First-in-Human clinical trial of a small molecule EBNA1 inhibitor, VK-2019, in patients with Epstein-Barr positive nasopharyngeal cancer, with pharmacokinetic and pharmacodynamic studies

Abstract

Purpose

A first-in-human phase one study was conducted in nasopharyngeal carcinoma (NPC) patients to assess the safety and tolerability of VK-2019, a small molecule selective inhibitor of Epstein-Barr virus Nuclear Antigen 1 (EBNA1).

Patients and Methods

Pharmacokinetic and pharmacodynamic studies, including circulating tumor EBV DNA plasma levels, were performed. Twenty-three patients received VK-2019 orally once daily at doses ranging from 60 to 1800 mg using an accelerated titration design, with cohort expansion at 1800 mg. EBV genome copy number and spatial transcriptomic analyses were conducted on biopsies collected from three patients at baseline and after treatment.

Results

VK-2019 was well tolerated. One patient achieved a partial response. Pharmacokinetic results demonstrated good systemic exposure, with high intersubject variability. Decreases in circulating tumor EBV DNA plasma levels were observed in some patients. VK-2019 reduced EBV genome copy number and viral gene expression in patient tumor samples and induced changes in immune cell markers.

Conclusions

VK-2019 at doses up to 1800 mg daily demonstrated an acceptable safety profile, achieved micromolar plasma concentrations, and showed on-target biological activity in tumors from patients with advanced EBV-positive nasopharyngeal carcinoma.

Translational Relevance

EBV is etiologically associated with several cancers including nasopharyngeal carcinoma. EBNA1 is the EBV-encoded nuclear protein essential for viral genome persistence in EBV+ tumors cells and is a candidate pharmacological target. Here we report the results of a First-In-Human clinical trial of VK-2019, an EBNA1 inhibitor, for the treatment of patients (n=23) with advanced EBV-positive nasopharyngeal carcinoma. Orally administered VK-2019 was well tolerated at doses up to 1800 mg per day. Although only 1 of 23 patients showed a partial clinical response, biopsied material from 3 patients taken before and after treatment revealed a decrease in EBV genome copy number and viral gene expression, along with changes in immune cell markers. With respect to the objectives of the phase one trial, VK-2019 demonstrated an acceptable safety profile, achieved micromolar plasma concentrations, and showed biological activity in patients with advanced EBV-positive nasopharyngeal carcinoma.

Introduction

Nasopharyngeal carcinoma (NPC) was diagnosed in approximately 130,000 people globally in 2020 and the death rate was approximately 80,000. The global number of cases and deaths from NPC are expected to increase to approximately 180,000 and 113,000 respectively by the year 2040 (1). The incidence of NPC is highest in Southern China, where virtually all cases are associated with Epstein- Barr Virus (EBV). To date, there is no EBV-targeted therapy to treat EBV associated NPC.

While early-stage local disease is highly curable with radiation therapy, advanced locoregional disease is fatal in approximately one third of all patients despite intensive chemotherapy and radiation treatments (2). There are no reliable cures for patients with metastatic disease. The standard systemic treatments for metastatic or locoregional recurrent disease is combination chemotherapy with agents such as gemcitabine and cisplatin and immune checkpoint inhibitors. Such combinations typically are associated with response rates of 67% to 78% and median progression free survivals of 8-21 months (3,4).

Evidence strongly implicates EBV as the primary etiologic agent in the development of NPC (5,6). Antibodies to viral antigens precede and predict the development of NPC. Nearly all NPC tumors harbor latent forms of EBV. EBV DNA in NPC tumor cells is monoclonal, indicating that EBV infection likely precedes tumorigenesis (7).

EBV persists in a latent state in all NPC tumors. During latency, EBV expresses a limited set of viral genes that stimulate cellular proliferation and promote survival. Epstein-Barr Nuclear Antigen-1 (EBNA1), is the only viral protein that is consistently expressed in all EBV positive tumors (8). EBNA1 is required for the stable maintenance of the EBV genome in latently infected cells and provides a survival function to host cells (9). EBNA1 localizes to the nucleus, binds to circularized EBV episomal DNA at a specific region known as the latent origin of replication (OriP). In addition to its roles in replication, EBNA1 also binds to a viral region called the Family of Repeats (FR) while simultaneously binding to cellular genomic DNA. Thus, EBNA1 acts as a tether to promote the distribution of viral DNA to each of the daughter cells. EBNA1 also acts as a transcription factor for other latent EBV genes including EBNA2, and LMP1.

