Randomized double-blind trial of amifostine versus placebo for radiation-induced xerostomia in patients with head and neck cancer.

<bold>Introduction: </bold>The role of the radioprotector amifostine in ameliorating radiotherapy side effects in head and neck squamous cell carcinoma (HNSCC) is controversial. This trial aimed to determine whether pretreatment with amifostine reduced the incidence of Radiation Therapy Oncology Group grade ≥2 acute and late xerostomia in patients receiving definitive or adjuvant radiotherapy for HNSCC, without reducing tumour control or survival.<bold>Methods: </bold>Between 14 September 2001 and 8 November 2004, 44 Royal Adelaide Hospital patients were randomized double-blind to receive amifostine (200 mg/m2 IV) or placebo (normal saline IV) 5 days/week, prior to standard radiotherapy (60-70 Gy), each having ≥75% of the parotids treated to ≥40 Gy. Side effects were assessed weekly during treatment, at 3 and 5 months after radiotherapy, then every 6 months until disease progression or death.<bold>Results: </bold>The accrual target was 200 patients over 4-5 years, but the trial closed prematurely when only 44 patients had been randomized after 3 years. Of 41 evaluable patients, 80% (16/20) in the amifostine arm had grade ≥2 acute radiation salivary toxicity versus 76% (16/21) in the placebo arm (P = 1.00). The rate of grade ≥2 late radiation salivary toxicity at 12 months was 66% in the amifostine arm and 82% in the placebo arm (estimated hazard ratio 1.61, 95% confidence interval 0.74-3.49, P = 0.22). Other toxicities tended to be worse in the amifostine arm: acute grade 3-4 skin 35% vs 5% and mucous membrane 40% vs 5%; grade ≥2 vomiting 35% vs 5%, hypocalcaemia 25% vs 5% and fatigue 85% vs 33%, with only the latter retaining statistical significance after adjusting for multiple comparisons. There were no significant differences in failure-free (P = 0.70) or overall survival (P = 0.86), with estimated 4-year rates of 48% vs 54% and 49% vs 59% for the amifostine vs placebo arms respectively.<bold>Conclusion: </bold>There was no clear evidence that pretreatment with amifostine made any difference to the incidence of grade ≥2 acute or late xerostomia. Other toxicity tended to be more severe with amifostine. There was no effect on failure-free or overall survival. Acknowledging the low statistical power, these results do not support the use of IV amifostine pre-radiotherapy in HNSCC. [ABSTRACT FROM AUTHOR]