RANDOMIZED TRIALS ON HYPERFRACTIONATED AND ACCELERATED FRACTIONATION IN HEAD AND NECK CANCERS

From 1978 to March 1998, (1867 patients) were accrued in head and neck trials comparing hyperfractionation (HF) and accelerated fractionatio n (AF) to classical fractionation (CF). Two randomized trials (867 pts ) led to positive conclusions in favour of the HF & AF arms: Il EORTC trial 22791 (356 patiens, 1980-87) compared CF (70 Gy/35-40 fr/7-8 wks ) to HF (80.5 Gy/70 fr/7 wks, using 2 fr x 1.15 Gy/day) in T2 T3, No-N 1<3 cm in oropharyngeal carcinoma. Locoregional control (LRC) was high er (p=0.01) in HF versus CF. At 5 years, 56 % of the patients are LRC free with HF versus 38 % with CF on the latest update (February 1998). This improvement of LRC also resulted in a significant overall surviv al (p=0.05). There was no difference in late normal tissue damage betw een the two treatment modalities. Overall, this is the largest improve ment documented in a randomised trial for oropharyngeal cancers during the past decade 2) EORTC trial 22851 (511 patients, 1985-1995) compar ed AF(72 Gy/45fr/5 wks) to CF(70 Gy/35 fr/7 wks) in T2 T3 T4 head & ne ck cancers (hypopharynx excluded). Acute and late toxicity were increa sed in the AF arm. Late severe sequelae occurred in 14 % of patients o f the AF ann versus 4 % in the CF arm. Two cases of radiation-induced myelitis occurred after doses of 42 and 48 Gy to the spinal cord The A F arm is significantly better for locoregional control (p=0.017) for t ime to progression (p=0.012) resulting in a 15 % locoregional gain at 5 years over the CF arm. This improvement is of larger magnitude inpat ients with poorer prognosis (N3 any T T4 any N) than inpatients with m ore favourable stage. Multivariate analysts confirmed AF as an indepen dent pronostic factor for local control (p=0.03). Specific survival sh ows a non significant advantage (p=0.06) in favour of the AF arm. This trial shows that accelerated radiotherapy is able to improve locoregi onal control in a large variety of head and neck squamous cell carcino mas. A less toxic scheme should however be investigated before moving AF schemes in standard practice. To conclude, these two schemes derive d from experimental radiobiology concepts resulted in a significant im provement of locoregional control. Hyperfractionation resulted in an i mproved locoregional and survival benefit. Although HF is presently th e most reliable regimen to improve locoregional control the validity o f the concept of AF is also confirmed. Better schemes of AF should now be evaluated to reduce late toxicity.