ABSTRACT We have shown that elective nodal irradiation (ENI) blunts systemic immunity in preclinical models of head and neck cancer (HNC), but without it, the risk of regional recurrence in the lymph nodes (LN) is high. Data from clinical trials showed that omission of ENI is feasible, but nodal progression or recurrence does vary by the choice of immunotherapy (IO) that is combined with the radiation (RT). Our previous trial showed that combination aPDL1 (Durvalumab) with tumor-only RT resulted in exceptional high pathological and clinical response rates and no nodal recurrences. However, in two other separate trials, when either another a- PDL1(Atezolizumab) or a-CTLA4 was added to Durvalumab-RT, it resulted in high rates of nodal recurrence leading to trial closures. This suggests that target cell activation or inhibition varies by the type of IO or combination IO used in the context of RT and dictates nodal outcome. Our preliminary data show that trial patient progressive nodes contained a high percentage of Tregs and, preclinically, genetic knockout of PD1 on Tregs increases their activation, which we have previously shown to be driven by STAT3. We demonstrate that nodal metastasis can be prevented by early depletion of Tregs prior to tumor implantation. This is accompanied by enhanced morphological maturation of high endothelial venules (HEVs), specialized postcapillary venules whose function within DLNs has been correlated with immune trafficking and activation. Localized within perivascular niches, and surrounded by Tregs and cancer cells, these HEVs, which express LTBR, are decreased in numbers and maturation on nodal recurrence, a phenomenon that is reversed by LTBR agonists. We also show that ST2 Tregs, a specific Tregs subset, are upregulated with RT and that genetic knockout of ST2 on Tregs results in marked tumor regression. Recombinant administration of IL33 (ST2's ligand), on the other hand, combined with RT and Treg depletion, leads to high rates of eradication of nodal recurrence. We hypothesize that activation of PD1 expressing Tregs in the RT-IO setting contributes to cancer immune evasion directly via STAT3 pathway activation, indirectly through LTA1B2-LTBR Treg-HEV interaction, thereby limiting the Teff infiltration into the nodal environment, and through the ST2-IL33 Treg-stromal interaction, abrogating Teff cells' expansion potential. In Aim1, we will examine the mechanistic impact of immune check point inhibition on Tregs; in Aim2, we will test how Treg-HEV interaction facilitates a tumor- promoting environment within the LN, leading to immunosuppression and cancer cell immune evasion; and in Aim 3, we will determine how ST2 expressing Tregs modulate tumor immune microenvironment in nodal progression. These will be tested using genetically engineered mouse models, multicompartmental mass cytometry, scRNA seq, metabolomic and proteomic analysis, and multispectral immunofluorescence preclinically and in clinical trial speci...