Background: Encouraging responses to chemoimmunotherapy with I/T/DIN/GM-CSF have been observed in trials for patients (pts) with relapsed/refractory HR-NBL, but factors associated with response have not been identified and duration of response has not been assessed. We aimed to evaluate timing and duration of response among pts with relapsed HR-NBL treated with I/T/DIN/GM-CSF and identify factors associated with response. Methods: We performed a multicenter retrospective cohort study of pts treated with I/T/DIN/GM-CSF. Eligibility criteria included: diagnosis of relapsed HR-NBL prior to age 30; objective response [OR; complete, partial, or minor response (CR, PR, or MR) by International Neuroblastoma Response Criteria (INRC)] or stable disease (SD) after initial therapy; receipt of I/T/DIN/GM-CSF for relapse or progression outside a clinical trial from 1/1/15-6/1/20. Logistic regression was used to identify factors associated with OR. Kaplan Meier analysis was used to determine progression-free survival (PFS). Results: We enrolled 143 pts with a median age at diagnosis of 51 months. Tumors were MYCN amplified in 52 (36%) and ALK was wild type in 73/94 (78% of tumors in which ALK status was known). 79 (55%) had received prior anti-GD2 therapy. I/T/DIN/GM-CSF comprised first relapse therapy in 96 pts (67%), second relapse therapy in 23 (16%) and subsequent therapy in 24 (17%). 70 (49%) achieved OR following I/T/DIN/GM-CSF therapy [29% CR, 15% PR, 5% MR], 30 (21%) achieved SD and 43 (30%) progressed. Median cycles received was 5 (range 1-31). 121 patients (85%) had their best response upon first disease evaluation. Later disease evaluations showed improved INRC classification in 14% of pts with initial SD, 33% with MR, and 41% with PR. Median time to OR was 2 months (range 1-21). Of the 105 relapse/progression events after starting I/T/DIN/GM-CSF (73% of pts), 59 (56%) occurred during therapy. Of the 42 pts who achieved CR with I/T/DIN/GM-CSF, 5 (12%) relapsed during I/T/DIN/GM-CSF and 17 (40%) relapsed after discontinuation. I/T/DIN/GM-CSF was discontinued in 83 pts (58%) due to suboptimal response or PD, and in 19 (13%) for toxicity. Median PFS among objective responders was 15.5 months. Among those in CR, median PFS after discontinuation of I/T/DIN/GM-CSF was 11.8 months (range 0.7-70.6). Multivariable models did not identify clinical or biologic factors associated with OR. Conclusions: 49% of pts receiving I/T/DIN/GM-CSF for relapsed HR-NBL achieved OR. Among responders, median response duration was 15.5 months. Pts with SD on first disease evaluation were unlikely to achieve OR, but > 1/3 of pts with MR/PR on first evaluation ultimately achieved CR. No identifiable clinical or biologic factors were associated with OR.