PubMed2002
UNLABELLED: It has been reported that cytarabine, acting by at least two different mechanisms, enhances the cytotoxic effect of cisplatin in in vitro systems. The aim of this open, prospective, randomised study was to estimate the eventual benefits from the inclusion of high-dose cytarabine in the cisplatin-5-fluorouracil (5-FU) regimen as first-line treatment of patients with advanced head and neck cancer. The study recruited successive patients with unresectable grade I/II head and neck cancer who were not suitable for irradiation treatment (T any N3 or T4 N2C), metastatic or previously irradiated. All patients gave their informed consent. A joint ear, nose and throat (ENT) oncological committee performed the selection. A total of 170 patients were included in the study. Patients randomised to arm A were given 1000 mg/m(2) cytarabine on day 1 preceding for 6 h cisplatin infusion, 30 mg/m(2)/24 h cisplatin intravenous (i.v.) bolus days 1-4 and 1000 mg/m(2)/24 h 5-FU in a 4-h infusion on days 1-4. Patients in arm B were given cisplatin and 5-FU in the same dosage and schedule as in arm A. Additional irradiation+/-surgery was performed if and when feasible. Patients in both arms were well balanced with regard to clinical variables. The following results were obtained: Arm A: 84 patients were included, 74 were evaluable for activity; RESPONSE: complete response (CR) 8 (11%), partial response (PR) 40 (54%), stable disease (SD) 11 (15%), progressive disease (PD) 15 (20%). The overall response rate (RR) based on the evaluable patients was 48/74 (65%, 95% confidence interval (CI) 54-75%); The RR based on an intent-to-treat analysis was 57%, 95% CI 47-67%; Median survival was 13 months; There were 50 episodes of granulocytopenia grade IV and 15 of febrile neutropenia per 316 cycles. Arm B: 86 patients were included, 80 were evaluable for activity; RESPONSE: CR 7 (9%), PR 29 (36%), SD 10 (12.5%), PD 34 (42.5%); The overall RR based on the evaluable patients was 36/80 (45%, 95% CI 35-56%); The RR based on an intent-to-treat analysis was 42%, 95% CI 32-52%; Median survival was 8 months; There were 14 episodes of granulocytopenia grade IV and 7 febrile neutropenias per 324 cycles. The RR was significantly higher in arm A (P=0.013), power (one-sided) 80%. The proportion of patients from the appropriate subset who achieved a clinical response making additional treatment feasible was higher in arm A (P=0.00015), as well as the proportion of patients with a performance status 2+3 achieving a response (P<0.0001). Using the Log-rank test, patients from arm A achieved a significantly longer survival (P=0.009), with the probability of survival at 12 months of 0.58 for patients in arm A and 0.28 for patients in arm B. Grade IV granulocytopenia and thrombocytopenia were more frequent in arm A. Due to its haematological side-effects, cytarabine might not be the ideal drug to modulate the cytotoxicity of cisplatin. However, other modulators of its activity could be of interest for further studies in head and neck cancer patients.