Background: Chemoresistant disease is an obstacle for cure of high-risk neuroblastoma (HR-NB). Anti-GD2 monoclonal antibodies (MoAb) dinutuximab and naxitamab in combination with cytokines are FDA-approved to consolidate remission and for chemorefractory osteomedullary HR-NB, but responses in progressive disease (PD) are rare. We investigated the combination of Humanized anti-GD2 MoAb naxitamab (Hu3F8), Irinotecan, Temozolomide and Sargramostim (GMCSF) in a phase II "HITS" protocol against resistant HR-NB (NCT03189706). Noteworthy differences between HITS and COG protocol ANBL 1221 included higher MoAb and temozolomide dosage and overlap of naxitamab with GMCSF. Methods: Patients were treated at Memorial Sloan Kettering (MSK) on protocol and at Hospital Sant Joan de Déu (HJSD) per protocol on compassionate basis. Salient eligibility criteria included evaluable or measurable chemoresistant disease. Prior anti-GD2 MoAb or irinotecan/temozolomide (IT) therapy was permitted. Each cycle, administered 3-5 weeks apart, comprised irinotecan 50 mg/m²/day intravenously (IV) plus temozolomide 150 mg/m²/day IV or orally (days 1-5); naxitamab 2.25 mg/kg/day IV, days 2,4,8 and 10, and GMCSF 250 mg/m²/day subcutaneously, days 6-10. Toxicity was measured by CTCAE v4.0 and responses by International Neuroblastoma Response Criteria. Objective responses (OR) were also noted. The primary endpoint of the phase II trial was complete (CR) and partial response (PR) after 4 cycles with a desirable rate of 40%; type I and II errors of 10% (undesirable=20%). Results: Of 90 heavily prior-treated patients (38 at MSK evaluated on trial, 52 at HJSD), 8 had HR-NB refractory to induction chemotherapy while 82 had up to 6 prior relapses (median=1). 503 cycles (median 5/patient) were administered. Toxicities included myelosuppression and diarrhea expected with IT, pain and hypertension expected with naxitamab, plus febrile neutropenia in 4%. No other >grade 2 unexpected toxicities occurred; treatment was outpatient. Primary endpoint was reached in the phase II trial: INRC response = 30.6%, lower boundary = 20.4%. In the entire cohort, best responses were CR (26%), PR (11%), mixed response (9%), stable disease (27%) and PD (27%). OR were noted in 64%, with soft tissue (48%) and skeletal MIBG uptake (66%). CR in BM was seen in 57%. OR occurred in patients with MYCN-amplified (25%), refractory (100%) and relapsed (61%) HR-NB; and patients who had previously received I/T (64%) or naxitamab (68%). In patients who had previously received dinutuximab/IT, OR rate to HITS was 42% (5/12). Human anti-human antibody did not develop in any patient (n=50). Conclusions: Naxitamab-based chemoimmunotherapy was safe without immunogenicity. It was effective against chemoresistant HR-NB in all disease compartments even in patients with multiple prior relapses, and in patients who previously received anti-GD2 MoAbs and/or IT.