JIN-A02, a Mutant-Selective Fourth-Generation EGFR Inhibitor, Overcomes C797S-Mediated Resistance and Demonstrates Intracranial Activity in NSCLC

Abstract

Purpose:

Epidermal growth factor receptor tyrosine kinase inhibitors (EGFR TKI) have revolutionized the treatment of non–small cell lung cancer (NSCLC) with activating EGFR mutations. However, acquired resistance—particularly the EGFR C797S mutation—remains a major clinical challenge. As no approved targeted therapies are available following disease progression on the third-generation EGFR-TKI osimertinib, this study aimed to evaluate JIN-A02, a novel fourth-generation EGFR-TKI, as a therapeutic strategy to overcome C797S-mediated resistance.

Experimental Design:

JIN-A02, a fourth-generation EGFR TKI, was evaluated for its antitumor efficacy and blood–brain barrier penetration in in vitro and in vivo models of NSCLC harboring EGFR C797S and T790M mutations.

Results:

JIN-A02 demonstrated potent antiproliferative activity in preclinical NSCLC models harboring EGFR C797S and T790M mutations, with superior inhibition of EGFR signaling compared with osimertinib. In both subcutaneous and orthotopic intracranial xenograft models, JIN-A02 elicited substantial tumor regression, indicating robust in vivo efficacy. The agent was well tolerated throughout the treatment period without notable toxicity. In line with preclinical data, early clinical trial data showed signs of efficacy, including three patients showing partial response.

Conclusions:

These findings highlight JIN-A02 as a promising therapeutic strategy to overcome C797S- and T790M-mediated resistance in EGFR-mutant NSCLC, including intracranial disease, and support its further clinical development.

Translational Relevance.

Acquired resistance to third-generation epidermal growth factor receptor tyrosine kinase inhibitors (EGFR-TKI), most commonly driven by the EGFR C797S mutation, represents a major unmet clinical need in patients withEGFR-mutant non–small cell lung cancer (NSCLC). At present, there are no approved targeted therapies that effectively overcome this resistance mechanism. In this study, we demonstrate that JIN-A02, a fourth-generation EGFR TKI, exhibits robust antitumor activity against EGFR C797S mutations across engineered and patient-derived cell lines, as well as in subcutaneous and intracranial xenograft models. In addition, early clinical observations from an ongoing phase I/II study indicate molecular and radiographic responses in patients harboring EGFR C797S. Together, these data provide a biological and pharmacologic rationale for the continued clinical development of JIN-A02 as a therapeutic option for patients with EGFR-mutant NSCLC.

Introduction

Lung cancer is the leading cause of cancer-related mortality worldwide (1). Non–small cell lung cancer (NSCLC), which accounts for approximately 85% of all lung cancer cases, has been associated with poor prognosis, with a 5-year relative survival rate of only 9.7% among patients with distant metastases (2, 3). The epidermal growth factor receptor (EGFR), a critical regulator of cell proliferation and survival, is frequently altered by activating mutations in NSCLC (4, 5). These mutations are identified in 40% to 55% of Asian and 5% to 15% of Western patients (6), with exon 19 deletions and exon 21 L858R substitutions being the most prevalent, accounting for approximately 50% and 40%, respectively, of all EGFR mutations (7).

Phase III clinical trials have shown that first-generation (gefitinib and erlotinib) and second-generation (afatinib) EGFR tyrosine kinase inhibitors (EGFR TKI) significantly improve response rates and progression-free survival (PFS) compared with platinum-based chemotherapy in patients with advanced NSCLC harboring activating EGFR mutations (8). However, acquired resistance develops in the majority of patients following an initial response. Among various resistance mechanisms, the EGFR T790M mutation is the most prevalent, accounting for approximately 50% to 60% of cases (9). To overcome T790M-mediated resistance, osimertinib, a third-generation EGFR TKI, was developed (10, 11). Osimertinib has shown PFS and overall survival benefit over first- or second-generation TKIs in the FLAURA trial (12). Although it induces strong initial responses, acquired resistance almost inevitably emerges during treatment (13, 14). The most prominent on-target resistance mechanism is the EGFR C797S mutation in exon 20, which prevents covalent binding of third-generation EGFR TKIs and confers drug resistance. This mutation is detected in approximately 10% to 26% of patients after second-line therapy and ∼7% following first-line osimertinib (15). Mechanistically, substitution of cysteine (Cys797) with serine in the ATP-binding pocket impairs covalent inhibitor binding, thereby reducing drug efficacy (16).

Currently, no targeted therapies have been approved to effectively overcome resistance mediated by the EGFR C797S mutation. After disease progression on osimertinib, platinum-based cytotoxic chemotherapy remains the standard of care, despite its considerable toxicities, including immunosuppression, myelosuppression, and hepatotoxicity. Moreover, many patients develop compound EGFR C797S mutations, for which no existing treatment can effectively suppress multiple concurrent resistance mechanisms (17). Thus, the development of next-generation EGFR TKIs capable of selectively targeting EGFR C797S and its associated alterations constitutes a critical unmet clinical need in EGFR-mutant NSCLC.

JIN-A02 is a novel, orally available fourth-generation EGFR TKI that selectively and reversibly inhibits resistance-associated EGFR mutations, including C797S and T790M. In this study, we demonstrated its potent, dose-dependent antitumor activity in cell lines and xenograft models harboring these mutations. Notably, JIN-A02 crosses the blood–brain barrier (BBB) and exerts robust efficacy in intracranial tumor models. These findings support JIN-A02 as a promising therapeutic strategy for patients with EGFR TKI–resistant NSCLC driven by C797S or T790M mutations. In this study, we present preclinical and early clinical trial results of JIN-A02.

