Results of the Eastern Co-operative Oncology Group (ECOG) 1684 trial of high-dose interferon (HDI) versus observation in resected stage IIB or III melanoma (primary tumour >4 mm in depth and/or lymph node involvement) were published in 1996 (Kirkwood et al, 1996). The study was the first to show a statistically significant improvement in relapse-free and overall survival in these patients with a high risk of relapse. This was hailed as a breakthrough in the USA: here at last, after many negative studies, was an effective treatment. In Europe the response was more muted, with many oncologists unwilling to submit patients to such an arduous regimen without further supportive evidence. The results of the Intergroup E1690 study were therefore disappointing: although the improvement in relapse-free survival with high-dose treatment seen in the ECOG 1684 trial was confirmed there was no improvement in overall survival (Kirkwood et al, 2000a).The difference in these results is due to an improvement in the outcome of patients kept under observation in the later trial, there being an increase in median survival after relapse from 1.8 years in ECOG 1684 to 4.3 years in E1690. The reasons for this remain unclear: salvage therapy with interferon at relapse accounts for some of the increase and there were differences between the two study populations. Whatever the explanation, these conflicting reports on the efficacy of HDI have led to only limited uptake of the regimen in the UK to date.A number of studies of low-dose interferon (LDI) in this indication have also been undertaken. The first major study, conducted under the auspices of the World Health Organization, showed no improvement in relapse-free or overall survival for patients with stage III disease at long-term follow-up (Cascinelli, 1999), a finding in common with the LDI arm of E1690. Two large trials have examined LDI treatment in patients with resected stage II disease, i.e. tumours >1.5 mm in depth without nodal metastases. An Austrian study of over 300 patients demonstrated increased disease-free survival for a year of LDI, compared with observation, but as yet no effect on overall survival (Pehamberger et al, 1998). The larger French study compared 18 months LDI with observation in 489 patients. Again, there was a clear advantage for LDI in terms of disease-free survival, and a trend towards improved overall survival (Grob et al, 1998). Similar results with only 6 months treatment are published in this issue by the Scottish Melanoma Group, although the relatively small numbers recruited to their study preclude meaningful statistical analysis (Cameron et al, 2001). The UKCCCR AIMHIGH study, which has recently closed to recruitment, will provide more evidence on the effects of LDI in the stage IIB and III population.Results from the latest Intergroup trial, E1694, are now available. This study randomized patients between HDI (according to the regimen in the earlier ECOG studies) and a GM2 ganglioside vaccine. The study was designed before the results of E1690 ...