Specifying the choice of EGFR-TKI based on brain metastatic status for advanced NSCLC with EGFR p.L861Q mutation

Abstracts

Background

In-depth insight into the genomic features of the uncommon EGFR p.L861Q mutant NSCLC is scarcely performed, and no consensus on the preferred treatment strategy has been established. Moreover, the therapeutic implications of EGFR-TKI stratified by clinical and molecular features remained largely unknown.

Methods

A multi-center NGS database comprising 44,993 NSCLC samples was utilized for the genomic landscape profiling of EGFR p.L861Q mutation. Furthermore, a real-world cohort of 207 patients harboring EGFR p.L861Q mutation with complete treatment history was curated for comprehensive clinical analysis.

Results

L861Q is prevalent in approximately 2.1% of EGFR-mutated NSCLC and is typically co-mutated with EGFR p.G719X on the same allele (20%) and exhibits co-occurrent EGFR copy number amplification in approximately 17% of cases. In the first-line setting, afatinib and third-generation EGFR-TKI have been shown to yield notably superior treatment outcomes compared to first-generation EGFR-TKI (1^st vs.2^nd vs.3^rd generations, ORR: 15.8% vs.56.5% vs.46.7%, P=0.01; median PFS: 6.4 vs.13.5 vs.15.1 months, P=0.002). This finding consistently held for patients without CNS metastases (1^st vs.2^nd vs.3^rd generations, median PFS:6.0 vs.18.2 vs.14.1 months, P=0.003). In contrast, third-generation EGFR-TKI demonstrated superior efficacy compared to afatinib or first-generation TKI among the subgroup of brain metastasis (Pooled 1^st/2^nd-generation vs.3^rd-generation TKI, brain ORR:0.00% vs.33.33%; median PFS:7.9 vs.19.3 months, P=0.021). Additional concurrent EGFR mutations or EGFR amplification did not yield a discernible impact on the efficacy of EGFR-TKI.

Conclusions

The present study comprehensively elucidates the molecular features of EGFR p.L861Q mutation and underscores the optimal therapeutic choice of first-line EGFR-TKI based on brain metastatic status.

Introduction

Epidermal growth factor receptor (EGFR) is one of the most prevalent genetic alterations in non-small cell lung cancer (NSCLC), which mutates in around 50% of Asian patients and 20% of Western patients, respectively[1,2]. Tyrosine kinase inhibitor (TKI) targeting EGFR has substantially revolutionized the treatment paradigm of NSCLC[3,4]. Clinical practice has shown that classical EGFR mutations, namely exon 19 deletion and exon 21 L858R substitution, respond well to the first to third-generation EGFR-TKI[[5], [6], [7]]. However, uncommon EGFR mutations represent a diverse subgroup that includes a range of rare mutations. The most frequent of these are G719X, L861Q, S768I, and exon 20 insertions, collectively referred to as “major uncommon mutations”[8]. The rarity of individual mutations presents a challenge to personalized treatment, as most previous studies assess uncommon EGFR mutations as a whole group to achieve the power to calculate statistical significance[9,10]. However, the specific mutation subtype greatly impacts the effectiveness of different EGFR-TKIs[11]. Consequently, each uncommon EGFR mutation is expected to be considered individually. Except for pharmacodynamics, the pharmacokinetics such as the central nervous system (CNS) distribution of different agents further complicates the clinical scenarios.

The EGFR p.L861Q mutation is the second most prevalent among non-classical EGFR mutations, following the p.G719X alteration[12,13]. Of note, similar to the classical EGFR p.L858R mutation, EGFR p.L861Q resides within the activation loop, distanced from the binding pocket structure, leading to enhanced EGFR heterodimerization and aberrant activation while leaving the binding activity unaffected[[14], [15], [16]]. The optimal treatment strategies for advanced NSCLC patients with EGFR p.L861Q mutation have yet to be formulated. Given the significant improvement in survival benefits, afatinib has garnered approval from the Food and Drug Administration and the European Medicines Agency regarding NSCLC with any sensitizing EGFR mutation, including both common and uncommon EGFR mutations[17]. Nevertheless, approximately 40% of patients treated with afatinib 40 mg daily necessitated a dose reduction among patients harboring uncommon EGFR mutations, as indicated by a post hoc analysis of the Lux-Lung 2/3/6 trials[17]. Concerns regarding drug tolerability have spurred the trial of third-generation EGFR-TKI in patients with uncommon EGFR mutations. A prospective phase II trial from South Korea unveiled compelling efficacy evidence for osimertinib, showcasing promising treatment responses in this subset of patients[18]. Consequently, both afatinib and osimertinib have been recommended as first-line treatment options by the National Comprehensive Cancer Network (NCCN) guidelines.