Materials and Methods

Chemicals and antibodies

Osimertinib was purchased from Selleckchem. JIN-A02 was provided by J INTS BIO.

Antibodies were obtained from the following manufacturers: phospho-EGFR (Tyr1068; no. 2234, RRID: AB_331701), EGFR (no. 4267, RRID: AB_2895042), phospho-ERK1/2 (Thr202/Tyr204; no. 4370, RRID: AB_2315112), ERK1/2 (no. 4696, RRID: AB_390780), phospho-AKT (Ser473; no. 9271, RRID: AB_329825), AKT (no. 9272, RRID: AB_329827), phospho-S6K (no. 4858, RRID: AB_916156), and S6K (no. 2217, RRID: AB_331355) were all purchased from Cell Signaling Technology. β-Actin (no. A3854, RRID: AB_262011) was purchased from Sigma.

Cell culture

PC9 (RRID: CVCL_B260), HCC827 (RRID: CVCL_2063), and H1975 (RRID: CVCL_1511) were purchased from the ATCC, and Ba/F3 cells (ACC300) were purchased from the Leibniz Institute DSMZ (German Collection of Microorganisms and Cell Cultures GmbH). Cells were maintained in RPMI-1640 medium supplemented with 10% FBS and 1% antibiotics in an incubator at 37°C with 5% CO₂. Culture medium was refreshed every 2 days, and cells were passaged once or twice per week depending on the proliferation rate of each cell line. All cell lines were routinely tested and confirmed to be negative for Mycoplasma contamination using the Myco-Read Mycoplasma Detection Kit (Biomax).

Ba/F3 cells expressing EGFR mutations

Ba/F3 cells were transduced with lentiviral particles encoding human wild-type (WT) or mutant EGFR sequences cloned into the pLVX-puro vector (RRID: Addgene_125839). The pLVX-puro vector contains a cytomegalovirus promoter to drive the expression of the transgene and a puromycin resistance gene under the PGK promoter, enabling efficient antibiotic selection. Stable cell lines were established by puromycin selection (1 µg/mL). Transduced Ba/F3 cells were cultured under standard conditions. On the other hand, the EGFR WT Ba/F3 cell lines were cultured in growth medium supplemented with human recombinant EGF (10 ng/mL, R&D Systems) and puromycin (0.5 µg/mL).

Generation of PC9_DC cells via CRISPR/Cas9

The PC9_DC (EGFR E19del/C797S) cell line was generated by introducing the C797S mutation into PC9 cells already carrying the EGFR E19del mutation using CRISPR/Cas9 genome editing. The CRISPR/Cas9 plasmid (no. 632601) was purchased from Clontech Laboratories, Inc., and the single-guide RNA (sgRNA) and single-stranded oligonucleotide (ssODN) repair template were synthesized by Macrogen (RRID: SCR_014454). In this system, Cas9, guided by the sgRNA (5′-GCTGCCTCCTGGACTATGTC-3′), introduced a double-strand break at the target locus, allowing homology-directed repair using the ssODN template (5′-TGTGCCGCCTGCTGGGCATCTGCCTCACCTCCACCGTGCAACTCATCACGCAGCTCATGCCCTTCGGCAGCCTCCTGGACTATGTCCGTGAACACAAAGACAATATTGGCTCCCAGTACC-3′) encoding the C797S mutation. After genome editing, single-cell cloning and stepwise validation were conducted to confirm the successful incorporation of the dual EGFR mutations. The resulting cell line was designated as PC9_DC.

Establishment of patient-derived preclinical models

Patient-derived cell lines (PDC; YU-1097, YU-1182, and YUX-1024) were established from malignant pleural effusions, as previously described (18), and are not associated with RRIDs. PDCs were cultured using standard methods and used within 40 passages. Flow cytometry with EpCAM staining confirmed >99% EpCAM-positive cells, indicating high epithelial tumor cell content. Genomic fidelity to the original tumors was confirmed by Sanger and whole-exome sequencing (WES; Supplementary Table S1).

Cell viability assays

Cells were seeded in 96-well plates at a density of 2 × 10³ cells per well and incubated overnight at 37°C to allow adherence. JIN-A02 and the comparator drug osimertinib were added at the indicated concentrations and incubated for 72 hours. Cell viability was assessed using the CellTiter-Glo Luminescent Cell Viability Assay (Promega), according to the manufacturer’s instructions. Dose–response curves, IC₉₀, and AUC values were determined using GraphPad Prism software (RRID: SCR_002798). All experiments were independently performed at least three times.

Colony formation assay

Cells were seeded in six-well plates and cultured in growth medium. After initial attachment, the medium was replaced every 3 days with fresh medium containing the indicated compounds. After 14 days of drug treatment, colonies were fixed with 4% paraformaldehyde and stained with 0.5% crystal violet for 1 hour at room temperature. Plates were gently washed with distilled water and air-dried before imaging.

Immunoblot analysis

Immunoblotting was performed to assess the phosphorylation status of kinases and the total protein levels of target molecules. Protein lysates were separated by SDS-PAGE and transferred onto nitrocellulose membranes. Membranes were blocked in 5% skim milk in TBS with Tween 20 and incubated with the indicated primary antibodies, followed by horseradish peroxidase (HRP)–conjugated secondary antibodies. β-Actin was used as a loading control. Chemiluminescent signals were detected using the SuperSignal West Pico Chemiluminescent Substrate (Thermo Fisher Scientific) and visualized by exposure to X-ray film in a dark room.

In vivo studies

All animal experiments were conducted in accordance with institutional guidelines and approved by the Institutional Animal Care and Use Committee of Yonsei University College of Medicine. The facility is fully accredited by the Association for Assessment and Accreditation of Laboratory Animal Care International. Mice were housed under specific pathogen-free conditions with no more than five animals per cage, and animal welfare (including food, water, and hygiene) was monitored daily. Female mice were used for all in vivo experiments for consistency across treatment groups.

