Surgical extirpation of the primary tumor and draining lymph nodes followed by histopathologic riskadapted adjuvant radiation or chemoradiation remains the standard of care for most patients with Head and Neck Squamous Cell Carcinoma (HNSCC). However, non-viral associated (HPV-) HNSCC is characterized by a high rate of therapeutic resistance: approximately 30-50% of these patients have local or distant recurrence following conventional treatment. Furthermore, there is some evidence to suggest that extirpative surgery itself is immunosuppressive and may promote cancer progression [1]. Identification and successful integration of novel immunotherapy into existing treatment paradigms of surgery and radiation for HNSCC is an evolving treatment approach that has the potential to overcome suppressive mechanisms within the tumor and lead to enhanced survival and decreased morbidity for patients with locally advanced or recurrent HNSCC [2]. Our research in preclinical models has shown that immune responses play an important role in local tumor control following surgical resection [3], and there is increasing evidence that the addition of local or systemic immunotherapy before (neoadjuvant) or after (adjuvant) surgery may enhance survival [3-5].

The tumor environment is the primary target site for anti-tumor immune responses, but commonly evolves during malignant progression to include a range of suppressive mechanisms. Information obtained from the surgical resection specimen may be leveraged to tailor immunotherapy interventions targeting the surgical site to eliminate minimal residual disease and minimize recurrence. Biomaterial platforms can be constructed to provide a local delivery system for such immunotherapies, which can be applied to the resection bed at the time of surgery, and be utilized to enhance the effectiveness of surgery. We recently demonstrated that cyclic-dinucleotides (CDN), which are ligands of STimulator of INterferon Genes (STING), incorporated into a simple biomaterial and placed into the resection cavity were able to eliminate residual disease in preclinical models of HNSCC [4]. CDN are naturally generated following cGAS recognition of cytoplasmic DNA from endogenous sources or following intracellular infection, and CDN binding to STING results in activation of IRF3 and transcription of type I IFN. STING therefore forms part of an endogenous nucleic acid sensing mechanism that can be exploited for cancer therapy [6]. Initial studies using direct injection