Thus, EBNA1 plays a critical role in several processes thought to promote and maintain transformation. The essential function of EBNA1 is further substantiated by siRNA knockdown, CRISPR knock-out and dominant negative mutants of EBNA1 that cause substantial inhibition of cellular growth and cell death in EBV+ cancer cell lines (10–16). For all the aforementioned activities of EBNA1, the site-specific DNA binding function of EBNA1 is required.

Small molecule inhibitors of EBNA1 which bind to a region of EBNA1 that is critical for EBNA1-DNA interaction have been developed. One of these small molecule inhibitors of EBNA1, VK-2019 (2-(1H-indol-6-yl)-3-((4-(((tetrahydro-2H-pyran-4-yl)oxy)methyl)phenyl)ethynyl)benzoic acid), binds to an important pocket of EBNA1 at the protein: DNA interface. In particular, the acidic group of VK-2019 binds to a region that overlaps with a critical phosphate of the DNA backbone. EBNA1 inhibitors thereby interfere with critical EBNA1 functions including recruiting cellular replication machinery to the viral origin of replication and the tethering of the viral DNA to the cellular DNA during mitotic segregation. The DNA binding domain is required for all known functions of EBNA1. VK-2019 selectively significantly inhibits the proliferation of EBV-positive tumors in EBV-dependent xenograft models (17). Such inhibition lowers several EBV-encoded genes. VK-2019 has undergone IND-enabling studies supporting a starting dose level of 60 mg orally once daily. The preclinical toxicology profile in those studies suggests that the modest adverse event profile of VK-2019 will be consistent with a small molecule inhibitor of a viral protein with no known human ortholog. The methodology for pharmacokinetic evaluation of VK-2019 in humans has been reported (18). This is the first report of a phase I trial of VK-2019 in patients with EBV associated NPC.

Patients and Methods

Ethical Considerations

The trial was approved by the Institutional Review Board of all participating sites and adhered to the ethical principles of the Declaration of Helsinki and followed Good Clinical Practice Guidelines (19,20).

Study Design and Overview

This is a report of an open-label clinical trial designed to assess the efficacy and safety of VK-2019 in subjects with EBV-associated NPC (NCT03682055; NCT04925544). The study includes a dose escalation phase followed by a dose expansion phase to evaluate clinical activity, pharmacokinetics (PK), pharmacodynamics (PD), and biomarkers. The primary objective is to characterize the safety profile and determine the recommended phase 2 dose of VK-2019 dosed once daily. In the expansion phase, the primary objective is to characterize the anti- tumor effect in patients with recurrent or metastatic NPC using RECIST response rates (21). Secondary endpoints include characterization of pharmacokinetics, pharmacodynamics, progression-free survival and overall survival.

Participants

Eligible participants included adults with confirmed EBV-positive NPC not amenable to curative treatment. At dose expansion, RECIST-evaluable disease was required. Prior therapies must have been completed at least four weeks before enrollment.

EBV positivity was defined as a high circulating tumor EBV (ct EBV) DNA in plasma (> 4000 genomes per μg plasma DNA) and/or biopsy tissue positive for EBV by in situ hybridization.

Subjects were required to have adequate organ function and performance status (ECOG 0-2) and have recovered from acute adverse events (AEs) from prior treatment. Informed written consent from each subject per institutional review board (IRB) procedures was obtained prior to any investigational interventions.

Intervention

Participants received VK-2019 orally using combinations of 30 mg and 200 mg capsules at dose levels presented in the results section. This trial was divided into three parts: Initial Dose Escalation, and two planned Dose Expansions. Dose escalation in this trial utilized both an accelerated titration, based in part upon a 3B design proposed by Simon et al., coupled with a “rolling six” design, using cumulative AE data for each patient collected during the first 28 days of agent dosing (22,23). After accrual to each dose level, data were reviewed to determine if the protocol defined maximum tolerated dose (MTD) or biologically or clinically active dose (BCAD) had been achieved, at which point a decision to expand a dose level to 16 patients was made. If at least two of these sixteen patients experienced RECIST response without safety concerns, an additional nine patients were to be accrued. Supplement Figure 1 depicts the dose escalation and expansion design.

