A comparative review of the Hammersmith (1971-75) and Edinburgh (1977-82) neutron therapy trials of certain cancers of the oral cavity, oropharynx, larynx and hypopharynx. Medical Research Council Neutron Therapy Working Group.

Because of apparent inconsistencies in the two trials of neutron therapy carried out under its auspices, the MRC in 1982 requested a comparative review of the data in the Hammersmith and Edinburgh trials in so far as the clinical material allowed of such comparison. It was clear that the two trials were not only not contemporaneous (Hammersmith 1971-75; Edinburgh 1977-82) but that they differed in site and stage of disease as well as in the parameters of radiation dosage. Only a fraction of the patients were comparable and this review, which sets out the differences in design and conduct of the two trials, focuses on squamous-cell carcinoma in the four sites--oral cavity, oropharynx, larynx and hypopharynx--which were common in both. The first 120 patients entered into the Edinburgh trial have been compared with the 95 patients selected from 161 in the Hammersmith trial who would have been eligible at Edinburgh by reason of the site of disease and histology. Thus, Hammersmith patients with tumours of salivary glands, sinuses and neck nodes were, of necessity, excluded. Within such limitations the data on case distribution, tumour size and stage, the parameters of radiation dosage and the outcome in terms of tumour control, survival and treatment morbidity have been subjected to rigorous assessment and analysis. The important aspects in which the two investigations had differed included the stage of disease (more advanced at Hammersmith, with 63% of patients having fixed nodes, compared with 27% at Edinburgh); radiation dosage (the median neutron dose being 5% lower at Edinburgh than at Hammersmith while the median photon dose was l0% higher at Edinburgh); the "photon arm" of the Edinburgh trial being on site whereas that at Hammersmith had been at multiple centres; and the years in which the trials had been carried out. The two series are comparable in terms of treatment failure (local recurrence or late radiation death) following local tumour control. The results are also consistent in associating an increased incidence of late severe radiation morbidity and of intercurrent deaths with neutron therapy. In three other respects the results of the two series differ. At Edinburgh there was advantage to the photon-treated patients in terms of overall mortality, whereas at Hammersmith the advantage was to those treated with neutrons. The incidence of tumour regression for patients randomised to photons was much lower in the Hammersmith trial than for neutrons, whereas in Edinburgh the two modalities gave similar results.