SLC-3010: another CD122-biased IL-2 complex shelved in solid tumors

What was tried

SLC-3010 is human interleukin-2 non-covalently complexed with the anti-IL-2 antibody TCB2, a construct written as hIL-2/TCB2c. The antibody binds the central part of the CD25 (IL-2Ralpha) interface on IL-2, which blocks high-affinity CD25 engagement and biases signaling toward CD122 (IL-2Rbeta) high CD8 T cells and natural killer cells over regulatory T cells (Selecxine preclinical reports, 2024). NCT05525247 was a Phase 1/2, open-label, multicenter, non-randomized, sequential dose-escalation and expansion study of SLC-3010 as monotherapy and in combination with gemcitabine in advanced solid tumors (ClinicalTrials.gov NCT05525247). Selecxine dosed SLC-3010 by intravenous infusion on day 1 of each 21-day cycle, with gemcitabine at 1000 milligrams per square meter on days 1 and 8. The study enrolled 27 participants and was terminated with the posted reason "business decision." No results were posted.

The biological hypothesis

High-dose IL-2 produces durable remissions in a minority of patients with melanoma and renal cell carcinoma, which proves that the pathway can drive antitumor immunity. Its use is limited by vascular leak and by preferential expansion of CD25-high regulatory T cells. The engineering thesis behind not-alpha or CD122-biased IL-2 is to strip away CD25 binding, preserve CD122 and CD132 signaling, and thereby expand effector CD8 and NK cells while sparing Tregs (Raeber and colleagues, 2022; Hashimoto and colleagues, 2022). SLC-3010 pursues that thesis through antibody complexation rather than mutation or pegylation. Open Targets returns no direct association score for IL2 against the generic neoplasm term, so a default of 0.5 was applied for the mechanism risk calculation, with the caveat that IL-2 pathway validity in cancer rests on the approved activity of aldesleukin rather than on target-disease genetics.

What actually happened

The trial stopped at 27 enrolled participants for a business decision, with no efficacy, safety, or pharmacodynamic results posted to the registry (ClinicalTrials.gov NCT05525247). Absent a public readout, the direct evidence is limited to the fact that a small biotech ended a first-in-human program before completing expansion. The relevant context is the class. Bempegaldesleukin, the most advanced CD122-biased IL-2, failed across the PIVOT-IO 001 melanoma program and related studies despite strong preclinical and early-phase pharmacodynamic data, and its development was halted in 2022. Several other engineered IL-2 assets have been discontinued or deprioritized over the same window (Raeber and colleagues, 2022, review of IL-2-based immunotherapies). SLC-3010 enters that record as another CD122-biased construct shelved before a registrational signal emerged.

Failure mechanism, best guess

For this specific trial the proximate cause was a sponsor-level business decision by a small company, which is an operational stop rather than a demonstrated mechanistic failure. The class-level pattern is the more informative signal, and it points to translational mismatch. CD122 bias increases effector expansion in preclinical models, but the same bias can drive systemic CD8 and NK proliferation without producing productive intratumoral cytotoxicity, and it can still trigger the eosinophilia and vascular effects that IL-2 signaling causes. Peripheral pharmacodynamic wins have repeatedly failed to predict tumor control for this class. The plausible mechanism is that antigen-independent effector expansion in blood does not translate into antigen-specific killing inside immune-excluded tumors, so the pharmacodynamic biomarker that these programs optimize is disconnected from the efficacy endpoint that matters.

How to prevent this next time

Two quantitative tools would have de-risked the program before expansion. First, a translational validation sequence anchored on intratumoral rather than peripheral readouts, with a prespecified requirement that paired biopsies show a defined increase in intratumoral CD8 density before expansion proceeds. Second, a Bayesian go-no-go framework tied to that biomarker and to early response, evaluating the predictive probability of success at each interim:

PP(success)=∫θ​P(final success∣θ,interim data)⋅π(θ∣interim data)dθ

With a beta prior centered on the class base rate and a go threshold set to a clinically meaningful response floor, the predictive probability would remain low until intratumoral activity was demonstrated, which prevents committing to a large expansion on peripheral pharmacodynamics alone. Historical base rates justify the discipline, since oncology assets show roughly a 3.4 percent probability of success from Phase 1 to approval against about 13.8 percent across all indications (Wong, Siah, and Lo, 2019), and the engineered IL-2 class has underperformed even that low bar. The single highest leverage change would have been ...

What this means for similar programs

The IL-2 field keeps producing constructs that separate effector expansion from regulatory expansion in mice, then fail to reproduce tumor control in patients. SLC-3010 does not add efficacy data, but its quiet termination is consistent with the class. Groups pursuing masked, cis-targeted, or receptor-biased IL-2 should treat peripheral CD8 and NK expansion as a necessary but not sufficient readout, require intratumoral evidence before expansion, and size early cohorts to detect response rather than pharmacodynamics. The approved activity of aldesleukin shows the pathway can work, so the engineering question is delivery and localization, not whether IL-2 signaling matters.

Open questions

Did SLC-3010 produce the intended CD122-biased pharmacodynamic profile in the dosed patients, and was any of it posted at a conference. Would the gemcitabine combination have changed the tumor microenvironment enough to matter. Is antibody complexation more or less stable in circulation than mutein or pegylation approaches to CD25 bias. Does any engineered IL-2 subclass show intratumoral CD8 expansion that tracks with response, which would define the enrichable population the class has lacked.

Sources

1. • ClinicalTrials.gov, NCT05525247, full record, terminated status, whyStopped "business decision," enrollment 27, no posted results: https://clinicaltrials.gov/study/NCT05525247 - Selecxine mechanism of SLC-3010 as hIL-2 complexed to anti-IL-2 antibody TCB2 blocking the CD25 interface, biasing to CD122-high CD8 and NK cells, summarized in a systematic review of IL-2-based immunotherapies, eBioMedicine, 2023: https://www.thelancet.com/journals/ebiom/article/PIIS2352-3964(23)00104-4/fulltext - Open Targets Platform, IL2 (ENSG00000109471), no direct association for the generic neoplasm term, default 0.5 applied: https://platform.opentargets.org/target/ENSG00000109471 - Engineering IL-2 for immunotherapy of autoimmunity and cancer. Nature Reviews Immunology.

2. - PD-1-cis IL-2R agonism yields better effectors from stem-like CD8 T cells. Nature.

3. - IL-2Ralpha-biased agonist enhances antitumor immunity. Nature Cancer.

4. - Wong CH, Siah KW, Lo AW. Estimation of clinical trial success rates and related parameters. Biostatistics. 2019;20(2):273-286.