Approximately 70% of patients with EGFR-mutated NSCLC will experience brain metastasis, which undermines the survival advantages conferred by EGFR-TKI[19]. Compared to afatinib, osimertinib, and other third-generation EGFR-TKI have a more favorable safety profile and superior ability of intracranial penetration[5,20,21]. Consequently, the clinical selection of EGFR-TKI necessitates a weighted consideration of clinical and genetic features, such as CNS metastases, concurrent mutations, and associated toxicities. Considering the strict inclusion criteria under highly controlled settings, patients with brain metastases and uncommon mutations tend to be under-represented in randomized controlled trials. Hence, this study aimed to comprehensively delineate the molecular features of EGFR p.L861Q mutant NSCLC and to explore the optimal treatment strategies for patient subsets with distinct clinical features.

Methods

Enrolled patient

In the Burning Rock genomic database, we screened formalin-fixed, paraffin-embedded (FFPE) tissue, pleural effusion or peripheral blood specimens of NSCLC submitted for comprehensive genomic profiling (CGP) during routine clinical care from multi-centers in China. Patients with EGFR p.L861Q mutation were further enrolled in the analysis to characterize its prevalence and genomic landscape.

In addition, a real-world cohort of NSCLC patients with EGFR p.L861Q mutation at the Sun Yat-Sen University Cancer Center (SYSUCC) was identified spanning from December 2018 to December 2023, with detailed clinical annotation accessible for this cohort. Patients with stage IIIB-IV NSCLC with complete follow-up and treatment efficacy data on the first-line EGFR-TKI were further enrolled to evaluate the clinical response of different EGFR-TKIs. Considering that the structural changes of EGFR p.L861Q mutation compared with wild-type EGFR are similar to those of EGFR p.L858R mutation, we also enrolled patients with classical EGFR p.L858R mutation from SYSUCC in the same period as a comparison cohort.

Written informed consent has been obtained from each patient before enrollment in the real-world cohort of SYSUCC. The current study was conducted under the Institutional Review Boards at SYSUCC. All research procedures conformed to the principles of the Helsinki Declaration.

Genetic analysis

GCP was performed in CLIA-certified labs using hybridization capture-based next-generation sequencing (NGS) in the Burning Rock genomic database. The utilized NGS panels, which covered 8, 68, 168, or 520 cancer-related genes, are inconsistent, but all panels have full coverage of the EGFR gene. Indexed samples were sequenced on Nextseq 500 (Illumina, Inc., CA, USA) with paired-end reads and an average sequencing depth of 1,000 × for tissue samples and 10,000 × for liquid biopsy samples. Library preparation, quality controlling, sequencing, and data analysis followed the same standardized pipeline for all samples. To avoid the impact of different gene counts between these inconsistent panels, genomic mutation analysis utilized the shared 163 genes from 168, and 520 panels (Supplementary Table 1).

For the SYSUCC cohort, targeting NGS detected in the SYSUCC Diagnostics Personalized Panel included a 1021-gene panel for solid tumors. A hybrid captured-based NGS assay covering approximately 1.1 megabases (Mb) of genomic sequences of 1021 cancer-related genes (GenePlus-Beijing, China) was used for the sequencing. Genomic alterations assessed in the SYSUCC Personalized Diagnostics Panel included SNVs, short Indel, copy number (CN) alterations, and rearrangement. Targeted capture sequencing required a minimal mean effective depth of coverage of 300x in tissue samples. The sequencing coverage and detailed description of the SYSUCC Diagnostics panel are summarized in Supplementary Table 2. In addition, a subset of patients in the real-world dataset was sequenced using other targeted NGS panels at other institutions or received the Amplification Refractory Mutation System-Polymerase Chain Reaction(ARMS-PCR) method for molecular analyses.

