Cancer of the oesophagus--how should tumour biology affect treatment?

An analysis of the biology of cancer of the oesophagus suggests that significant dissemination may not have taken place at the time of diagnosis in approximately one-third of cases. Positive nodal histology in the mediastinum or abdomen is the best available evidence of disseminated disease. Patients without such nodal involvement therefore represent a group in whom the disease is most likely to be localized. It is suggested that only these patients be subjected to aggressive local therapy. The effects of no treatment, orth- and megavoltage radiotherapy, surgery and combinations of surgery and radiotherapy are reviewed. The need for cooperative randomized clinical trials to test the biological hypothesis outlined is discussed. Megavoltage radiotherapy as a primary treatment for oesophageal cancer followed by radical surgery in those patients without evidence of dissemination from nodal histology is proposed as a basis for such clinical trials.