Discordant treatment response in primary tumors and lymph node metastases after four weeks of preoperative PD-1 blockade in head and neck squamous cell carcinoma (HNSCC).

person Adam Luginbuhl Thomas Jefferson University Hospital, Department of Otolaryngology, Philadelphia, PA info_outline Adam Luginbuhl, Jennifer Maria Johnson, Madalina Tuluc, Stacey Mardekian, Larry Harshyne, Benjamin Leiby, Ralph Zinner, Joseph M. Curry, David M. Cognetti, Ulrich Rodeck, Athanassios Argiris

Authors person Adam Luginbuhl Thomas Jefferson University Hospital, Department of Otolaryngology, Philadelphia, PA info_outline Adam Luginbuhl, Jennifer Maria Johnson, Madalina Tuluc, Stacey Mardekian, Larry Harshyne, Benjamin Leiby, Ralph Zinner, Joseph M. Curry, David M. Cognetti, Ulrich Rodeck, Athanassios Argiris Organizations Thomas Jefferson University Hospital, Department of Otolaryngology, Philadelphia, PA, Thomas Jefferson University Hospital, Philadelphia, PA, Thomas Jefferson University, Philadelphia, PA, Thomas Jefferson University, Department of Neurological Surgery, Philadelphia, PA, Thomas Jefferson University, Department of Pharmacology and Experimental Therapeutics, Philadelphia, PA, Thomas Jefferson University Hospital, Department of Medical Oncology, Philadelphia, PA, Thomas Jefferson University, Department of Dermatology and Cutaneous Biology, Philadelphia, PA Abstract Disclosures Research Funding Pharmaceutical/Biotech Company Background: Discordant radiographic responses are described in other tumor types in response to immunotherapy with response at some anatomic sites and progression in others. Here we determined the frequency of discordant treatment effects (TE) in HNSCC patients treated with immunotherapy in the context of a neoadjuvant trial. Methods: 23 Patients with resectable primary HNSCC were 1:1 randomized to receive nivolumab (240 mg IV Q 2 weeks x 2) or nivolumab and tadalafil 10 mg daily. Surgery was performed 4 weeks after the first nivolumab infusion. Resection specimens were graded histopathologically by two pathologists. Areas exhibiting TE (defined by fibrosis with chronic inflammation, foamy macrophage reaction and multinucleated giant cells) were expressed relative to the total tumor area. This was assessed in the primary tumor and all lymph nodes (LN). Each primary lesion and individual LN was defined as a) no response 0%TE, b) minimal response 1-19%TE, c) response 20-99% or d) complete response 100%. Concordance was defined if primary lesion and LNs were in the same ordinal data set. Results: 11/23 (48%) of subjects experienced concordant ΤΕ in the primary tumor and LNs. Within this cohort, 3 patients had a complete pathologic response both at the primary site and LNs. In contrast, 12/23 patients (52%) revealed discordant ΤΕ between the primary tumor sites (average of 17% TE) and involved LNs (average of 62% TE), (p= 0.018; signed rank test). Interestingly, in the discordant group, TE effects in LNs were invariably greater than in primary lesions. In 5 of 11 patients with multiple involved LN, the TE varied between nodes. This included patients with adjacent LNs demonstrating 0% and 100% TE in the same level. Systemic and local immune parameters as they relate to concordant and discordant TEs in individual patients will be presented including a type 1 immune bias. Conclusions: Early histologic evaluation of TE in patients with HNSCC receiving immunotherapy demonstrate a wide variety of response between the primary tumor and LNs. Further investigations will lend insight into complex interactions of cancer cells with the microenvironment. Clinical trial information: NCT03238365