Data collection and response evaluation

Patients’ genetic information and baseline clinicopathologic characteristics (including age, gender, smoking status, pathology, clinical TNM stage, and brain metastasis status) and treatment histories were retrospectively collected from electronic medical records at SYSUCC. The clinical outcomes of interest included objective response rate (ORR) and progression-free survival (PFS) of first-line EGFR-TKI. Two investigators independently assessed the radiographic tumor response following the Response Evaluation Criteria in Solid Tumors version 1.1. Subsequent to the administration of EGFR-TKI, patients typically underwent computed tomography (CT) scan in the first month and regular imaging every 8-12 weeks thereafter to assess the treatment efficacy. ORR was defined as the sum of patients with complete response (CR) or partial response (PR) out of all evaluable patients. PFS refers to the time from commencing treatment to objective tumor progression or death from any cause. Besides, for patients with CNS lesions, two investigators independently evaluated the response assessment in neuro-oncology brain metastases (RANO-BM).

Statistical analysis

Differences in the baseline clinicopathologic features and treatment response between patient groups were compared via χ² or Fisher exact tests. PFS was assessed using the Kaplan-Meier method with a log-rank test. Waterfall plots were used to visualize the treatment efficacy for individual patients. All statistical analyses were performed in the R-4.3.1 software, with all tests being two-sided and statistical significance set at P<0.05.

Results

Genomic characteristics of EGFR p.L861Q mutation in NSCLC

CGP results for 44,993 NSCLC patient samples in the Burning Rock genomic database were queried. Among these patients, 24,507 individuals had tumors harboring EGFR mutations, of which genetic alteration at L861Q occurred in 506 patients, representing a prevalence rate of 2.1% (Fig. 1A). Unlike other uncommon EGFR mutations (e.g EGFR exon 18 mutation)[22,23], EGFR p.L861Q tends to mutate independently (63.3%) and is less frequently identified as compound mutations. Detailed information on the distribution of co-occurring EGFR mutation in patients with EGFR p.L861Q mutation was illustrated in Fig. 1B. Within the L861Q cohort, 19.57% (99/506) also presented with concurrent EGFR p.G719X mutation (G719A: n=52, G719S: n=30, G719C: n=14, G719D: n=2 and G719R: n=1). Furthermore, the coexisting p.G719X mutation displayed a comparable variant allele frequency (VAF) to the p.L861Q mutation, indicating that these two EGFR mutations may be concomitant mutations existing on the same allele (Fig. 1C). Besides, other frequent co-occurring EGFR mutations included EGFR p.L858R/M/Q (2.37%), EGFR p.R776C/G/H/S (1.98%) and EGFR p.L833W/F. In terms of the co-occurring somatic alterations, TP53 ranked at the top (56%) among all genes, followed by MYC (8%), RBM10 (7%), and PIK3CA (6%) (Fig. 1D). Approximately 17% of EGFR p.L861Q mutant NSCLC cases exhibited co-occurrent EGFR copy number amplification.

Fig. 1.

Genomic characteristics of EGFR p.L861Q mutation in NSCLC—Burning Rock genomic database.

(A) EGFR p.L861Q mutation rate in Chinese patients diagnosed with NSCLC. B) Distribution of concurrent EGFR mutation in patients with EGFR p.L861Q mutation. C) Relative variant allele frequency (VAF) of EGFR p.G719X compared to EGFR p.L861Q mutation. D) Genomic landscape of patients with EGFR p.L861Q mutation.

