants still alive at 24 months. To provide context for this single-arm trial, a trial-matched real-world analysis was conducted on existing therapies from the Flatiron Health NSCLC database, which are described in the Data Supplement.

Previous Treatment Subgroup
At a median follow-up of 15.4 months (range, 0.3-30.8), the median duration of treatment with later-line amivantamab-lazertinib was 8.9 months (range, 0.2-29.9). The investigator-assessed ORR was 48% (95% CI, 35 to 62). PR was observed in 27 (48%) participants, SD in 19 (34%) participants, and PD in six (11%) participants (Fig 1 and Table 2). The median DoR was 11.0 months (95% CI, 4.5 to NE), and the CBR was 75% (95% CI, 62 to 86). The median PFS was 7.8 months (95% CI, 5.4 to 11.1), and the median OS was 22.8 months (95% CI, 16.9 to NE).

Safety
In the overall population, the safety profile of amivantamab-lazertinib was consistent with that previously reported (Table 3).37-39 Individual treatment-emergent AEs (TEAEs; by preferred term) most commonly reported by participants were rash (67%), paronychia (67%), hypoalbuminemia (59%), and infusion-related reactions (IRRs; 56%) and were mostly grade 1 and 2. The majority (95%) of IRRs occurred during the first infusion and were grade 1 and 2. Overall, grade ≥3 TEAEs were reported in 73 (70%) participants. Fifty (48%) participants experienced grade ≥3 treatment-related AEs. The most frequent grade ≥3 treatment-related AEs reported by participants were rash (13%) and hypoalbuminemia (8%). Serious TEAEs were reported for 53 (50%) participants, 22 (21%) of which were considered related to treatment. Dose interruptions, reductions, and discontinuations of any study agent because of TEAEs were seen in 73 (70%), 52 (50%), and 29 (28%) participants, respectively. Discontinuation of all study agents because of treatment-related AEs occurred in seven (7%) participants. Death occurred in 12 (11%) participants, one (1%) of which was considered related to treatment.
Venous thromboembolism (VTE) was reported in 31 (30%) participants; most events occurred in the first 4 months of treatment (71% [22 of 31]). Grade 3 VTE was observed in 10 (10%) participants; no grade 4 and 5 events occurred. At baseline, 16% (17 of 105) of all participants received anticoagulation for prophylaxis or based on the medical history of VTE. Most participants (94% [29 of 31]) were not on anticoagulation at the time of first VTE.

Exploratory Analyses

Biomarker Analysis
Among participants treated with first-line amivantamab-lazertinib and among those treated with later-line amivantamab-lazertinib, most participants had solitary EGFR mutations. Among participants treated with first-line amivantamab-lazertinib, the confirmed ORR was 63% (95% CI, 44 to 79) for those with solitary mutations and 47% (95% CI, 23 to 72) for those with compound mutations (nominal P = .299). The corresponding values among those treated with later-line amivantamab-lazertinib were 52% (95% CI, 37 to 67.5) and 33% (95% CI, 10 to 65), respectively (nominal P = .244). The presence of solitary or compound EGFR mutations had no major impact on ORR, as indicated by the 95% CIs (Fig 1).
Of the 73 participants who had analyzable baseline ctDNA data, 65 (89%) participants had detectable ctDNA and 59 (81%) participants had any detectable pathogenic alterations. Of these participants, TP53 comutations were detected in 37 (63%) participants. The ORR was 54% (95% CI, 37 to 71) among participants with TP53 mutations and 46% (95% CI, 28 to 66) among those without TP53 mutations.
Participants were further classified using the structure-based approach described by Robichaux et al.42 In the overall population, P-loop and αC-helix compressing (PACC) mutations were observed in 38 participants, classical-like mutations in 14 participants, and T790M-like mutations in three participants. The ORRs in these groups were 45% (95% CI, 29 to 62), 64% (95% CI, 35 to 87), and 67% (95% CI, 9 to 99), respectively (Fig 3).

