Background: Several 5-HT₃ antagonists are available to prevent chemotherapy-induced nausea and vomiting (CINV); when control is inadequate with one agent, another may be used. Using data from a randomized phase III trial (Grous et al. ASCO 2009, #9627), we examined the efficacy of APF530, a sustained delivery formulation of the 5-HT₃antagonist granisetron, in patients (pts) who failed to achieve a complete response (CR; no emesis or rescue medication) with palonosetron (PALO) in preventing acute (0-24 h) and delayed (24-120 h) CINV in pts receiving MEC or HEC. Methods: 1,428 pts receiving single doses of MEC or HEC were randomized to APF530 250 mg (5 mg granisetron) subcutaneously (SC), APF530 500 mg (10 mg granisetron) SC, or PALO 0.25 mg intravenously (IV) in cycle 1 (C1). Prior to C2, pts who received PALO in C1 and remained on study were re-randomized to APF530 250 mg or 500 mg SC. CR rates in C2 were assessed for pts receiving APF530 500 mg who did not achieve CR in C1 with PALO. Results: 446 pts received PALO in C1 (208 MEC; 238 HEC). Of these, 194 (43.5%) were overall (0-120 h) failures (100/208 [48.1%] MEC; 94/238 [39.5%] HEC). Of 194 C1 PALO failures, 72 were re-randomized prior to C2 to APF530 500 mg (38 MEC; 34 HEC). Of 38 MEC PALO failures who received APF530 in C2, overall CR was 39.5% (57.9% acute; 38.2% delayed). Of 34 HEC PALO failures who received APF530 in C2, overall CR was 41.2% (58.3% acute; 45.5% delayed). In the acute phase, > 50% of MEC and HEC pts who failed PALO in C1 achieved CR to APF530 500 mg in C2. CR rate for pts receiving MEC or HEC was slightly less in the delayed vs acute setting. Conclusions: APF530 500 mg demonstrated substantial activity (ie, CR) in pts receiving MEC or HEC who had failed PALO in C1. Failure to achieve an initial CR to PALO 0.25 mg IV does not predict failure of APF530 500 mg SC in subsequent MEC or HEC cycles. Further studies are needed to confirm these observations. Clinical trial information: NCT00343460.