DK210 (EGFR) Induces Immune Response Without Increased CRS or Regulatory T Cells in Solid Tumors

The next-generation, dual immunocytokine Diakine DK2¹⁰ (EGFR) targeting wild-type interleukin (IL)-2 and IL-10 engendered potent, balanced, and targeted immune activation without driving significant systemic toxicity or increases in regulatory T cells (Tregs) in patients with relapsed/refractory solid tumors known to express EGFR, according to findings from an ex vivoassay of immune response biomarkers used during a first-in-human phase 1 trial (NCT05704985).

Findings presented at the 2024 SITC Annual Meeting showed that among the first 31 patients evaluated during the study, DK210 (EGFR) administration stimulated antitumor cytotoxic molecules, including interferon-𝛾 (IFN-𝛾), granzyme B, and perforin in a dose-dependent manner without significantly increasing TNF-alpha (TNF-ɑ) levels. Notably, the magnitude of IFN-𝛾 induction exceeded benchmarks for wild-type IL-2. These data demonstrated that combining IL-10 with wild-type IL-2 uncouples IFN-𝛾 induction from cytokines associated with cytokine release syndrome (CRS), including IL-6, IL-1𝛽, and TNF-ɑ.

Asymptomatic eosinophilia was observed across all cohorts, but no clinical interventions were necessary. Furthermore, treatment with DK210 (EGFR) induced several biomarkers of wild-type IL-2 across all cohorts and increased granzyme B and perforin levels. Notably, low-frequency, low-grade CRS and hypertension were reported, and CRS-associated cytokine levels did not increase. These findings corroborate patient response data from the first-in-human phase 1 trial.

“The immune response biomarker profile in patients in the response assay to DK2¹⁰ (EGFR) and the on-treatment assessment demonstrated that coupling wild-type IL-2 with IL-10 and targeting within the tumor microenvironment results in potent immune activation without inducing cytokines that drive significant systemic toxicity or [a] statistically significant increase in Tregs,” presenting author Abdul R. Naqash, MD, of Stephenson Cancer Center of the University of Oklahoma in Oklahoma City, and colleagues wrote in a poster presentation of the data. “This proof of mechanism supports further clinical evaluation of DK2¹⁰ (EGFR) in renal cell carcinoma (RCC) and non–small cell lung cancer (NSCLC), and [it] validates the Diakine platform.”

Study Rationale, Design, and Methodology

DK2¹⁰ (EGFR) is comprised of wild-type IL-2 coupled to a high-affinity variant of IL-10 that binds EGFR. This coupling is intended to reduce toxicities associated with wild-type IL-2, increase the potency of this molecule, and target it more precisely to the tumor cell surface for improved efficacy.

In the IFN-𝛾 predictive response assay, CD8-positive T cells were model antigen bulk activated with anti-CD3 and anti-CD28 for 3 days before being exposed to DK2¹⁰ (EGFR) for 3 days. INF-𝛾 secretion from T-cells was then induced through the administration of soluble anti-CD3, resulting in weak antigen presentation when activated T cells encounter the tumor. Responders were defined as patients with high levels of secreted INF-𝛾. The genetic differences between high and low responders are accordingly under investigation.

Investigators hypothesized that INF-𝛾 induction through this approach would not be accompanied by significant upregulation of other cytokines indicative of vascular leak syndrome or CRS; rather, they posed that INF-𝛾 induction would result in increased interferon γ–induced protein 10, wild-type IL-2 receptor subunit alpha (IL2Rɑ), IL-18, and IL-18 binding protein levels. Wild-type IL-2 was also expected to induce IL-5, resulting in eosinophilia. Immune system reprogramming would allow peripheral T-cell and natural killer (NK) cell proliferation to be induced in the absence of Treg upregulation, which investigators hypothesized could be measured by an increase in new T-cell clones.

The study enrolled patients with relapsed/refractory solid tumors with known EGFR expression. During dose escalation, DK2¹⁰ (EGFR) was self-administered subcutaneously at 1 of 4 dose levels: 2 mg, 4 mg, 8 mg, or 16 mg. All patients were treated three times a week on days 1, 3, and 5 during 3-week cycles. In the dose-optimization portion of the study, patients received DK2¹⁰ (EGFR) at either 6 mg thrice weekly or twice weekly; 8 mg twice weekly; or 12 mg thrice weekly or twice weekly.

Pharmacodynamic and pharmacokinetic sampling was performed on days 1 through 5 of cycle 1, days 1 and 2 of cycle 2, and concurrent with response evaluation thereafter. Response evaluation by CT scan or MRI was performed every 9 weeks. Fold changes were evaluated at baseline up to the highest change or day 22.

The data cutoff for the current analysis was September 20, 2024. The median age of patients was 65 years (range, 45-79) and most patients were male (52%). All had previously received chemotherapy, and 48% had prior exposure to a checkpoint inhibitor. Over half of patients were White (55%), and the remainder were African American (13%), Asian (6%), Native Hawaiian (3%), or did not have race reported (23%). Tumor types included RCC (36%), colorectal cancer (CRC; 29%), NSCLC (19%), and pancreatic ductal adenocarcinoma (PDAC; 16%). In terms of mutational burden, all 14 patients with CRC/PDAC were microsatellite stable.

Additional Pharmacokinetic and Pharmacodynamic Data

Evidence of wild-type IL-2 signaling was reported at all DK210 (EGFR) dose levels and minimal therapeutic exposure was achieved at dose levels ranging from 4 mg to 8 mg 3 times per week. With a target of area under the curve (AUC) 150 ng/mL per hour, the 2-mg dose cohort achieved an AUC exposure of approximately 145 ng/mL per hour with confirmed 6-month stable disease (SD). Evaluation of on-treatment patient plasma showed a 20-fold increase in INF-𝛾 in cohorts 1, 2, and 3; however, this increase plateaued at the 8-mg dose.

Although DK2¹⁰ (EGFR) was found to induce CD3-positive T-cell and NK cell expansion, it did not increase Tregs among patients with SD. Between day 0 to day 22, granzyme B and perforin were also upregulated in CD8-positive T cells and NK cells, along with Ki-67, which is indicative of proliferation.

Notably, a subset of patients with immune activation also had clonal expansion and enhanced repertoire diversity beginning on day 5. Accordingly, TCR Beta Next Gen Sequencing was performed in this subgroup. Changes in peripheral repertoire correlated with results from precision patient selection assays.

“Further exploration of DK2¹⁰ (EGFR) to optimize monotherapy dose selection is ongoing before proceeding to evaluate clinical activity in expansion cohorts and relevant combinations,” study authors concluded.

Reference

Naqash A, Ayanambakkam A, Spira A, et al. Immune biomarkers from phase 1 first-in-human trial treating advanced cancer patients with DK2. Presented at: 2024 SITC Annual Meeting; November 6-10, 2024; Houston, TX. Abstract 43.