A platform for locoregional T-cell immunotherapy to control HNSCC recurrence following tumor resection

Abstract

Surgical resection of head and neck squamous-cell carcinoma (HNSCC) is associated with high rates of local and distant recurrence, partially mitigated by adjuvant therapy. A pre-existing immune response in the patient's tumor is associated with better outcomes following treatment with conventional therapies, but improved options are needed for patients with poor anti-tumor immunity. We hypothesized that local delivery of tumor antigen-specific T-cells into the resection cavity following surgery would direct T-cells to residual antigens in the margins and draining lymphatics and present a platform for T-cell-targeted immunotherapy. We loaded T-cells into a biomaterial that conformed to the resection cavity and demonstrated that it could release T-cells that retained their functional activity in-vitro, and in a HNSCC model in-vivo. Locally delivered T-cells loaded in a biomaterial were equivalent in control of established tumors to intravenous adoptive T-cell transfer, and resulted in the systemic circulation of tumor antigen-specific T-cells as well as local accumulation in the tumor. We demonstrate that adjuvant therapy with anti-PD1 following surgical resection was ineffective unless combined with local delivery of T-cells. These data demonstrate that local delivery of tumor-specific T-cells is an efficient option to convert tumors that are unresponsive to checkpoint inhibitors to permit tumor cures.

Keywords: T-cell; biomaterial; head and neck cancer; immunotherapy; intratumoral.

Figure 1. Biomaterial loading of T-cells.

(A) Matrigel pellets containing purified CD8 T-cells from naïve C57BL/6 mice were seeded to 24 well Matrigel-coated plates and covered with media. (B) Images show placement and degradation of the pellet over time. (C) Culture supernatant and Matrigel were isolated over time and the number of viable T-cells in each compartment weredetermined. Graphs show (i) number of T-cells in supernatant and pellet over time, (ii) viability of T-cells in the supernatant and pellet over time. (D) Functional response of loaded T-cells. The well was coated with Moc1 cells engineered to express the model antigen SIINFEKL (Moc1-ova), and the Matrigel pellet was loaded with OT1 T-cells specific for SIINFEKL, 2C T cells with irrelevant specificity, or no T-cells (NT). Graph shows confluency of cancer cells over time.

Figure 2. Tumor control following injection of antigen-specific T cell biomaterial.

(A) (i) MOC1-ova tumors were established in immune competent C57BL/6 mice along with anti-CD40L to block implantation-related immune responses, and randomized to receive intravenous (IV) or intratumoral (IT) adoptive T cell transfer to established tumors. Mice received 1 × 10⁶ tumor-specific OT1 T-cells or non-specific 2C T-cells IV in suspension, or in 30 μl of Matrigel into the tumor (IT). Graphs show (ii) tumor growth following IV transfer, or (iii) tumor growth following IT transfer. (B) Overall survival of groups treated as in (A), showing (i) IV treatment groups, and (ii) IT treatment groups. (iii) comparison of overall survival of mice treated with OT1 T cells IV versus IT. Experiments incorporated 6–8 mice per group and the displayed experiment is representative of 3 independent repeats. Abbreviation: NS: not significant. ^* p < 0.05, ^** p < 0.01.

Figure 3. Characterization of circulating T-cells following biomaterial injection.

(A) (i) MOC1-ova tumors were established in immune competent C57BL/6 mice and randomized to receive intravenous (IV) or intratumoral (IT) adoptive T-cell transfer. Mice received 1 × 10⁶ tumor-specific OT1 T-cells or non-specific 2C T-cells IV in suspension, or in 30 μl of Matrigel into the tumor (IT). (ii) Representative flow cytometry plots show whole blood 14 days following adoptive T-cell transfer showing identification of CD3+CD8+CD90.1+ OT1 T cells or CD3+CD8+CD45.2+ 2C T-cells. (B) Quantitative analysis of the number of OT1 T-cells and 2C T-cells in the peripheral blood over time following transfer (i) IV as a suspension or (ii) IT as a biomaterial. Phenotypic analysis of Effector (CD44+CD62L-, Memory (CD44+CD62L+), or Naive (CD44-CD62L+) among gated OT1 T-cells over time following transfer (i) IV as a suspension, or (ii) IT as a biomaterial. (C) Phenotypic analysis of Effector (CD44+CD62L–, Memory (CD44+CD62L+), or Naive (CD44–CD62L+) among gated OT1 T-cells over time following transfer (i) IV as a suspension, or (ii) IT as a biomaterial. Abbreviation: NS: not significant, ^* p < 0.05.

Figure 4. Characterization of the tumor environment following biomaterial injection.

MOC1-ova tumors were established in immune competent C57BL/6 mice and randomized to receive 1 × 10⁶ tumor-specific OT1 T-cells or non-specific 2C T-cells in 30 μl of Matrigel into the tumor. Tumors were harvested 7 days later for analysis of infiltrating immune cells. (A) Representative images following multiplex IHC for infiltrating immune cells in the tumor following injection of 2C or OT1 biomaterial. (B) (i) Identification of congenic CD90.1+CD90.2- OT1 and CD90.2+CD45.1+ 2C in the tumor from infiltrating CD8 T-cells. (ii) Graph of mean and standard deviation of infiltrating CD8 T-cells and the proportion that are host, transferred OT1, or transferred 2C. (C) (i) Representative flow cytometry plots show gating for major myeloid populations in the tumor. (ii) Proportions of Neutrophils, TAM, and monocytes in tumors treated with 2C or OT1 biomaterials Abbreviation: NS: not significant. ^* p < 0.05, ^*** p < 0.001.

Figure 5. T-cell biomaterial control of recurrence following surgical resection.

(A) (i) MOC1-ova tumors were established in immune competent C57BL/6 mice and underwent IT administration of tumor-specific OT1 T-cells or non-specific 2C T-cells in 30 μl of Matrigel into the tumor, or subcomplete resection and administration of the same biomaterials into the resection cavity. Graphs show (ii) average tumor size for all groups (iii) recurrence for resection groups, and (iv) overall survival of all treated mice. (B) (i) Immune competent mice bearing MOC1-ova tumors underwent surgical resection and administration of T cell biomaterials into the resection cavity as per a). Mice were randomized to receive no further treatment or 3 weekly doses of adjuvant anti-PD1 starting 5d following resection. Graphs show average tumor size for (ii) T-cells alone or (iii) T-cells plus adjuvant anti-PD1. Graphs show (iv) recurrence following resection, and (v) overall survival of treated mice. Abbreviation: NS: not significant. ^* p < 0.05, ^** p < 0.01.