Clinical and molecular characteristics of NSCLC patients with EGFR p.L861Q mutation

A secondary cohort of 207 patients harboring EGFR p.L861Q mutation in real-world practice was identified from the SYSUCC. The detailed basic characteristics of all enrolled EGFR p.L861Q mutant patients with lung cancer were summarized in Table 1. The cohort's median age was 62.0 (IQR, 56.0-69.0) years, with 52.7% of patients being women, and 68.1% being non-smokers. Most of the cases were adenocarcinoma (96.6%), and the number of stage IV, stage IIIB/C, or early-stage patients were 115 (55.6%), 9 (4.3%) and 83 (40.1%), respectively. Brain metastasis was recorded in 39 cases (18.8%) prior to initiating EGFR-TKI. EGFR p.L861Q mutation was detected by NGS testing in 40.6% of patients, while the remaining 59.4% of patients were detected using ARMS-PCR methods. In addition, a total of 632 patients with EGFR p.L858R mutation who received NGS testing via the SYSUCC Personalized Diagnostics Panel before the administration of EGFR-TKI were also included as the comparison cohort. The baseline demographic and clinical characteristics among patients with EGFR p.L861Q or EGFR p.L858R mutation were shown in Supplementary Table 3. Similar to patients with EGFR p.L861Q mutation, individuals of the female gender, non-smokers, and those with adenocarcinoma constituted the majority of the EGFR p.L858R mutation cohort.

Table 1.

Basic characteristics of all enrolled EGFR p.L861Q mutant patients with lung cancer.

Characteristics	Patients(n=207)
Age
Median [IQR], years	62.0[56.0,69.0]
≤60 years	86(41.5%)
>60 years	121(58.5%)
Sex, No. (%)
Female	109(52.7%)
Male	98(47.3%)
Smoking Status, No. (%)
Never	141(68.1%)
Current/ever	66(31.9%)
ECOG/PS, No. (%)
0	140(67.6%)
1	67(32.4%)
Pathology, No (%) ADC SCLC+ADC SCC ADC+SCC	200(96.6%) 3(1.4%) 2(1.0%) 2(1.0%)
TNM Stage, No. (%)
Stage I-IIIA Stage IIIB/IIIC	83(40.1%) 9(4.3%)
Stage IV	115(55.6%)
Extra thoracic metastasis, No. (%)
Yes	79(38.2%)
No	128(61.8%)
CNS metastasis, No. (%)
Yes	39(18.8%)
No	168(81.2%)
Detection Methods, No. (%) ARMS-PCR NGS	123(59.4%) 84(40.6%)

ADC, adenocarcinoma; ARMS-PCR, Amplification Refractory Mutation System-Polymerase Chain Reaction; NGS, next-generation sequencing; SCLC: small-cell lung cancer; SCC: squamous cell carcinoma.

Among patients identified through ARMS-PCR, the EGFR p.L861Q mutation typically manifests as a single mutation among up to 80% of patients. Co-occurrent EGFR p.G719X, EGFR p.L858R, and EGFR p.S768I were observed in 13.0%, 3.3%, and 1.6% of EGFR p.L861Q mutant patients, respectively (Fig. 2A). Similarly, for patients identified through NGS testing, patients with solely EGFR p.L861Q driver mutation comprise the major proportion of cases (72.6%) (Fig. 2B). EGFR p.G719X was the most prevalent co-occurring EGFR mutation with p.L861Q(13.1%), followed by EGFR p.L858R/M (6.0%), EGFR p.R776H (2.4%) and EGFR p.L833W/F (2.4%). Of note, the relative VAF of EGFR p.L861Q and concurrent EGFR mutation are comparable (Fig. 2C). Considering the inconsistent NGS panels in the real-world cohort, only the genomic results derived from the SYSUCC diagnostic personalized panels are adopted to analyze co-occurring somatic alteration events to avoid the interference of different gene counts. The most frequent co-occurring genetic alteration was TP53 (73.8%), RBM10 (14.3%), TERT (11.9%), LRP1B (11.9%), RB1 (7.1%), EPHA5 (7.1%), EPHA3 (5.7%), ASXL1 (7.1%) and EP300 (7.1%) (Fig. 2D).

Fig. 2.

Molecular characteristics of NSCLC patients with EGFR p.L861Q mutation in SYSUCC cohort.

