Document resource
Background Interleukin-15 (IL-15) is a potent anti-tumour cytokine with strong immunotherapeutic potential, particularly when complexed with IL-15 receptor alpha to form a superagonist (IL-15SA). 1 However, intravenous delivery induces widespread immune activation and dose-limiting toxicities, thereby narrowing its therapeutic window and restricting clinical use to local administration. Directing IL-15SA activity to the tumour microenvironment (TME) is therefore highly desirable to fully harness its therapeutic potential. To overcome this, we have engineered a tumour-targeted IL-15SA fused to collagen-binding domain (CBD) called CBD IL-15SA to facilitate enhanced localisation to collagen-rich extracellular matrix (ECM) of solid tumours.2 3 We hypothesised that this strategy would enhance intratumoural immune cell activation while reducing systemic exposure to improve the therapeutic index and reduce adverse effects associated with intravenous administration of IL-15SA.Methods IL-15SA and CBD IL-15SA were recombinantly expressed in HEK293F cells, purified via metal-immobilised affinity chromatography and size exclusion chromatography. Therapeutic efficacy of intravenously administered CBD IL-15SA and unmodified IL-15SA was evaluated in subcutaneous murine colon carcinoma (CT26, n=5/group) and pancreatic adenocarcinoma (KPC, n=10-11/group) models. Plasma interferon gamma levels (IFNγ) and spleen weight were assessed in tumour-bearing mice (n=5-8/group). Serum biochemistry was evaluated in non-tumour bearing mice (n=2-3/group). Flow cytometry was performed on CT26 tumours to assess immune cell infiltration.Results We found CBD IL-15SA, unlike unmodified IL-15SA, significantly prolonged survival in both CT26- and KPC-bearing mice compared to PBS, with a durable complete remission observed in 20% of CT26-bearing mice treated with CBD IL-15SA ( figure 1). Crucially, plasma IFNγ was undetectable following intravenous administration of CBD IL-15SA in contrast to IL-15SA treatment which induced a significant elevation. Furthermore, unmodified IL-15SA induced a significant increase in splenomegaly (41%, p<0.05) compared to PBS, while CBD IL-15SA caused a minimal increase (18%, p>0.05), indicating reduced systemic immune activation. In naïve mice, serum alanine aminotransferase and glucose levels remained comparable to PBS following CBD IL-15SA, unlike unmodified IL-15SA, which suggests CBD fusion reduces systemic toxicity (figure 2A). Furthermore, CBD IL-15SA treatment resulted in a marked reduction in CT26 tumour volumes with flow cytometry analysis indicating significantly higher numbers of NK, NKT, CD8+ and CD4+ T effector cells in the tumour compared to PBS, which suggests effective TME remodelling (figure 2B).Conclusions CBD fusion to IL-15SA enables potent, tumour-localised immune activation while mitigating systemic toxicity. This strategy may enhance the therapeutic index of IL-15SA for intravenous administration, supporting the development of safer and more effective immunotherapies with improved clinical benefit.Acknowledgements We would like to thank Cancer Research UK’s Convergence Science Centre for funding this research, and the technical staff in the Department of Bioengineering and the Central Biomedical Services at Imperial College London for their support.References Guo Y, Luan L, Patil NK, Sherwood ER. Immunobiology of the IL-15/IL-15Rα complex as an antitumor and antiviral agent. Cytokine Growth Factor Rev. 2017;38:10–21.Ishihara J, Ishihara A, Sasaki K, et al. Targeted antibody and cytokine cancer immunotherapies through collagen affinity. Sci Transl Med. 2019;11(487):eaau3259.Mansurov A, Ishihara J, Hosseinchi P, et al. Collagen-binding IL-12 enhances tumour inflammation and drives the complete remission of established immunologically cold mouse tumours. Nat Biomed Eng. 2020;4(5):531–543.Ethics Approval All the animals were handled in accordance with the 1986 Animal Scientific Procedures Act and under a United Kingdom Government Home Office-approved project license and overseen by ethical committees of Imperial College London.Abstract 858 Figure 1CBD IL-15SA improves survival in CT26 and KPC bearing mice. (A) Kaplan-Meier survival curves of Balb/c mice (female, 8-12 weeks) bearing subcutaneous CT26 (5x105 cells, back skin) treated with PBS (n=5), IL-15SA (n=5) or equimolar CBD IL-15SA (n=5) via intravenous injection on day 6 and 10. Single experiment. (B) Survival curves of C57BL6/J mice (female, 8-12wks) bearing subcutaneous KPC (5x105 cells, back skin) treated with PBS (n=10), IL-15SA (n=11) or equimolar CBD IL-15SA (n=11) via intravenous injection on day 10 and 14. Data from two independent experiments. Mice were culled when humane endpoints were met. Statistical analysis was performed using a log-rank (Mantel-Cox) test using GraphPad prism (v9.5.1)Abstract 858 Figure 2CBD IL-15SA demonstrates improved toxicity profile and induces immune cell infiltration in CT26 tumours. (A) Toxicity profiles of treatments evaluated in tumour-bearing and non-tumour bearing mice after treatment are shown. Plasma IFNy (i) from KPC-bearing mice 24 hours post treatment quantified by ELISA and spleen weights (ii) from CT26-bearing mice sacrificed 24 hours after second dose are shown. For biochemistry analysis, C57BL6/J mice (female, 8-12wks) were treated with PBS (n=3), IL-15SA (n=2) or equimolar CBD IL-15SA (n=3) twice four days apart, 24 hours later mice were sacrificed via cardiac puncture and glucose (iii) and alanine aminotransferase (iv) were quantified via Skylar veterinary biochemistry analyser. (B) Immune cell infiltration study was conducted on mice bearing CT26 tumours treated with PBS (n=5), IL-15SA (n=8) or equimolar CBD IL-15SA (n=8) and sacrificed 24 hours after the second dose. Tumour volumes (i) on day of sacrifice were measured with digital calipers and calculated as Volume (mm³) = length (mm) x width (mm) × depth (mm). Flow cytometry analysis on tumour cell suspension showing the number of CD8+ T cell (i), CD4+ T cell (ii), NK (iii) and NKT cells (iv) per mg of tumour tissue. Data presented as average ± SEM, statistical analyses were performed using one-way ANOVA for toxicity parameters and tumour volume comparison, and a student’s T-test for immune cell infiltration performed using GraphPad prism (p<0.05)