To establish patient-derived xenograft models, YU-1097 cells (5 × 10⁶ in 100 µL PBS) were subcutaneously implanted into the flanks of 6-week-old female BALB/c nude mice (Orient Bio Inc., RRID: IMSR_RJ:BALB-C-NUDE). Once tumors reached an average volume of 200 mm³, mice were randomized (at least five mice per group) to receive one of the following treatments via oral administration: (i) vehicle control, (ii) osimertinib (25 mg/kg, once daily), (iii) JIN-A02 (10 mg/kg, once daily), or (iv) JIN-A02 (30 mg/kg, once daily).

Tumor size was measured three times weekly using digital calipers and calculated as 0.532 × length × width. Percent tumor volume change was calculated as (V_t − V₀)/V₀ × 100. Tumor growth inhibition (TGI) was calculated using two formulas described by Drilon and colleagues (19):

$$\text{If}\mathrm{\hspace{0.17em}}{\mathrm{V}}_{\mathrm{t}}\mathrm{\hspace{0.17em}}\mathrm{>}\mathrm{\hspace{0.17em}}{\mathrm{V}}_0\mathrm{,}\mathrm{\hspace{0.17em}}\text{TGI}\mathrm{\hspace{0.17em}}\mathrm{=}\mathrm{\hspace{0.17em}}100\mathrm{\hspace{0.17em}}\mathrm{\times }\mathrm{\hspace{0.17em}}\left[1\mathrm{\hspace{0.17em}}\mathrm{‐}\mathrm{\hspace{0.17em}}\mathrm{(}{\mathrm{V}}_{\mathrm{t}}\mathrm{\hspace{0.17em}}\mathrm{‐}\mathrm{\hspace{0.17em}}{\mathrm{V}}_0\mathrm{/}\mathrm{\hspace{0.17em}}{\text{CV}}_{\mathrm{t}}\mathrm{\hspace{0.17em}}\mathrm{‐}\mathrm{\hspace{0.17em}}{\text{CV}}_0\mathrm{)}\right]$$

$$\text{If}\mathrm{\hspace{0.17em}}{\mathrm{V}}_{\mathrm{t}}\mathrm{\hspace{0.17em}}\mathrm{<}\mathrm{\hspace{0.17em}}{\mathrm{V}}_0\mathrm{,}\mathrm{\hspace{0.17em}}\text{TGI}\mathrm{\hspace{0.17em}}\mathrm{=}\mathrm{\hspace{0.17em}}100\mathrm{\hspace{0.17em}}\mathrm{\times }\mathrm{\hspace{0.17em}}\left[2\hspace{0.17em}\mathrm{‐}\mathrm{\hspace{0.17em}}\mathrm{(}{\mathrm{V}}_{\mathrm{t}}\mathrm{\hspace{0.17em}}\hspace{0.17em}\mathrm{‐}\mathrm{\hspace{0.17em}}\hspace{0.17em}{\mathrm{V}}_0\mathrm{)}\right]$$

in which V₀ and V_t represent tumor volumes at baseline and study endpoint in the treatment group and CV₀ and CV_t denote the corresponding volumes in the control group. Tumor measurements were performed by a single investigator who was unaware of the treatment group assignments.

No animals were excluded from the analysis, and all enrolled mice completed the planned treatment and evaluation schedule.

Intracranial tumor model

Intracranial tumor models were established by stereotactic implantation of YU-1097-luc cells or PC9-luc cells into the right frontal lobe of 6-week-old female BALB/c nude mice. Mice were anesthetized and positioned in a stereotactic apparatus. A 0.5-mm burr hole was drilled into the right frontal bone, and a guide screw was inserted. A total of 5 × 10⁵ cancer cells suspended in 5 µL PBS were injected over 5 minutes through the guide screw using a microsyringe at a depth of 2.5 mm. The skin incision was closed using surgical glue.

Tumor growth was monitored using the IVIS Spectrum System (Xenogen, PerkinElmer). Mice were intraperitoneally injected with 150 mg/kg D-luciferin, and imaging was performed 15 minutes after injection using Living Image software (PerkinElmer, RRID: SCR_014247). One week after tumor implantation, mice exhibiting detectable bioluminescent signals were randomized (n = 12 per group) to receive one of the following oral treatments: (i) vehicle control, (ii) osimertinib (25 mg/kg, once daily), or (iii) JIN-A02 (30 mg/kg, once daily).

Immunohistochemistry

Immunohistochemistry (IHC) was performed on 4-µm-thick formalin-fixed, paraffin-embedded tissue sections. Slides were heat-treated and deparaffinized in xylene, followed by rehydration through graded alcohols. Antigen retrieval was carried out in 1 mmol/L EDTA buffer (pH 8) at 125°C for 30 seconds. Slides were then pretreated with a peroxidase-blocking reagent (Dako) for 5 minutes and washed with 50 mmol/L Tris-Cl (pH 7.4). Nonspecific binding was blocked with normal goat serum (Dako), after which slides were incubated for 1 hour with an anti-Ki67 antibody (no. 9027, Cell Signaling Technology, RRID: AB_2636984). Slides were washed with 50 mmol/L Tris-Cl (pH 7.4) and incubated for 30 minutes with SignalStain Boost IHC detection reagent (HRP, rabbit; no. 8114; Cell Signaling Technology, RRID: AB_10544930). After additional washes, immunoperoxidase staining was developed for 5 minutes with 3,3′-diaminobenzidine (DAB; Dako). Slides were counterstained with hematoxylin, dehydrated through graded alcohols and xylene, mounted, and coverslipped. Histoscore (H-score) was calculated by inForm 2.6 Tissue Analysis Software (Akoya Biosciences, RRID: SCR_019155).