A) Distribution of co-occurring EGFR mutation in patients with EGFR p.L861Q mutation detected by ARMS-PCR. B) Distribution of co-occurring EGFR mutation in patients with EGFR p.L861Q mutation detected by next-generation sequencing. C) Relative variant allele frequency (VAF) of EGFR p.G719X and other concurrent EGFR mutations compared to EGFR p.L861Q mutation. D) Genomic landscape of EGFR p.L861Q mutant patients with lung cancer.

Meanwhile, NSCLC with EGFR p.L858R mutations showed similarities with EGFR p.L861Q mutation in their genomic landscapes, namely high mutational frequency in TP53, RBM10, LRP1B, PIK3CA, CTNNB1 and RB1(Supplementary Fig. 1A). Besides, no significant difference was found in the mutational frequency of co-occurring somatic hypermutations between these two groups except a higher prevalence of TP53 mutation in EGFR p.L861Q mutant cases (Supplementary Fig. 1B). Of note, tumor mutation burden and PD-L1 expression level associated with EGFR p.L861Q mutations were not significantly different compared to EGFR p.L858R mutations (Supplementary Fig. 1C-D).

Treatment efficacy of different EGFR-TKI for NSCLC patients harboring EGFR p.L861Q mutation

The efficacy analysis included a cohort of EGFR p.L861Q mutant NSCLC patients who received the first-generation (n=19), second-generation (specifically afatinib, n=46), and third-generation (n=30) EGFR-TKI as the first-line targeted therapy with available follow-up data. Waterfall plots that depicted the maximum changes in the cumulative size of tumor target lesions from baseline were presented in Fig. 3A. Both afatinib and third-generation EGFR-TKI demonstrated notably superior treatment responses compared to the first-generation EGFR-TKI (1^st-generation EGFR-TKI vs. Afatinib vs. 3^rd-generation EGFR-TKI, ORR: 15.8% vs. 56.5% vs. 46.7%, P=0.01) (Fig. 3B). Furthermore, Kaplan-Meier survival analysis revealed that patients receiving afatinib or third-generation EGFR-TKI derived significantly greater survival benefits than those treated with first-generation EGFR-TKI (1^st-generation EGFR-TKI vs. Afatinib vs. 3^rd-generation EGFR-TKI, median PFS: 6.4 vs. 13.5 vs. 15.1 months, P=0.002, Fig. 3C). However, no statistically significant difference was found between the afatinib and third-generation EGFR-TKI group.

Fig. 3.

Treatment efficacy of different EGFR-TKIs for NSCLC patients harboring EGFR p.L861Q mutation.

A) Waterfall plots depicted the maximum changes in the sum of tumor target lesion size from baseline. B) Comparison in the overall response rate of different EGFR-TKIs among EGFR p.L861Q mutant patients. C) Comparison in the survival benefits of different EGFR-TKIs among EGFR p.L861Q mutant patients. D) Literature review: response to EGFR-TKIs (first-generation EGFR-TKIs, Afatinib or Osimertinib) in EGFR p.L861Q-mutant patients with advanced NSCLC.

Given the absence of head-to-head comparison of first-generation EGFR-TKI, afatinib, or third-generation EGFR-TKI for patients with EGFR p.L861Q mutation, we summarized the efficacy data of different EGFR-TKIs from previously reported single-arm studies. As demonstrated in Fig. 3D, the response rate varied from 39.6% to 60.0%, and the mPFS ranged from 6.0 to 8.9 months among those treated with first-generation EGFR-TKI[10,24,25]. In contrast, a more favorable treatment response appeared evident in those treated with afatinib or third-generation EGFR-TKI. For afatinib, the ORR of patients treated with afatinib ranged from 50.0% to 56.3%, with the mPFS ranging from 8.2 months to 15.6 months[17,[26], [27], [28]]. Meanwhile, the response rate to osimertinib ranged from 41.2% to 78.0%, and the mPFS ranged from 9.0 months to 22.7 months in previously reported data[18,[29], [30], [31]]. Nevertheless, despite comparable treatment effectiveness, various confounding factors merit consideration in the selection of an optimal drug between afatinib and third-generation TKI.