Safety and Monitoring

Safety assessments including complete blood count, serum chemistry, and toxicity were evaluated at baseline, followed by frequent clinic visits during the first 2 cycles, and then every 4 weeks afterward. Electrocardiograms (ECGs) were performed at specified times before and after VK-2019 administration during the first two cycles on days 1, 14, and 28.

Adverse events were monitored and graded according to National Cancer Institute Common Toxicity Criteria for Adverse Events (CTCAE) version 5.0 by local investigators at each clinical site (24). An independent safety and monitoring board of the Stanford University Research Compliance Office reviewed trial conduct and data.

A dose-limiting toxicity (DLT) was defined as a drug-related adverse event or abnormal laboratory result occurring during the first cycle of treatment that met specific criteria. These included Grade 4 neutropenia lasting longer than seven days, Grade 4 thrombocytopenia, or Grade 3 thrombocytopenia associated with bleeding or requiring platelet transfusion, Grade 3 febrile neutropenia, and Grade 4 anemia or anemia requiring blood transfusion. Additionally, DLT encompassed any Grade 3 or higher non-hematological toxicity, with several exceptions: toxicities related to disease progression or intercurrent illness, Grade 3 AST/ALT levels that return to Grade 1 or lower within one week, asymptomatic laboratory abnormalities, Grade 3 fever without neutropenia that returns to Grade 1 or lower within three days and is not associated with hemodynamic compromise, Grade 3 diarrhea, nausea, vomiting, or abdominal pain that returns to Grade 1 within three days or to baseline with medical intervention, Grade 3 fatigue that returns to Grade 1 or baseline within seven days, and Grade 3 tumor flare (defined as pain, irritation, or rash localizing to sites of known or suspected tumor) that returns to Grade 1 or baseline within seven days.

DLT also included laboratory findings indicative of acute liver injury, including doubling of AST or ALT levels to no less than four times the upper limit of normal (ULN) for more than one week, or signs of increasing functional liver impairment with more than a two-fold increase in total bilirubin from baseline. Increased AST or ALT levels above eight times ULN, suspected acute liver injury per Hy’s law, any adverse event resulting in a subject missing less than seven days of dosing with VK-2019, and any adverse event not otherwise meeting the criteria or timing of a DLT but declared a DLT by the Data and Safety Monitoring Committee (DSMC) were also considered DLTs (25).

Clinical Pharmacology and ct EBV DNA Methods

Plasma samples for PK were collected on Cycle 1 Day 0 at predose, 30 minutes, 1 hour, 2 hours, 3 hours, 4 hours, 6 hours, 8 hours and 24 hours post-dose. To measure potential drug accumulation, plasma samples were also collected on Cycle 1 Day 14 at the aforementioned timepoints. Samples were processed within 30 minutes of collection, and all specimens were stored at −70°C until further analysis. PK was analyzed using 100 uL of plasma collected in a K₂EDTA tube by a validated LC-MS/MS method with positive ESI-MRM mode that had a linear range of 0.5 to 500 ng/mL by the Analytical Laboratory at Johns-Hopkins University (RRID:SCR_003391) (18). If samples were above the range, dilutions of 1:10 were accurately quantified. Deuterated VK-2019 (VK-2019-d6) was used as an internal standard. Pharmacokinetic parameters for VK-2019 were estimated through noncompartmental analysis using Phoenix WinNonlin version 8.3 (RRID:SCR_024504). Plasma ct EBV DNA, as the primary biomarker, was measured at baseline, serially during treatment, and at end of treatment to assess changes in ct EBV DNA load in response to VK-2019 treatment (26).

Biopsy Analysis

Patients provided additional informed consent for the optional collection of tumor biopsies. Tumor samples were divided into three sections for EBER-ISH staining, for gene expression and EBV genome copy number (in RNA later) and for spatial transcriptomic analysis (in 10% formalin). The formalin-fixed samples were embedded into one paraffin block and cut into 10 μm sections and mounted on adhesive glass slides by the Wistar Histotechnology Core Facility (RRID:SCR_010206). Each slide contained paired baseline and treatment NPC sections from each of the 3 patients (6 samples in total).

