Background: Targeting the immune checkpoint CTLA-4 in addition to PD-1/PD-L1 alone did not increase efficacy in HNSCC, whereas this has not been studied in combination with chemotherapy. Induction chemo-immunotherapy followed by pathologic response-based patient selection for chemotherapy-free radioimmunotherapy was efficient in locally advanced HNSCC (J Immunother Cancer. 2022 Jan;10(1):e003747). The expansion cohorts of the CheckRad-CD8 trial studied safety and efficacy of induction chemo-immunotherapy with increased dose or without CTLA-4 inhibition. Methods: Patients with previously untreated stage III-IVB (AJCC 8^th edition) HNSCC were eligible for this multicenter phase II trial. Induction chemo-immunotherapy of the main cohort (MC) consisted of a single cycle of cisplatin 30mg/m² d1-3, docetaxel 75mg/m² d1, durvalumab 1500mg fix dose d5 and tremelimumab 75mg fix dose d5. Patients in expansion cohort 1 (EC1) received this combination with high dose tremelimumab 300mg fix dose d5 and patients in expansion cohort 2 (EC2) received no tremelimumab. In EC1 and EC2 prophylactic G-CSF was recommended. Patients with at least 20% increase of intratumoral CD8+ immune cell density or pathological complete response (pCR) in the re-biopsy entered chemotherapy-free radioimmunotherapy up to a total dose of 70Gy. The current analysis focuses on toxicity and pathologic response after induction chemo-immunotherapy. Results: Between Sep 2018 and Sep 2021, 80 patients were enrolled in the MC (one excluded), 20 in EC1 and 20 in EC2 (one excluded) subsequently. In the MC, EC1 and EC2 a total of 56%, 50%, 58% were stage IV and 29%, 30%, 26% had p16 positive oropharyngeal tumors. Baseline median intratumoral CD8+ immune cell density was 395/mm², 505/mm² and 763/mm² in MC, EC1 and EC2. After induction chemo-immunotherapy 41 (52%), 12 (60%) and 11 (58%) of the patients had pCR in the re-biopsy in MC, EC1 and EC2. Patients with residual tumor after induction therapy had a median intratumoral CD8+ immune cell density of 670/mm², 781/mm² and 1605/mm², which was a median increase by factor 3.0, 2.1 and 4.8 in the corresponding patients’ tissue samples. In the cohorts MC, EC1 and EC2 the overall rate of grade 3-4 adverse events per patient was 1.38, 1.35 and 0.58. The corresponding rate of non-hematologic adverse events per patient was 0.84, 0.95 and 0.37, respectively. Conclusions: Neither increase of tremelimumab dosage nor its omission did significantly affect pathologic response to induction chemo-immunotherapy with cisplatin/ docetaxel/ durvalumab. Non-hematologic toxicity was slightly increased for high dose tremelimumab and clearly decreased without tremelimumab. The role of concomitant administration of tremelimumab with radiotherapy cannot be assessed until the final study analysis. Clinical trial information: NCT03426657.