Effect of CNS metastasis and concurrent mutations on survival benefits from EGFR-TKI in NSCLC patients harboring EGFR p.L861Q mutation

CNS metastasis typically deteriorated the clinical outcomes for patients harboring EGFR p.L861Q mutation when treated with first-line EGFR-TKI (with vs. without brain metastasis, 7.9 vs. 15.1 months, P=0.0003) (Fig. 4A). It is noteworthy that this detrimental effect of CNS metastasis maintains significance in patients treated with afatinib (with vs. without brain metastasis, 7.9 vs. 18.2 months, P<0.0001) (Fig. 4B-C), whereas no significant difference was observed in patients treated with third-generation TKI (with vs. without brain metastasis, 19.3 vs. 14.1 months, P=0.56) (Fig. 4D). In other words, third-generation TKI may abolish the prognostic detrimental effect of CNS metastasis.

Fig. 4.

Effect of CNS metastasis and concurrent mutations on survival benefits from EGFR-TKIs in NSCLC patients harboring EGFR p.L861Q mutation

A) Detrimental effect of brain metastasis on the survival benefits of EGFR-TKIs for patients harboring EGFR p.L861Q mutation. B) Detrimental effect of brain metastasis for EGFR p.L861Q mutant patients treated with 2nd-generation EGFR-TKIs.C) Detrimental effect of brain metastasis for EGFR p.L861Q mutant patients treated with 3rd-generation EGFR-TKIs. D) Comparison in the treatment efficacy of different EGFR-TKIs among EGFR p.L861Q mutant patients without brain metastasis. E) Comparison in the treatment efficacy of different EGFR-TKIs among EGFR p.L861Q mutant patients with brain metastasis. F) Intracranial objective response rate per response assessment in neuro-oncology brain metastases and comparison in patients treated with different EGFR-TKIs. G) Impact of concurrent EGFR mutation on treatment response to EGFR-TKIs in patients with EGFR p.L861Q mutation. H) Impact of concurrent EGFR p.G719X mutation on treatment response to EGFR-TKIs in patients with EGFR p.L861Q mutation. I) Impact of concurrent EGFR amplification on treatment response to EGFR-TKIs in patients with EGFR p.L861Q mutation.

In the subgroup of patients without CNS lesions, both afatinib and the third-generation EGFR-TKI demonstrated enhanced survival benefits when compared to the first-generation EGFR-TKI (1^st- vs. 2^nd- vs. 3^rd-generation EGFR-TKI, median PFS: 6.0 vs. 18.2 vs. 14.1 months, P=0.003) (Fig. 4D). In contrast, the third-generation EGFR-TKI performed better than the first or second-generation EGFR-TKI among the subgroup of patients with brain metastasis (1^st/2^nd-generation EGFR-TKI vs. 3^rd-generation EGFR-TKI, median PFS: 7.9 vs. 19.3 months, P=0.02) (Fig. 4E). Furthermore, among patients with assessable or measurable disease in CNS lesions, the RANO-BM response rate is considerably higher when treated with third-generation EGFR-TKI than afatinib or first-generation EGFR-TKI (brain ORR: 33.33% vs. 0.00%) (Fig. 4F).

In terms of the concurrent EGFR mutations, no significant disparity in survival outcome to EGFR-TKI was observed whether the patients carried any compound EGFR mutations, EGFR p.G719X mutation or EGFR CN amplification (compound L861Q vs. single L861Q: 14.1 vs. 11.7 months, P=0.44; G719X mutant vs. G719X wild-type: 15.1 vs. 11.7 months, P=0.66; EGFR CN amplification vs. No EGFR CN amplification: 15.5 vs. 13.2 months, P=0.44, respectively) (Fig. 4G-I).