The GeoMx Digital Spatial Profiler instrument (Bruker Spatial Biology, Inc., RRID:SCR_021660) and the associated commercial software (RRID:SCR_023424) were utilized to analyze the NPC samples in the Wistar Genomics Core Facility (RRID:SCR_010205). NPC sections were stained with DAPI and a cocktail of fluorophore-tagged human antibodies against PanCK, CD3, and CD19/20, each conjugated with a synthetic oligonucleotide via a UV-cleavable cross-linker. The stained slides were scanned using the GeoMx instrument, generating three-channel immunofluorescent (IF) digital images to visualize nuclei (DAPI), epithelial cells with PanCK (Novus Cat# NBP2-33200AF488, RRID:AB_3284601), T-cells with CD3 (Novus Cat# NBP2-54392AF594, RRID:AB_3334641), and B-cells with CD19 (Abcam Cat# ab196515, RRID:AB_3663030) and CD20 (Abcam Cat# ab198943, RRID:AB_2905499). Additionally, EBER-ISH staining on a neighboring section cut from the same FFPE block was overlaid on top of the IcF images to aid the selection of 50 circular regions of interest (ROIs). The size of the ROIs was identical to ensure that similar numbers of cells would be collected. These ROIs were then serially illuminated with UV light to release the photocleavable RNA to create libraries for next-generation sequencing (NGS) analysis. Paired baseline and on-treatment samples available from three patients at 1800 mg daily dose level were analyzed using NanoString GeoMx Digital Spatial Profiling (DSP) technology to map gene expression changes after treatment. DSP is a platform for high-plex spatial transcriptional profiling using formalin-fixed, paraffin-embedded (FFPE) tissue samples (27). The three patients were: patient 10 (tissue from baseline about 2.5 years before enrollment and on-treatment cycle 2), patient 13 (tissue from baseline, on-treatment cycle 5, on-treatment cycle 15), and patient 14 (tissue from baseline, on-treatment cycle 2 and at end of treatment, post cycle 4).

EBER-ISH

Tumor sections were fixed in neutral buffered formalin solution at room temperature before paraffin embedding. The EBER-ISH assay was performed using the EBER PNA Probe/Fluorescein and the One-Day In Situ Hybridization Procedure for RNA Detection Using PNA Probes, per manufacturer’s instruction (Dako, Denmark).

Digital Droplet PCR (ddPCR)

DNA was extracted from snap frozen tumor biopsies using the Qiagen DNeasy blood and tissue kit (Qiagen, Hilden, Germany) according to the manufacturer’s instructions. DNA concentrations were calculated using NanoDrop 2000 Spectrophotometer (ThermoFisher, RRID:SCR_018042) prior to ddPCR (Bio-rad QX200 Droplet Digital PCR (ddPCR) System, RRID:SCR_019707) using the following primer probe sets: EBV-BamHI W primers and FAM/MGB probe: (Forward: CTTCTCAGTCCAGCGCGTTT; Reverse: CAGTGGTCCCCCTCCCTAGA; Probe: 6FAM CGTAAGCCAGACAGCAGCCAATTGTCAG MGBNFQ), EBV-LMP1 primers and FAM/MGB probe (Forward: AAGGTCAAAGAACAAGGCCAAG, Reverse: GCATCGGAGTCGGTGG ; Probe:FAM- AGCGTGTCCCCGTGGAGG; were designed for ddPCR using NCBI Primer Blast and Primer3Plus. For the duplex ddPCR, the housekeeping gene, Ribonuclease P protein subunit 30 (RPP30) primers and VIC/MGB probe (Forward: GATTTGGACCTGCGAGCG, Reverse: GCGGCTGTCTCCACAAGT; Probe: VIC-CTGACCTGAAGGCTCT) was used to determine the cellular quantities.(28)

Data Analysis of tumor biopsies

NanoString Data Normalization: Oligonucleotides in each ROI were quantified and processed using NanoString’s bioinformatics pipeline, which normalized the data for subsequent bioinformatic analyses.

Statistical Analysis: All analyses and plots were conducted using the GeoMx software and associated statistical packages (RRID:SCR_023424) and GraphPad Prism (RRID:SCR_002798). Differences in continuous variables between groups were tested using nonparametric Mann-Whitney tests, with significance set at a two-tailed p-value of 0.05 unless otherwise specified.

Differential Expression and Pathway Analyses: Differential expression analysis and pathway analysis was conducted using the GeoMx software package (RRID:SCR_023424). Genes with a log fold change greater than ±1 and a p-value < 0.05 were considered differentially expressed.