Whole exome sequencing

Genomic DNA purity and concentration were measured using the PicoGreen dsDNA assay (Invitrogen) and agarose gel electrophoresis. A genomic fragment library was prepared using the SureSelect v5 Kit (Agilent Technologies) for whole exome sequencing (WES) on the Illumina HiSeq 2500 platform. Sequencing reads were aligned to the human reference genome (hg19) using the Burrows–Wheeler Aligner (RRID: SCR_010910), and postalignment processing was performed with the Genome Analysis Toolkit. Somatic variants were detected with Mutect2 (RRID: SCR_026692) and annotated using Oncotator (RRID: SCR_005183). After sequencing, 171 cancer-associated genes were analyzed.

Statistical analysis

Statistical analyses were performed using GraphPad Prism (version 7). In vitro experiments were independently repeated at least three times with technical triplicates. Data are presented as mean ± SEM for cell-based assays and mean ± SD for animal studies. Group comparisons were made using two-way ANOVA with a Bonferroni post hoc test or Kruskal–Wallis with a Dunn test, as appropriate. P < 0.05 was considered statistically significant. Sample size was determined based on prior experience with similar xenograft models and was sufficient to detect biologically meaningful differences between treatment groups.

Patients

All patient samples were obtained from patients with EGFR-mutant NSCLC treated at Severance Hospital, Yonsei University, either before or after disease progression on EGFR TKI therapy. The study was approved by the Institutional Review Board of Severance Hospital (no. 4-2016-0001), and all patients provided written informed consent. This study was conducted in accordance with the ethical principles of the Declaration of Helsinki (World Medical Association) and the Belmont Report (U.S. Department of Health and Human Services).

Phase I study of JIN-A02

The first-in-human phase I study was conducted to evaluate the safety of JIN-A02, including a dose-escalation phase, to characterize dose-limiting toxicities (DLT) and determine the recommended phase II dose (NCT05394831); preliminary data have been presented elsewhere (20–27). Key eligibility criteria included advanced-stage NSCLC with an activating EGFR mutation and disease progression on at least one prior EGFR inhibitor. A measurable lesion was required for exploratory evaluation of efficacy. Starting from 12.5 mg once daily, JIN-A02 dose levels were escalated after evaluating dose–DLT relationships and toxicity probabilities during the first 21 days of dosing. Blood sampling for ctDNA testing was performed at screening, 4 weeks, and end of treatment (EoT). Intrapatient dose escalation to the next dose level was allowed. The trial was conducted in accordance with the Declaration of Helsinki and Good Clinical Practice guidelines. All applicable regulatory requirements were fulfilled, and the protocol was approved by an Ethics Committee at all participating sites. All participants provided written informed consent before undergoing any study procedures or sharing any data, imaging, and tissue for the study.

Results

JIN-A02 exhibits potent in vitro activity against EGFR C797S–driven resistance models

To evaluate the in vitro efficacy of JIN-A02 against C797S-mediated EGFR resistance, we established a panel of Ba/F3 cell lines harboring clinically relevant EGFR mutations (Table 1; Supplementary Fig. S1A–S1D). These included EGFR E19del/C797S and EGFR L858R/C797S, which reflect resistance mechanisms following first-line osimertinib therapy, as well as EGFR E19del/T790M/C797S and EGFR L858R/T790M/C797S, modeling resistance that arises after second-line osimertinib treatment. Cell viability assays were conducted using these engineered Ba/F3 cells, with osimertinib included as a reference comparator. We used IC₉₀ as it more accurately reflects the inhibition level associated with meaningful antitumor effects. JIN-A02 exhibited potent antiproliferative activity across all tested models, with IC₉₀ values of 117 nmol/L (EGFR E19del/C797S), 231 nmol/L (EGFR L858R/C797S), 192 nmol/L (EGFR E19del/T790M/C797S), and 359 nmol/L (EGFR L858R/T790M/C797S; Supplementary Fig. S1A–S1D). Immunoblot analysis demonstrated that JIN-A02 substantially inhibited EGFR phosphorylation in all C797S-mutant models, even upon long exposure observation, corroborating its target-specific activity (Supplementary Fig. S1E–S1H). Furthermore, JIN-A02 demonstrated antiproliferative activity comparable with that of osimertinib in Ba/F3 cells engineered to express EGFR E19del, EGFR L858R, EGFR E19del/T790M, and EGFR L858R/T790M mutations (Supplementary Fig. S2A).

Table 1.

Cell viability for treatment of JIN-A02 in Ba/F3 cell lines with EGFR mutations and WT.

Compound	IC90 nmol/L
	D	L	DT	LT	DC	LC	DTC	LTC	WTa
Osimertinib	6	7	5	8	398	ND (12.1%b)	ND (64.2%b)	ND (68.2%b)	295
JIN-A02	76	86	56	45	117	231	192	359	ND (17.6%c)

Abbreviations: D, E19del; L, L858R; T, T790M; C, C797S; ND, not detected; WT, wild-type.

^aSupplemented with recombinant human EGF (10 ng/mL).

^bCell viability at 1,000 nmol/L osimertinib.

^cCell viability at 1,000 nmol/L JIN-A02.

JIN-A02 demonstrates potent antitumor activity in EGFR C797S preclinical models, including PDCs

To validate the findings from Ba/F3 models, we evaluated the antitumor efficacy of JIN-A02 in patient-derived tumor cell lines harboring EGFR C797S and in additional models harboring EGFR activating mutations and EGFR T790M (Table 2; Supplementary Fig. S2B and S3A–S3C).

Table 2.

