[Granisetron (per os) compared with ondansetron (per os) in the prevention of nausea and vomiting induced by mildly emetogenic chemotherapies. Groupe de Recherches en Cancerologie du Nord].

It is a randomised cross-over multicenter study comparing the efficacy and the tolerance of granisetron (Gra) 1 mg and ondansetron (Ond) 8 mg, oral, given during two consecutive cycles to 188 naive patients scheduled to receive a moderately emetogenic chemotherapy. The antiemetic treatment is given one day per course, 1 hour before chemotherapy and the second administration from 8 to 12 hours after the beginning, during each of the two cycles; alternatively according to the randomisation. Five criteria are assessed; nausea (ordinal and visual analogic scales), emeric episodes (vomiting orland retching), complete response (minor or no nausea, no emetic episode and no rescue treatment), patient preference and tolerance. The intent to treat analysis showed no significant difference at cycle 1 between Gra and Ond; at cycle 2, there is no significant difference in the number of emetic episodes; for the prevention of nausea, the ordinal scale shows a significant difference (p = 0.028 in favour of Gra at day 1 (D1) but not from D2 to D5. Gra induced more complete response than Ond at D1 (p = 0.028), but not from D2 to D5. The cross-over study did not show any period or order effect, whereas a treatment effect on Ond was significant in favour of Gra (p = 0.01). There is no significant patients preference in favour of Gra or Ond. In conclusion, Gra was more efficient in preventing nausea and obtaining complete response on the first day of treatment, significantly at the second cycle. Both Gra and Ond had a good antiemetic activity for moderately emetogenic chemotherapy with complete response rates always over 50% on day 1; delayed emesis remain less weli controlled.