Statistical Analysis

For the dose expansion cohort, a sample size of 16 evaluable patients was planned, with the intention of proceeding to a second stage if greater than one RECIST partial response or better was observed, with a plan for a total of 25 patients to be treated in this expansion cohort to allow the null hypothesis response rate of 0.1 to be rejected if greater than 5 responses occurred in these 25 patients.

Data Availability

All versions of the clinical trial protocol and data collected during the conduct of this study are stored in the Stanford University eProtocol database and Stanford Cancer Institute OnCore clinical trials database respectively. Protocols and anonymized data summaries are available upon request.

Results

Between April 4, 2019, and October 2, 2023, 23 patients were accrued at 4 clinical sites to six dose levels (Figure 1). See Supplement Tables 1 and 2 for a summary of patient characteristics. The accelerated titration design allowed dose escalation to proceed from 60 mg/day to 1800 mg/day, with one patient added at 120 mg/day to replace a patient who was non-compliant, and one patient added at 460 mg/day to replace a patient because of highly variable pharmacokinetics. A maximally tolerated dose (MTD) was not determined. (Table 1). All VK-2019-related adverse events (AEs) were manageable and reversible. Eleven serious adverse events (SAEs) occurred. Nine were unrelated to study drug. A grade 4 allergic reaction of less than 24 h duration occurred in one patient at 1800 mg/day and was assessed as related to study drug. One hypokalemia episode was assessed as probably related to study drug. The most common VK-2019 attributed AEs were grade 1 diarrhea (n=11), grade 1 nausea (n=4), and grade 2 rash (n=3) (Table 2). The decision to expand accrual at 1800 mg daily was made based upon preliminary PK evidence (Figure 2) suggesting target exposures associated with clinical activity in preclinical models had been met at 920 mg daily and above, and the impracticality of ingesting more than nine pills at a single daily dose.

Figure 1.

Time on treatment and Treatment response. Individual patient duration on treatment and time of best RECIST response: Partial Response (PR), Stable disease (SD), Progressive Disease (PD), Not Evaluated (NE). Black dot denotes the time that the response was recorded.

Table 1.

Dose levels treatment responses and adverse events

Dose VK-2019, once daily dosing, mg	Number of patients enrolled	Best responses	Adverse events attributed to VK-2019 (grade)
60	1	SD	None
120	2*	PD (2)	None
230	1	PD	Dyspepsia (1)
460	2*	PD (2)	Fatigue (1), Urticaria (1)
920	1**	SD	Nausea (1), Dizziness (1), AST increased (1), ALT increased (2), ALP increased (1), Blood bilirubin increased (1)
1800	16	PR (1) SD (5) PD (9) NE (1)***	See table 2

SD-Stable Disease, PR-Partial Response, PD-Progressive Disease, NE-Not Evaluated

^*One patient was added at 120 mg dose level to replace a patient who was non-compliant, and one patient added at 460 mg dose level to replace a patient who had highly variable drug levels.

^**Patient 18 completed 3 cycles at the 920 mg dose level. Subsequent cycles were at the 1800 mg dose level.

^***Patient 20 did not have treatment response assessed, as they withdrew consent prior to first response assessment.

Table 2.

Summary of treatment-related adverse events in 16 patients according to CTCAE Version 5.0 (24) at the expansion level of 1800 mg daily, N=16 (%).

Adverse Event	G1	G2	G3	G4	G5
Abdominal cramping	1 (6)
Abdominal pain		1 (6)	1 (6)
ANC decreased	2 (12)
Alkaline phosphatase increased	2 (12)	1 (6)
Allergic reaction				1 (6)
Anemia	2 (12)	1 (6)
Bilirubin increased	1 (6)	1 (6)
Body aches	1 (6)
Chills	1 (6)
Diarrhea	11 (69)	2 (12)	1 (6)
Dry mouth	1 (6)
Dyspepsia	1 (6)
Fatigue	2 (12)
Fever	1 (6)	2 (12)
Flu-like Symptoms	1 (6)
Hypoalbuminemia	1 (6)
Hypocalcemia	1 (6)
Hypokalemia	2 (12)		1 (6)
Hyponatremia	1 (6)	1 (6)
Itchy eyes	1 (6)
Jaw pain	1 (6)
Lip infection	1 (6)
Malaise	1 (6)
Nausea	4 (24)*
Oral mucositis	1 (6)
Platelet count decreased	1 (6)
Pruritus		1 (6)
Rash	1 (6)	2 (12)	1 (6)
Rash maculo-papular		1 (6)
Vomiting	2* (12)
White blood cell decreased	1 (6)

^*One instance of nausea and one instance of vomiting occurred in patient who had intrapatient dose escalation from 920 to 1800 mg/day in cycle 4.