DISCUSSION
In Cohort C of the CHRYSALIS-2 study, after a median follow-up of 16.1 months, amivantamab-lazertinib demonstrated clinically meaningful and durable antitumor activity in participants with atypical EGFR-mutated advanced NSCLC; OS was NE at the time of analysis. Among participants who were treatment-naïve, the ORR by investigator review was 57%, with a median PFS of 19.5 months, which is substantially longer than the <1-year PFS reported for participants receiving EGFR-TKI monotherapy.10,15 Participants from CHRYSALIS-2 Cohort C had a 24-month OS rate of 77%. Amivantamab-lazertinib also showed benefits among participants who received previous treatments in Cohort C, with an ORR rate of 48% and a median PFS of 7.8 months. These results suggest that amivantamab-lazertinib could be a treatment option for patients with atypical EGFR-mutated advanced NSCLC.
Afatinib, the only approved agent in this setting, demonstrated a median PFS of 10.7 months and an ORR of 71% in a post hoc analysis of data from three trials involving 38 participants with atypical EGFR mutations.15 In a prospective study of 40 participants with atypical EGFR-mutated NSCLC, osimertinib demonstrated a median PFS of 9.4 months and an ORR of 55%. Although osimertinib demonstrated an ORR of 86% among participants who harbored solitary L861Q mutations (n = 7), the ORR among those who harbored G719X solitary mutations (n = 10) was only 30%.10 In a retrospective, multicenter study of participants with atypical EGFR-mutated NSCLC in Germany, afatinib and osimertinib demonstrated a median PFS of 12.0 and 5.0 months, respectively (independent of the treatment line).43 However, these studies also included participants with tumors harboring atypical mutations with common compound EGFR mutations (Ex19del or L858R), potentially influencing the observed benefits. In another phase II study conducted in participants with tumors harboring EGFR mutations other than Ex19del, L858R, T790M, and Ex20ins in Korea, osimertinib demonstrated an ORR of 50% (n = 36) and a PFS of 8.2 months.44
Although atypical mutations vary from other common EGFR mutations and structure-based classifications suggest that a structure-function–based approach could predict drug sensitivity to targeted therapies,42 the sensitivity of each mutation to EGFR-TKIs is still not fully understood. Nonetheless, data from CHRYSALIS-2 Cohort C showed that amivantamab-lazertinib is effective against all atypical mutations, including single and compound atypical mutations, irrespective of the structural implication of the mutation locus in the EGFR gene. Among all participants with PACC mutations, amivantamab-lazertinib demonstrated an ORR of 45%, regardless of previous treatment. Firmonertinib is another EGFR-TKI that has demonstrated efficacy in TKI-naïve participants with PACC mutations, with an investigator-assessed ORR of 52% for the 160 mg once daily dose and 82% for the 240 mg once daily dose. Amivantamab-lazertinib has also demonstrated efficacy irrespective of the presence of high-risk features including baseline liver metastases, TP53 comutations, and detectable baseline circulating tumor DNA.45
The complementary mechanisms of action of amivantamab and lazertinib30,32 expand coverage against additional EGFR (eg, C797S) and MET (eg, amplification) TKI resistance mutations,38 potentially delaying disease resistance and prolonging disease control beyond that observed with other EGFR-TKIs alone. These mechanisms may explain the improved efficacy of amivantamab-lazertinib in this setting. In addition, next-generation sequencing of ctDNA samples at baseline showed that the presence of TP53 mutations was not associated with a lower response rate with amivantamab-lazertinib.
The safety profile of amivantamab-lazertinib was consistent with previous reports, with no new safety signals.37-39 IRRs were mostly grade 1 and 2 and primarily occurred during the first infusion cycle. IRRs can be further reduced with prophylactic oral dexamethasone 8 mg twice daily plus standard prophylaxis before first intravenous amivantamab infusion.46 VTE was reported in 30% of participants; none were grade 4 and 5, and most events occurred in the first 4 months of treatment. The majority of participants were not receiving anticoagulation at the time of first VTE. Prophylactic anticoagulation is now recommended for the first 4 months of treatment with amivantamab-lazertinib in all ongoing trials and in the prescribing information.37,41,47,48 In the PALOMA-3 study, prophylactic anticoagulation was safely implemented and effective in reducing the risk for VTE. Among all participants in PALOMA-3, VTE rates for those who received prophylactic anticoagulation was 10% versus 21% for those who did not receive anticoagulation. The subcutaneous formulation of amivantamab may further reduce overall VTE rates.47
There are some limitations in our study. This was a phase I proof-of-concept study conducted in a limited population, and ORR was not analyzed by blinded independent central review. However, the totality of the efficacy data presented, including ORR, DoR, PFS, and OS, demonstrates that amivantamab-lazertinib can be used as a clinically meaningful treatment in a patient population with very few available options. As shown in our real-world analyses (Data Supplement, Methods and Results), most participants treated with first-line physician-selected EGFR-TKI monotherapy do not receive second-line treatment. Furthermore, amivantamab-lazertinib demonstrates improved survival outcomes compared with physician-selected EGFR-TKIs, including afatinib and osimertinib, in the real-world setting after applying key trial eligibility criteria and adjusting for confounding differences. Of note, the real-world analyses may be subject to potential selection biases, variability in treatment approaches, and confounding effects of subsequent therapies. However, rigorous statistical methods were used, such as propensity score methods, to ensure that the comparative analyses were robust and reflective of true clinical scenarios. Key eligibility criteria from the trial were used to identify the target population, and an e-value analysis was performed to measure the impact of unmeasured confounding.
In conclusion, amivantamab-lazertinib demonstrated clinically meaningful and durable antitumor activity while maintaining safety in participants with atypical EGFR-mutated advanced NSCLC. To our knowledge, this is the largest, single-cohort, prospective study of atypical EGFR-mutated advanced NSCLC. Our results contribute to findings of other studies that demonstrated the efficacy and safety of amivantamab-based regimens in participants with advanced NSCLC harboring common EGFR mutations,49
EGFR Ex20ins,50 and now atypical EGFR mutations.