2770 - Hypofractionated Radiation and T Cells in Oral Cancer on a Phase I Clinical Trial of Preoperative Radiation

M. J. Awan¹, J. Zenga², A. Frei¹, J. Foeckler³, A. Miller³, T. Kearl³, V. Jin³, S. Wong⁴, and H. A. Himburg¹; ¹Department of Radiation Oncology, Medical College of Wisconsin, Milwaukee, WI, ²Department of Surgery, Medical College of Wisconsin, Milwaukee, WI, ³Medical College of Wisconsin, Milwaukee, WI, ⁴Department of Medical Oncology, Medical College of Wisconsin, Milwaukee, WI

Purpose/Objective(s): To determine the immunologic effects of hypofractionated radiation (RT) pre-operatively on patients with oral squamous cell carcinomas (OSCC) slated to undergo definitive resection. Materials/

Methods: HyPR-HN (NCT05538533) is a Phase I trial to determine the safety of three isoequivalent (EQD2 70 Gy at a/b = 3) but increasingly condensed preoperative RT regimens (46 Gy in 10 fractions, 40 Gy in 7 fractions and 35 Gy in 5 fractions) for patients with OSCC using a Time-to-event Continuous Reassessment Methodology (TiTE-CRM) for dose finding. Three patients enrolled on the HyPR-HN trial and assigned to the first dose level (46 Gy in 10 fractions) underwent biopsy prior to RT (3 patients), immediately following RT on the day of the last radiation fraction (3 patients) and at the time of surgery (2 patients) six weeks after completion of RT. Flow cytometry of immune cell populations and single cell RNA sequencing was performed at each time point for all three patients. T-cell receptor sequencing was performed on one patient at all three time points. Single cell RNA data was imported into R (v.4.3.2) and analyzed using the following packages: Seurat (v5.0) and scRepertoire (v2.0.0).
Results: On flow cytometry, CD8 T cell populations were diminished on the last fraction of RT compared to before RT and failed to recover at the time of surgery in all three patients. This suggests that post-RT immune checkpoint inhibition may be less effective due to T cell depletion. Single cell analysis showed the presence of canonical exhausted and regulatory T cell populations prior to RT, however after radiation these populations were diminished. Additionally, a proliferative CD8+ T cell population (T_{Prolif_Tox}) was readily identified in all three patients prior to RT, but unlike the canonical exhausted and regulatory populations, T_{Prolif_Tox} persisted after RT. T_{Prolif_Tox} co-expressed several cytotoxic markers (GZMB, IFNG), exhaustion markers (TIGIT, TIM3, LAG3, CTLA4, PD1) and cell proliferation markers (STMN, MKI47, CENPF, TOP2A, TOX), suggestive of persistent T cell stimulation in vivo. Analysis of a benign oral mucosa dataset did not demonstrate the presence of a T_{Prolif_Tox} population, suggesting tumor specificity. Further, clonal network analysis both pre- and post-RT demonstrated cycling between the T_{Prolif_Tox} cluster and canonical pre-exhausted clusters suggesting repopulation of the immune microenvironment with this T_{Prolif_Tox} cluster.
Conclusion: RT rapidly and persistently depletes T cells. A proliferative T cell cluster may contain tumor specific T cells and repopulates the tumor microenvironment after RT.

Abstract 2770 – Table 1. Percent of Live Cells on Flow Cytometry

Patient	Pre RT CD8	Pre RT CD4	Post RT CD8	Post RT CD4	Surgery CD8	Surgery CD4
1	6.2	11.0	0.3	1.7	0.9	5.6
2	8.6	8.5	0.1	0.5	2.7	7.6
3	11.0	7.4	0.1	0.4	~	~