Figure 2.

Plasma concentrations following single (cycle 1 day 0) and multiple dose (cycle 1 day 14) of VK-2019. A. Mean concentrations for each dose cohort after single dose. B. Mean concentrations for each dose cohort after multiple doses. Error bars indicate standard deviation. C. Plot of Cmax after single dose and multiple doses at different dose levels. D. Plot of AUC0-24h after single dose and multiple doses at different dose levels. Solid circles are from the single dose (Cycle 1 Day 1) and open circles are from multiple dose (Cycle 1 Day 14) PK data.

Sixteen patients were treated at the 1800 mg daily dose. At this dose one patient experienced two DLTs, a grade 4 allergic reaction that resolved within a day, and grade 3 abdominal pain that resolved within 4 days. Another patient with massive hepatic involvement with NPC experienced grade 3 bilirubin increase in the first cycle and had VK-2019 held. Initial attribution as possibly related to VK-2019 defined a DLT. However, subsequent CT scan done 4 days after the increase in bilirubin was noted demonstrated substantial tumor progression, so attribution was changed to cancer progression. The median number of cycles of VK 2019 received at 1800 mg was 2. Patient 13 achieved a RECIST PR at cycle 6 and received a total of 14 cycles before progressing (Figure 1). This patient exhibited a reduction in target tumor size exceeding 79% (Figure 3A), but no other patients experienced tumor regression (Supplement Figure 2C). Biopsies taken from the patients 10, 13, and 14 at baseline and during treatment revealed a decrease in viral genome copy numbers per cell over time, as indicated by qPCR analysis with two different primer sets, BamHI and LMP1 (Figure 3B). In addition, EBER-ISH showed a decrease in the number of EBV-positive tumor cells (Figure 3C). Because the prespecified response rate of interest of at least two patients in the first stage of expansion was not met, accrual was stopped.

Plasma concentrations obtained from 23 patients on C1D0 and 22 on C1D14 were analyzed in a non-compartmental pharmacokinetic analysis based on the actual dose administered at sampling. The results are presented in Supplement Table 3 and Figure 2. Although the number of patients in each dosing group does not allow for a statistical comparison, the increase in C_max and AUC _0-24h appears to be linear with increase in the administered dose (C_max=39 and AUC_0-24= 202.7 for 920mg vs C_max=82.5 and AUC_0-24= 340.2 for 1800mg) and no accumulation was observed with repeated dosing. The T_max is reported to be around 4 hours with a half-life around 8 hours for most patients. Changes in plasma EBV DNA during VK-2019 treatment are shown in Figure 3D. Among the 23 patients studied, 21 had serial sampling of circulating tumor EBV DNA done at the institution’s reference labs. Two patients had only baseline EBV DNA testing done, without further on-treatment testing, due to being taken off study treatment early. Thirteen patients showed either no change or an increase in EBV plasma levels from the start to the end of treatment and eight patients exhibited a decrease in EBV plasma levels (Figure 3D).

Figure 3.

Response to treatment of select patients and biomarker analysis. A. Tumor volume of patient 13 at 1800 mg daily dose who experienced RECIST PR during treatment. B. Change in EBV genome copy number per cell from biopsies taken from patients 10, 13 and 14. C. EBER-ISH of tumor material taken from patient 14. D. Changes in plasma EBV DNA of patients at the 1800 mg/day dose level. Baseline level = 1.

Analysis of gene expression changes between baseline and post-treatment samples revealed distinct patterns of expression for various viral and immune-related genes (Fig. 4A). Viral genes such as EBER-1 showed lower expression after VK-2019 treatment compared with baseline. Additionally, several genes involved in immune cell regulation exhibited lower expression after treatment relative to baseline (Fig. 4B). These genes included NF-κB complex component (NFKB1A), lymphocyte antigen 6 complex, locus E (LY6E), C-X-C motif chemokine ligand (CXCL) 10 (CXCL10, also known as interferon-γ inducible protein 10, IP-10), IFNα-inducible protein 27 (IFI27), and the HLA-DR antigen–associated invariant chain (CD74) (Fig. 4C). We observed an increased expression of the neutrophil-attracting chemokine CXCL8 (IL8) following treatment compared with baseline.

