Objectives: Immunotherapy added to standard chemoradiation (CRT) seeks to improve outcomes for patients with node-positive cervical cancer. Optimal sequencing of CRT with immunotherapy is unknown. NRG GY-17 (NCT03738228) is a randomized phase I/Ib trial of the anti-PD-L1 antibody, atezolizumab, before and concurrent (Arm A) or concurrent with CRT (Arm B) to determine the best sequence of therapy to result in immune activation, as determined by clonal expansion of T Cell Receptor Beta (TCRB) repertoires and tumor-associated T cell clones in peripheral blood. Secondary objectives included toxicity and the predictive value of T cell repertoire parameters for clinical outcomes. Methods: A total of 40 patients were randomized; 36 patients with locally advanced cervical cancer with positive lymph nodes were randomized to three doses of atezolizumab (1200mg) on day -21, 0, 21 (Arm A) versus day 0, 21, 42 (Arm B). All eligible patients for analysis received CRT and ≥1 dose of atezolizumab. Tumor biopsies were obtained prior to and during therapy, and peripheral blood was collected. T cell repertoire metrics were evaluated by Adaptive immunoSEQ assay. Dose-limiting toxicities were assessed during and up to 30 days post CRT. Comparison of arms and pre- to post-treatment comparisons were performed using Wilcoxon rank-sum or signed-rank test. Correlations of T cell clonality and diversity with clinical outcomes and biopsy results were explored. Results: Of the 40 patients, four patients were not assigned study treatment and were excluded from the analysis. For the 36 eligible and treated patients, the median follow-up time was 20 months. The median age was 48 years; most of the patients were non-Hispanic White and White and had performance status of 0, squamous cell carcinoma, and FIGO stage IIB. Seventy-five percent of all patients completed all protocol therapy, 86% received >4 cycles of cisplatin with RT. Of the 36 patients, 30 were DLT evaluable according to the protocol: Arm A: 16 patients with no DLTs; Arm B: 14 patients with three patients reported to have a DLT (8%): one immune-related event as colitis (3%), non-immune-related colitis, and thrombocytopenia with cisplatin delay >2 wks. Overall, three patients in Arm A and seven patients in Arm B experienced a grade 3 or higher treatment-related adverse event; except one, all grade 3 events were deemed non-immune-related. Thirty-one patients had on-treatment tumor biopsy at the first brachytherapy; ten patients (28%) showed no residual tumor on biopsy. There was an increase in peripheral blood TCR clonal expansion and expansion of tumor-associated T cell clones between baseline and day 21 in Arm A (p=0.0001) and Arm B (p=0.001). There was no statistically significant difference between the two treatment arms for either T cell clonal expansion or expansion of tumor-associated T cell clones. Patients with higher pre-treatment TCR diversity had an increased likelihood of complete pathologic response in on-treatment biopsy (p= 0.049). Overall, at 12 months, the DFS for the entire cohort is 72%. Conclusions: Our data indicate that atezolizumab as a primer and concurrent with CRT is safe and shows immune-modulating activity in women with locally advanced cervical cancer. There was no difference in change in T cell clonality between the arms. Favorable DFS was observed in both arms at 12 months for this high-risk patient population. Correlation between the treatment schedule, T cell repertoire parameters, and clinical outcomes will be performed as the follow-up becomes more mature. Objectives: Immunotherapy added to standard chemoradiation (CRT) seeks to improve outcomes for patients with node-positive cervical cancer. Optimal sequencing of CRT with immunotherapy is unknown. NRG GY-17 (NCT03738228) is a randomized phase I/Ib trial of the anti-PD-L1 antibody, atezolizumab, before and concurrent (Arm A) or concurrent with CRT (Arm B) to determine the best sequence of therapy to result in immune activation, as determined by clonal expansion of T Cell Receptor Beta (TCRB) repertoires and tumor-associated T cell clones in peripheral blood. Secondary objectives included toxicity and the predictive value of T cell repertoire parameters for clinical outcomes. Methods: A total of 40 patients were randomized; 36 patients with locally advanced cervical cancer with positive lymph nodes were randomized to three doses of atezolizumab (1200mg) on day -21, 0, 21 (Arm A) versus day 0, 21, 42 (Arm B). All eligible patients for analysis received CRT and ≥1 dose of atezolizumab. Tumor biopsies were obtained prior to and during therapy, and peripheral blood was collected. T cell repertoire metrics were evaluated by Adaptive immunoSEQ assay. Dose-limiting toxicities were assessed during and up to 30 days post CRT. Comparison of arms and pre- to post-treatment comparisons were performed using Wilcoxon rank-sum or signed-rank test. Correlations of T cell clonality and diversity with clinical outcomes and biopsy results were explored. Results: Of the 40 patients, four patients were not assigned study treatment and were excluded from the analysis. For the 36 eligible and treated patients, the median follow-up time was 20 months. The median age was 48 years; most of the patients were non-Hispanic White and White and had performance status of 0, squamous cell carcinoma, and FIGO stage IIB. Seventy-five percent of all patients completed all protocol therapy, 86% received >4 cycles of cisplatin with RT. Of the 36 patients, 30 were DLT evaluable according to the protocol: Arm A: 16 patients with no DLTs; Arm B: 14 patients with three patients reported to have a DLT (8%): one immune-related event as colitis (3%), non-immune-related colitis, and thrombocytopenia with cisplatin delay >2 wks. Overall, three patients in Arm A and seven patients in Arm B experienced a grade 3 or higher treatment-related adverse event; except one, all grade 3 events were deemed non-immune-related. Thirty-one patients had on-treatment tumor biopsy at the first brachytherapy; ten patients (28%) showed no residual tumor on biopsy. There was an increase in peripheral blood TCR clonal expansion and expansion of tumor-associated T cell clones between baseline and day 21 in Arm A (p=0.0001) and Arm B (p=0.001). There was no statistically significant difference between the two treatment arms for either T cell clonal expansion or expansion of tumor-associated T cell clones. Patients with higher pre-treatment TCR diversity had an increased likelihood of complete pathologic response in on-treatment biopsy (p= 0.049). Overall, at 12 months, the DFS for the entire cohort is 72%. Conclusions: Our data indicate that atezolizumab as a primer and concurrent with CRT is safe and shows immune-modulating activity in women with locally advanced cervical cancer. There was no difference in change in T cell clonality between the arms. Favorable DFS was observed in both arms at 12 months for this high-risk patient population. Correlation between the treatment schedule, T cell repertoire parameters, and clinical outcomes will be performed as the follow-up becomes more mature.