Cell viability for treatment of JIN-A02 in NSCLC and PDCs with EGFR mutations containing C797S.

Compound	IC90 nmol/L
	PC9 (D)	YUX-1024 (L)	H1975 (LT)	PC9_DC (DC)	YU-1097 (DTC)
Osimertinib	ND (19.3%a)	ND (34.5%a)	ND (33.8%a)	ND (63.4%a)	ND (69.2%a)
JIN-A02	229	501	245	275	624

Abbreviations: D, E19del; L, L858R; T, T790M; C, C797S; ND, not detected.

^aCell viability at 1,000 nmol/L osimertinib.

YU-1097 is a PDC established from the pleural effusion of a patient initially harboring an EGFR E19del mutation. Following progression on first-line gefitinib and subsequent osimertinib treatment, the tumor acquired EGFR T790M and C797S mutations, resulting in an EGFR E19del/T790M/C797S genotype. JIN-A02 demonstrated potent antiproliferative activity in this model, with an IC₉₀ of 624 nmol/L, whereas osimertinib failed to achieve 90% growth inhibition at the highest concentration (1,000 nmol/L), precluding IC₉₀ determination, and resulted in a maximum growth inhibition of approximately 74.2% (Supplementary Fig. S3A). PC9_DC cells were generated via CRISPR/Cas9-mediated knock-in of the EGFR C797S mutation into parental PC9 cells harboring EGFR E19del. Similarly, IC₉₀ values for osimertinib could not be determined in PC9 and PC9_DC cells because 90% growth inhibition was not achieved at concentrations up to 1,000 nmol/L, with maximum growth inhibition of approximately 19.3% and 63.4%, respectively. In contrast to osimertinib, JIN-A02 retained potent antiproliferative activity in both cell lines, with IC₉₀ values of 229 nmol/L and 275 nmol/L, respectively (Supplementary Fig. S3B). Moreover, JIN-A02 showed robust IC₉₀ and AUC values in EGFR activating mutation cell lines, including HCC827 (EGFR E19del), PDC YUX-1024 (EGFR L858R), and T790M harboring H1975 (EGFR L858R/T790M; Table 2; Supplementary Fig. S2B; Supplementary Table S2).

To assess the effects of JIN-A02 under long-term exposure, we performed colony formation assay in EGFR C797S–harboring models. In YU-1097 (EGFR E19del/T790M/C797S), JIN-A02 at 10 nmol/L markedly reduced colony number and size and at 30 nmol/L nearly abolished colonies. A similar pattern was observed in PC9_DC (EGFR E19del/C797S), in which JIN-A02 at 10 or 30 nmol/L visibly decreased clonogenic growth. In YU-1182 (EGFR L858R/C797S), although 90% growth inhibition was not achieved in short-term viability assays at concentrations up to 1,000 nmol/L (Supplementary Fig. S3C), prolonged exposure reduced colony number and size with JIN-A02 at 10 or 30 nmol/L. By contrast, colonies persisted with osimertinib at 30 or 100 nmol/L in all EGFR C797S–harboring cell lines. (Supplementary Fig. S3D–S3F).

In addition, immunoblot analysis revealed dose-dependent inhibition of EGFR phosphorylation and downstream signaling phosphorylation by JIN-A02 in all EGFR C797S–harboring models. In contrast, osimertinib failed to suppress EGFR phosphorylation and downstream signaling (Supplementary Fig. S3G–S3I).

Efficacy of JIN-A02 in a triple-mutant EGFR E19del/T790M/C797S xenograft model

To assess the in vivo efficacy of JIN-A02 against EGFR C797S–mediated resistance, we established a xenograft model using YU-1097 cells harboring EGFR E19del/T790M/C797S. Mice were administered vehicle, osimertinib (25 mg/kg), or JIN-A02 (10 or 30 mg/kg) orally once daily. Consistent with in vitro results, JIN-A02 induced robust and durable tumor regression, resulting in TGI rates of 115.9% and 168.2% at 10 and 30 mg/kg, respectively, whereas osimertinib achieved a TGI of 49.3% (Fig. 1A). No significant changes in body weight were observed throughout the treatment period (Fig. 1B). On day 29, waterfall plot analysis revealed that the majority of mice treated with JIN-A02 exhibited tumor shrinkage. By contrast, most animals in the vehicle and osimertinib groups showed progressive disease (PD; Fig. 1C). IHC of xenograft tumors showed intense p-EGFR staining and numerous Ki67-positive nuclei in vehicle-treated samples. Treatment with JIN-A02 at 30 mg/kg led to a significant reduction in the H-scores of phosphor-EGFR and Ki67 in tissue sections, with decreases of 70% and 52%, respectively, compared with the vehicle group (Fig. 1D and E). These results demonstrate that JIN-A02 effectively inhibited EGFR E19del/T790M/C797S activation in vivo, resulting in significant tumor suppression of YU-1097.

Figure 1.

In vivo activity of JIN-A02 in EGFR C797S–mediated osimertinib-resistant models. A, Tumor growth curve of YU-1097 xenografts in response to JIN-A02. Statistical analysis was performed using Kruskal–Wallis with a Dunn post hoc test: ***, P < 0.001 versus vehicle; #, P < 0.05 versus osimertinib; ###, P < 0.001 versus osimertinib; $, P < 0.05 versus JIN-A02 10 mg/kg; ns, not significant. Data represent the means ± standard deviation. B, Body weight of the mice during the experimental period. C, Waterfall plot representing the percentage of tumor volume change in mice after 4 weeks of treatment with the indicated drugs. D, Representative images of IHC staining for p-EGFR and Ki67 of tumor sections. E, H-score for phospho-EGFR and Ki67 of tumor sections. Statistical analysis was performed using one-way ANOVA, with a Tukey honest significant difference post hoc test: ***, P < 0.001 versus vehicle; **, P < 0.01 versus vehicle; *, P < 0.05 versus vehicle. Data represent the means ± standard deviation. p-EGFR, phosphorylated EGFR.