Potential resistance mechanisms of EGFR-TKI and subsequent treatment for patients with EGFR p.L861Q mutation

A total of 23 drug-resistance data were collected, including nine cases of first-generation EGFR-TKI, ten cases of afatinib, and four cases of third-generation EGFR-TKI. Around 39.1% of patients did not display precise or recognized acquired resistance genetic alterations. In line with previous reports[28,32], among patients with EGFR p.L861Q mutation, the emergence of secondary EGFR p.T790M mutations (26.3%) in the context of resistance appears relatively lower compared to classical EGFR mutations (∼50%). Besides, EGFR p.L718Q/V was detected in one EGFR p.L861Q mutant patient after failure of the third-generation EGFR-TKI. MET copy number amplification encompasses the majority of EGFR-independent resistance mechanisms of EGFR-TKI for patients with rare EGFR p.L861Q mutation (Supplementary Table 4).

In addition, a total of 11 patients with complete follow-up data who received platinum-based chemotherapy(n=8) or immunotherapy (n=3) after progression on EGFR-TKI were enrolled. Only two patients receiving chemotherapy achieved PR and five patients achieved stable disease (SD), while one experienced progression to disease. As for patients who received immunotherapy in combination with chemotherapy, PR was achieved in one patient, and SD was observed in two patients. The detailed treatment outcomes are shown in Supplementary Table 5.

Discussion

Due to the relatively lower prevalence of the EGFR p.L861Q mutation in NSCLC, a comprehensive exploration of the impact of compound genetic alterations and clinical features on the EGFR-TKI treatment outcome has not been conducted. Therefore, no consensus on the preferred treatment strategy has been established. In particular, the capacity of different EGFR-TKIs for CNS control of EGFR p.L861Q mutant NSCLC remains largely unknown. Our study systematically elucidates the molecular features of EGFR p.L861Q mutant NSCLC and initiatively formulates the optimal first-line therapeutic approach by incorporating CNS metastasis status into the therapeutic decision-making framework (Fig. 5).

Fig. 5.

Proposed therapeutic decision-making process for EGFR p.L861Q mutant NSCLC.

In concordance with the classical EGFR mutation profile, individuals of the female gender, non-smokers, and those with adenocarcinoma constituted a higher proportion of the EGFR p.L861Q mutation cohort[1,5]. With a prevalence of approximately 2.1% of EGFR mutant cases, EGFR p.L861Q usually mutates independently. However, EGFR p.G719X mutation can be the concurrent mutation on the same allele as EGFR p.L861Q mutation, observed in roughly 20% of cases. Besides, rare EGFR p.L861Q mutant cases exist as compound mutations with EGFR p.L858R/M/Q, p.R776C/G/H/S, and p.L833W/F mutation. Except for the driver mutations, genetic alterations in TP53 and RBM10 are the two most frequent concurrent somatic mutation events. Given that uncommon EGFR mutations, such as EGFR p.G719X and p.S768I mutations, tend to mutate in the form of compound mutations, EGFR p.L861Q mutations were less frequently identified as compound mutations[23]. These results could be attributed to the relatively tumorigenic solid potential of EGFR p.L861Q mutant cells, in which the enhancement of another EGFR-activating mutation is unnecessary.

Thus far, a high proportion of clinical evidence on treating uncommon EGFR mutations is obtained from the administration of afatinib[17,27,33,34]. Nevertheless, no validated differences in terms of clinical responsiveness among different generations of EGFR-TKI are available in the form of head-to-head comparison. While both afatinib and osimertinib are endorsed as the preferred treatment option by the NCCN guideline, the foundation of their relative clinical efficacy is somewhat limited[17,18]. Our findings from the sizeable real-world dataset suggested that afatinib and the third-generation EGFR-TKI may notably yield a more favorable treatment outcome compared to the first-generation EGFR-TKI for patients harboring uncommon EGFR p.L861Q mutation. Furthermore, findings from a literature review also affirm the relatively encouraging treatment responses observed with afatinib or third-generation EGFR-TKI among individuals with EGFR p.L861Q mutation[35]. Located in the exon 21 of EGFR, similar to EGFR p.L858R, the L861Q mutation is categorized as the “classical-like” type based on the structure classification[16]. Of note, our results also suggested that patients with nonclassical EGFR p.L861Q mutation and classical EGFR p.L858R showed similarities in their clinical and genomic features. Osimertinib has been confirmed to achieve better treatment outcomes than first-generation EGFR-TKI in the FLAURA study for common EGFR mutations[5], and afatinib has been conferred to provide a substantial improvement in survival benefits versus gefitinib revealed by the LUX-Lung 7 trial[36]. Consequently, the striking resemblance between the classical EGFR mutation and EGFR p.L861Q could partly explain the relatively favorable tumor response of afatinib or third-generation EGFR-TKI compared to the first-generation EGFR-TKI.

