PURPOSE: Radiation-induced emesis is one of the most disturbing side effects of total body irradiation (TBI). To evaluate the efficacy and to determine the best schedule of granisetron (a selective 5-hydroxytryptamine3 serotonin receptor antagonist) administration in the prevention of radiation-induced nausea and vomiting, we conducted a trial involving patients receiving single-dose TBI before bone marrow transplantation (BMT). METHODS AND MATERIALS: Thirty-six patients with non-Hodgkin's lymphoma (n = 12), multiple myeloma (n = 8), acute lymphoblastic leukemia (n = 7), acute nonlymphoblastic leukemia (n = 6), and chronic myeloid leukemia (n = 3) referred to our department between March 1992 and February 1994 were enrolled in this study to assess the efficacy of granisetron during single-dose TBI before autologous BMT (n = 26), allogeneic BMT (n = 8), or syngeneic BMT (n = 2). The male-to-female ratio was 22:14 (1.57), and the mean age was 41 +/- 11 years (range 16-58). Before TBI, conditioning chemotherapy consisted of cyclophosphamide (CY) alone (60 mg/kg per day on 2 successive days) in 24 patients, CY combined with other drugs in 6, and combinations without CY in 6. All patients received single-dose TBI (10 Gy administered to the midplane at L4, and 8 Gy to the lungs). The mean instantaneous and average dose rates were 0.039 +/- 0.012 Gy/min (range 0.031-0.058), and 0.025-0.006 Gy/min (range 2.08-3.96), respectively. Granisetron was administered 30-45 min before TBI according to two different modalities: a total dose of 3 mg as a 5-min intravenous (i.v.) infusion (Treatment A, n = 15; 42%) or the same treatment plus 3 mg of granisetron as a 24-h continuous i.v. infusion (total dose: 6 mg, Treatment B, n = 21; 58%). Depending on the BMT teams, hyperdiuresis was continued (n = 19, 53%) or suspended (n = 17, 47%) during TBI. Nausea and vomiting were assessed during the TBI session and the following 12 h, and were scored as follows: S1 = no nausea or vomiting; S2 = moderate nausea; S3 = severe nausea and/or single episode of vomiting; and S4 = multiple episodes of vomiting. RESULTS: During TBI, 18 (50%) patients were scored as complete responders (S1), 1 (3%) as a major responder (S2), 9 (25%) as minor responders (S3), and 8 (22%) as nonresponders (S4). During the following 12 h, 28 (78%) patients were free of severe nausea and vomiting (S1 or S2), whereas 8 (22%) vomited (S3 or S4). In univariate analyses, the 12-h probability of emesis was significantly higher in patients undergoing hyperdiuresis (63% vs. 30%; p = 0.05), and in patients older than 45 years (65% for age > 45 vs. 33% for age < or = 45; p = 0.05). The probability of S3 or S4 emesis was 50% with Treatment A and 47% with Treatment B (p = 0.86). Sex, body weight, and type of conditioning chemotherapy did not influence the 12-h probability of emesis. Multivariate analysis revealed that hyperdiuresis (p = 0.02) and Treatment A (p = 0.04) were independently associated with radiation-induced emesis, whereas sex (p = 0.85), body weight (p = 0.13), age (p = 0.12), and type of conditioning chemotherapy (p = 0.92) were not. No early toxicity related to granisetron was observed. CONCLUSION: Granisetron is a well-tolerated and effective antiemetic agent that can be used as monotherapy during single-dose TBI. Good control of nausea and vomiting is obtained with this antiemetic drug, and its effect is increased when hyperdiuresis is suspended during TBI.