JIN-A02 confers intracranial efficacy in an EGFR C797S brain metastasis model

Brain metastases are observed in approximately 20% to 25% of patients with EGFR-mutant NSCLC at diagnosis and increase in frequency with disease progression. Thus, the ability of EGFR TKIs to penetrate the BBB and achieve sufficient intracranial exposure is critical for improving clinical outcomes. To assess the intracranial efficacy of JIN-A02, we established an orthotopic brain metastasis model by directly injecting YU-1097 cells harboring EGFR E19del/T790M/C797S into the mouse brain. Bioluminescence imaging demonstrated that JIN-A02 (30 mg/kg, orally, once daily) rapidly reduced tumor signals by week 1 and maintained sustained suppression through week 4 (Fig. 2A). Quantitative analysis confirmed significant inhibition of tumor growth by JIN-A02 compared with vehicle and osimertinib (25 mg/kg; Fig. 2B). No significant changes in body weight were observed in any treatment group (Fig. 2C). On day 28, waterfall plot analysis showed tumor progression in most mice receiving vehicle or osimertinib, whereas nearly all JIN-A02–treated mice exhibited tumor regression (Fig. 2D). Histopathologic analysis further confirmed marked reduction of intracranial tumor lesions in the JIN-A02 group, whereas substantial tumor burdens remained in the control groups (Fig. 2E). Furthermore, JIN-A02 showed antitumor activity comparable with osimertinib in the PC9 intracranial model (Supplementary Fig. S4A and S4B).

Figure 2.

Antitumor activity of JIN-A02 in YU-1097 cell–driven intracranial tumor models. A, IVIS imaging of intracranial YU-1097-luc tumors following treatment with osimertinib and JIN-A02. B, Change in tumor burden derived from bioluminescent signal. Statistical analysis was performed using Kruskal–Wallis with a Dunn post hoc test: *, P < 0.05 versus vehicle. Data represent the means ± standard deviation. C, Body weight of the mice during the experimental period. D, Waterfall plot representing the percent change in tumor burden after 4 weeks of the indicated treatments. E, Representative image of hematoxylin and eosin staining of brain slices at the EoT. H&E, hematoxylin and eosin. T, tumor region.

JIN-A02 shows early signs of efficacy in phase I/II trial

As of the data cutoff date (July 22, 2025), 23 patients have been treated with JIN-A02 with a once-daily schedule in part A dose-escalation cohort (Fig. 3). The median age was 64 years (range, 52–78 years) and the proportion of female patients was 56.5% (Supplementary Table S3). JIN-A02 has demonstrated good overall tolerance and a safety profile (Supplementary Table S4). The longest duration of treatment was 651 days in an EGFR E19del/T790M/C797S patient (initial dose: 25 mg and current dose: 150 mg). There were three patients with partial response (PR) and seven patients with stable disease (SD). Three PRs were observed in the 50, 100, and 300 mg cohorts, respectively. In the 300 mg cohort, one patient's target lesion size was reduced by 39.7% at cycle 3 day 1 and cycle 5 day 1 and by 44.9% at cycle 7 day 1 compared with baseline (Fig. 4A). In this patient, the size of brain metastasis also reduced by 25% at cycle 5 day 1 and cycle 7 day 1 (Fig. 4B). Exploratory ctDNA analyses were conducted in a subset of participants as this assessment was incorporated mid-phase. These favorable clinical findings align with the molecular effects of JIN-A02 observed in the ctDNA analysis, which demonstrated complete clearance of C797S and E19del and reduction of T790M after JIN-A02 treatment (Fig. 4C). In the 300 mg cohort, one SD patient exhibited detectable E19del ctDNA at cycle 1 day 1, which became undetectable at EoT. Within the same cohort, two PD patients also demonstrated reductions in EGFR-mutant ctDNA; however, both individuals carried baseline TP53 co-mutations (variant allele frequency of 12.63% and 3.84%, respectively; Supplementary Fig. S5). These observations indicate that although JIN-A02 produced molecular activity against EGFR mutations, disease progression in these patients may have been influenced by non-EGFR co-mutations such as TP53. In the pharmacokinetic analysis, the plasma concentration of JIN-A02 reached a relatively steady state for a certain period after dosing, suggesting drug accumulation and the potential for more stable plasma levels with repeated administration (Fig. 4D). A reduction in brain metastases was first observed at the 100 mg dose. Part of the brain lesions showed complete disappearance in another 300 mg cohort patient. Treatment-related adverse events of grade 3 occurred in two patients (Supplementary Table S4). Gastrointestinal toxicities (e.g., diarrhea, nausea, and vomiting) were the most common adverse events. Skin toxicities were mild and no grade 3 or higher events have been reported. At the data cutoff date, dose escalation is ongoing at 300 mg once daily.

Figure 3.

Swimmer plot showing treatment response in patients receiving JIN-A02 (data cutoff: July 22, 2025). Each horizontal bar represents an individual patient, with the length of the bar indicating treatment duration (days). Colors indicate best overall response [red: PR, blue: SD, gray: PD, and green: to be determined (TBD)]. Arrows indicate patients labeled as ongoing treatment. The X-axis represents treatment duration and the Y-axis lists individual patients along with their EGFR mutation status.

Figure 4.