Despite the recommendation from NCCN guidelines, the FDA has not approved the third-generation EGFR-TKI, including osimertinib in uncommon EGFR mutations. Owing to their high blood-brain barrier penetration and milder toxic effects, the third-generation EGFR-TKI is now considered the initial treatment of choice for classical EGFR-mutant patients[5,37,38]. Therefore, oncology clinicians are encouraged to investigate third-generation EGFR-TKI in the context of uncommon EGFR mutations, especially in the setting of CNS metastasis. Expectedly, our results revealed that CNS metastasis detrimentally affects the survival benefits of EGFR-TKI for patients harboring EGFR p.L861Q mutation. Both afatinib and the third-generation EGFR-TKI could be used for uncommon EGFR p.L861Q mutant patients without CNS metastasis. On the contrary, in the presence of intracranial metastasis, third-generation EGFR-TKI instead of afatinib is recommended as the optimal treatment. The retrospective study conducted by Jair Bar et al. supports our proposal, in which osimertinib exhibited encouraging activity against CNS lesions[29]. However, given the relatively severe toxicity associated with afatinib, it is reasonable to recommend third-generation EGFR-TKIs as the favored therapeutic option for advanced NSCLCs with uncommon EGFR p.L861Q mutation. Furthermore, the low prevalence of secondary T790M mutation rates in resistant samples may suggest that the third-generation EGFR-TKI should be administered as a front-line therapy rather than being reserved for subsequent use following first- or second-generation EGFR-TKI.

This study is limited by its observational and retrospective nature. However, the low incidence of EGFR p.L861Q mutation poses a significant obstacle to conducting prospective trials. Moreover, patients in the treatment efficacy cohort come from a single center, which warrants further validation from other studies, although the relatively large sample size may assure the generalizability of our study results.

In conclusion, the present study comprehensively elucidates the molecular features of NSCLC with EGFR p.L861Q mutation. It underscores the third-generation EGFR-TKI as the optimal first-line treatment option for EGFR p.L861Q mutant patients with advanced NSCLCs, especially for patients with intracranial metastasis. Further multi-center studies with a prospective design are warranted to confirm and validate our findings.

Ethical Statement

This study was conducted by the Declaration of Helsinki Declaration (as revised in 2013) and approved by Sun Yat-sen University Cancer Center IRB (B2023-468-01). Written informed consent was obtained from all patients.

Funding

This work was financially supported by the Chinese National Natural Science Foundation Project (82173101 and 8237262), the Cancer Innovative Research Program of Sun Yat-sen University Cancer Center, and the 308-Program for Clinical Research of Sun Yat-sen University Cancer Center.

Data availability statement

The data supporting this study's findings are available on request from the corresponding author.

CRediT authorship contribution statement

Lan-Lan Pang: Writing – review & editing, Writing – original draft, Visualization, Software, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Wei-Tao Zhuang: Writing – review & editing, Writing – original draft, Visualization, Formal analysis, Data curation. Jun-Jun Li: Writing – review & editing, Writing – original draft, Visualization, Formal analysis, Data curation. Bing Li: Writing – review & editing, Writing – original draft, Visualization. Yi-Hua Huang: Writing – review & editing, Writing – original draft, Visualization, Methodology, Investigation. Jun Liao: Writing – review & editing, Writing – original draft, Visualization. Meng-Di Li: Writing – review & editing, Writing – original draft, Visualization. Li Zhang: Writing – review & editing, Writing – original draft, Visualization, Resources. Wen-Feng Fang: Writing – review & editing, Writing – original draft, Visualization, Validation, Supervision, Resources, Project administration, Funding acquisition, Conceptualization.

Declaration of competing interest

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 paper.