Figure 4.

A. Digital spatial profiling analysis of gene expression of biopsies before treatment (baseline) and after treatment (treatment) from three patients. A. Heatmap of differentially expressed genes. B. Volcano plot depictions of the log2(fold-change) in gene expression on the x-axis and negative log10(p value) on the y-axis for gene expression. C. Box and whisker plots displaying gene expression counts from three patients. Statistical testing was performed using Mann-Whitney non-parametric T-tests.

Discussion

The safety profile of VK-2019 seen in this first-in-human trial is likely due to the unique nature of its target, EBNA1, which has no orthologs in the human proteome. The targeted action of VK-2019 ensures that it disrupts EBV-related processes without disruptions of human cellular functions. The low level of serious adverse reactions supports the hypothesis that VK-2019’s safety profile is a result of its high specificity for EBNA1. The mild nature of the most common AEs, such as diarrhea, nausea, and rash, also underscores the favorable tolerability of VK-2019.

VK-2019 demonstrated good oral bioavailability, with observed micromolar concentrations of the drug in plasma. However, there was moderate to high variability, particularly at the highest dosage of 1800 mg/day. This variability may have contributed to only one tumor response observed in this trial. Optimizing the drug formulation and administration schedule might reduce pharmacokinetic variability, ensuring more consistent systemic exposure and potentially leading to more robust tumor responses.

In the patient who achieved a PR at 1800 mg daily, there was a modest decrease in circulating EBV DNA from serially amples blood that was concomitant with a decrease in tumor volume. The decrease in circulating EBV seen in this patient was modest, bringing into question whether or not levels of circulating EBV DNA represent a valid biomarker in this situation. Our hypothesis that tumor EBV copy number and circulating EBV DNA levels may serve as therapeutic biomarkers for VK- 2019 will need to be validated with additional study of patients dosed with higher levels of VK-2019 concurrent with serial tumor and blood evaluation of EBV DNA.

The high baseline expression of genes associated with immune cell regulation, antigen presentation, and antiviral restriction, including NFKB1A, lymphocyte antigen 6E (LY6E), CXCL10, IFI27, MX1, and CD74 (HLA-DR antigen associated invariant chain) seems to have been modulated by VK-2019 treatment. Moreover, the direction of gene expression changes observed during VK-2019 treatment is consistent with perturbation of EBV infection and improved cancer outcomes. CD74, for example, is downregulated by EBV-BZLF-1 during lytic activation and upregulated in EBV-associated gastric cancers (29). CXCL family gene expression, also affected by VK-2019 treatment, has been shown to have prognostic value in EBV-associated and other cancers, with CXCL10 expression markedly elevated in EBVaGC tumors (30). Moreover, the observed increase in CXCL8 expression after treatment is particularly noteworthy; CXCL8 plays a key role in neutrophil recruitment and increased CXCL8 expression is associated with better overall survival in patients with EBVaGC (30). In contrast, overexpression of LY6E, which broadly functions as an antiviral restriction factor, is associated with increased cancer cell proliferation, metastasis, and poor survival outcomes for several malignancies (31). Overall, the gene expression changes observed in VK-2019 treated tumors suggest that VK-2019 mediated disruption of EBV genome maintenance is marked by decreased EBER expression and reduced EBV copy number in treated tumors that corresponds with a concomitant decrease in the host antiviral response, suggesting that the overall effect of VK-2019 may be both antiviral, and anti-tumorigenic.

It is important to acknowledge that these biological results are derived from only three matched pairs of biopsies. While the findings are promising, they represent a limited sample size and must be interpreted with caution. Further studies with larger cohorts are necessary to validate these observations. We will be accruing additional patients to twice daily cohorts with further dose escalation in order to attempt to raise the highly variable trough levels seen because we hypothesize that constant exposure to VK-2019 and resultant uninterrupted abrogation of EBNA1 DNA binding will be essential to optimize the anti-cancer effect for patients with EBV associated NPC and to observe persistent downregulation of tumoral and circulating EBV DNA.