Activity of JIN-A02 in patients with NSCLC (cohort 6 300 mg). A, Target response after JIN-A02 300 mg treatment in a patient with the EGFR E19del/T790M/C797S mutation. Orange arrow indicates the tumor lesion. B, Intracranial response after JIN-A02 300 mg treatment in a patient with the EGFR T790M mutation. Orange arrow indicates the tumor lesion. C, Changes in EGFR-mutant ctDNA levels before and after JIN-A02 administration in patients with PR. This figure shows the changes in variant allele frequency levels of EGFR mutations from predose to postdose time points. The relative percentage after dose was calculated by setting predose variant allele frequency levels as 100%. D, Plasma concentrations of JIN-A02 from subject 06-001 were analyzed before dose (0) and at 1, 2, 4, 6, 8, 12, and 24 hours after dose following administration of 300 mg JIN-A02 during cycle 1 and cycle 15. C1D1, cycle 1 day 1; C1D15, cycle 1 day 15; C7D1, cycle 7 day 1.

Discussion

On-target resistance mediated by the EGFR C797S mutation remains a major clinical challenge in the management of EGFR-mutant NSCLC (15, 28, 29). This mutation disrupts the covalent binding of third-generation EGFR TKIs, such as osimertinib (30), thereby rendering them ineffective and significantly limiting subsequent therapeutic options. The clinical complexity is further compounded when C797S emerges alongside T790M, particularly in the cis configuration, in which currently available EGFR TKI combinations have shown limited or no clinical benefit (31).

Extensive efforts are underway to develop next-generation EGFR inhibitors capable of overcoming C797S-mediated resistance. The allosteric inhibitor EAI-045 demonstrated antitumor efficacy in EGFR L858R/T790M/C797S mouse models when combined with cetuximab (32), whereas its analogues, JBJ-04-125-02 (33) and JBJ-09-063 (34), exhibited similar mechanisms of action. ATP-competitive inhibitors such as LS-106 (35), CH7233163 (36), and HCD3514 (37) showed high potency and selectivity against triple-mutant EGFR in preclinical models. Alternative approaches, including targeted protein degradation via the PROTAC molecule HJM-561 (38) and the macrocyclic inhibitor BI-4020 (39), have also demonstrated preclinical efficacy. Despite these advances, most agents remain at the preclinical stage. Multiple agents targeting the EGFR C797S mutation are currently under clinical investigation. BLU-945 (NCT04862780) and BBT176 (NCT04820023) have completed early-phase trials (40, 41), whereas WSD0922-FU (NCT06868485) has entered early-phase trials (42) for EGFR C797S–mutant NSCLC. Notably, BH-30643 (NCT06706076), which demonstrated excellent efficacy in preclinical models harboring EGFR C797S and atypical EGFR mutations, including EGFR E20ins, EGFR G719A/S768I, and EGFR G719A/L861Q, has also begun clinical trials (43).

In this study, we demonstrated that JIN-A02, a fourth-generation EGFR TKI, exhibits potent and selective antitumor activity in preclinical models harboring EGFR C797S mutations. JIN-A02 effectively inhibited cell proliferation and suppressed EGFR phosphorylation in both engineered and patient-derived models, including triple-mutant variants containing EGFR E19del/T790M/C797S and EGFR L858R/T790M/C797S. These effects were consistent across in vitro and in vivo models, supporting the target specificity of JIN-A02 (Fig. 1; Supplementary Fig. S1 and S2). In patient-derived and CRISPR-engineered models, JIN-A02 retained strong antiproliferative activity against EGFR C797S–mediated resistance. YU-1097 cells, established from patients with acquired EGFR triple mutations following osimertinib therapy, were highly resistant to osimertinib but remained sensitive to JIN-A02. Likewise, JIN-A02 effectively inhibited growth and downstream signaling in PC9_DC cells harboring the EGFR E19del/C797S mutation (Table 2; Supplementary Fig. S4). These findings validate the efficacy of JIN-A02 across diverse models that represent clinically relevant resistance mechanisms. We report the cellular IC₉₀ values to emphasize that durable control of EGFR-driven tumors requires deep pathway suppression sufficient to eliminate resistant subpopulations.

It is noteworthy that the effect of JIN-A02 in the triple-mutant model was greater than in the double-mutant model in the colony formation assays whereas JIN-A02 exhibited greater effects on cell viability in the double-mutant model. There may be several explanations underlying this pattern. Colony formation assays (using the triple-mutant YU-1097 model) reflect the long-term proliferative capacity of cancer cells whereas cell viability tests (using the double-mutant PC9 model) exhibit short-term cytotoxicity. The PC9 double-mutant model possesses robust clonogenic recovery capacity, meaning that a subset of persister cells survives drug treatment and can eventually regenerate colonies. Indeed, it has been reported that PC9 cells form drug-tolerant persister populations capable of surviving EGFR inhibition, undergoing reversible cell-cycle arrest, and subsequently resuming proliferation to reconstitute colonies (44). By contrast, the triple-mutant YU-1097 model is a patient-derived cell, which may generally have low intrinsic clonogenic potential, reduced DNA repair fidelity, and high cellular heterogeneity (45). These characteristics of the patient-derived cell may have led to the greater effect of JIN-A02 in the triple-mutant model than in the double-mutant model in the long-term colony formation assays.

In xenograft models established with YU-1097 cells harboring EGFR E19del/T790M/C797S mutations, JIN-A02 induced marked and durable tumor regression in a dose-dependent manner. Compared with osimertinib, JIN-A02 achieved greater TGI, with several mice showing complete tumor regression at the 30 mg/kg dose. Importantly, this antitumor effect was achieved without significant changes in body weight, indicating a favorable therapeutic window. Given the clinical relevance of brain metastases in EGFR-mutant NSCLC, we further evaluated the intracranial efficacy of JIN-A02 in an orthotopic brain metastasis model. In this model, JIN-A02 significantly suppressed tumor growth as evidenced by bioluminescence imaging, volumetric analysis, and histopathologic examination. Importantly, JIN-A02 treatment resulted in near-complete tumor regression in most animals and was associated with marked tumor clearance in brain tissues.

In the early analysis of a phase I trial involving patients with EGFR-mutant NSCLC who experienced disease progression on third-generation EGFR TKI, treatment with JIN-A02 demonstrated favorable safety profile with clinically meaningful activity. Overall, these data support the continued development of JIN-A02 among patients previously treated with EGFR TKI.

In summary, this study demonstrated the preclinical efficacy of JIN-A02 against EGFR C797S–mediated resistance, a major clinical obstacle after failure of third-generation EGFR TKIs. Consistent and durable activity across intracranial models highlighted its therapeutic potential. In addition, based on initial analyses from the phase I clinical trial, these findings provide a rationale for continued evaluation of JIN-A02 as a fourth-generation EGFR TKI across diverse clinical settings in treatment-refractory EGFR-mutant NSCLC. A limitation of this study includes the single-arm study design and thus the lack of a standard-of-care comparator arm. However, participants had all failed standard-of-care regimens. In addition, the sample size is small to draw a firm conclusion on the activity and tolerability.

Supplementary Material

Supplementary Figure S1

Cell viability curve for treatment of JIN-A02 in Ba/F3 cell lines with EGFR mutations containing C797S

Supplementary Figure S2

Cell viability curve for treatment of JIN-A02 in Ba/F3 cell lines and tumor cell lines with EGFR activating mutations and containing T790M mutation

Supplementary Figure S3

Cell viability curve for treatment of JIN-A02 in EGFR-C797S mediated osimertinib-resistant cells

Supplementary Figure S4

Antitumor activity of JIN-A02 in PC9 cells-driven intracranial tumor models

Supplementary Figure S5

Changes in EGFR-mutant ctDNA Levels Before and After JIN-A02 Administration in Patients with Stable Disease and Progressive Disease

Supplementary Table S1

The EGFR mutation status of patient-derived models using whole exome sequencing analysis

Supplementary Table S2

Area under the curve quantification of TKI efficacy in C797S-negative EGFR-mutant cell lines.

Supplementary Table S3

Patients Characteristics.

Supplementary Table S4

Summary of Treatment Related AEs

Data Availability

The WES data generated from PDCs in this study contain information that could compromise patient privacy and are subject to institutional and regulatory restrictions but are available from the corresponding author upon reasonable request. Processed data relevant to the results, including mutant allele frequency summaries, are provided in the Supplementary Table S1. Additional data supporting the findings of this study were generated by the authors and are available from the corresponding author upon reasonable request.

Authors’ Disclosures

K.H. Lee reports grants from Merck and personal fees from Bristol Myers Squibb, MSD, Pfizer, AstraZeneca, Eli Lilly and Company, Yuhan, Daiichi Sankyo, Boehringer Ingelheim, Johnson & Johnson/Janssen, and Amgen outside the submitted work. B.Y. Shim reports grants from Yuhan outside the submitted work. M. Nagasaka reports personal fees from AstraZeneca, Daiichi Sankyo, Eli Lilly and Company, Genentech, Regeneron, Boehringer Ingelheim, Caris Life Sciences, Johnson and Johnson, Pfizer, Bristol Myers Squibb/Mirati Therapeutics, and Takeda and personal fees and other support from AnHeart/Nuvation Bio outside the submitted work. E. Seah reports employment with the Sponsor Company. B.C. Cho reports other support from Champions Oncology, Crown Bioscience, Imagen, Yonsei University Health System, and PearlRiver Bio GmbH; grants from CJ Bioscience, Cyrus, ImmuneOncia, Janssen, J INTS BIO, MSD, Yuhan, Dong-A ST, and LigaChem Bioscience; personal fees from Amgen, ArriVent BioPharma, AstraZeneca, Bristol Myers Squibb, Boehringer Ingelheim, CJ Bioscience, Cyrus Therapeutics, Gilead Sciences, GSK, Regeneron, Janssen, MSD, Yuhan, KANAPH Therapeutics Inc., J INTS Bio, Novartis, and Pfizer; and personal fees and other support from DAAN Biotherapeutics outside the submitted work. S.M. Lim reports grants from Amgen, Yuhan, Beigene, Boehringer Ingelheim, BridgeBio Therapeutics, Oscotec, Roche, GSK, Jiangsu Hengrui, AstraZeneca, Lily, Takeda, Daiichi Sankyo, and J Ints Bio. No disclosures were reported by the other authors.

Authors’ Contributions

E.J. Lee: Data curation, software, formal analysis, visualization, writing–review and editing. J.A. Ko: Conceptualization, resources, data curation, software, formal analysis, supervision, funding acquisition, validation, investigation, methodology, writing–original draft, project administration, writing–review and editing. M.-j Kim: Formal analysis. J.S. Cho: Data curation, software, formal analysis, validation, writing–original draft. J.-Y. Han: Resources, formal analysis, methodology. S.W. Kim: Resources, methodology. K.H. Lee: Resources, methodology. B.Y. Shim: Resources, methodology. J.-M. Sun: Resources, methodology. M. Nagasaka: Resources, methodology. S. Park: Formal analysis. S.Y. Oh: Formal analysis. M.H. Hong: Resources, funding acquisition, methodology. J.B. Lee: Resources, funding acquisition, methodology. A. Jo: Resources, methodology, writing–review and editing. E. Seah: Resources, methodology, writing–review and editing. B.C. Cho: Resources, funding acquisition, methodology, writing–review and editing. S.M. Lim: Conceptualization, resources, data curation, supervision, funding acquisition, investigation, methodology, writing–original draft, project administration, writing–review and editing.