The surprising discovery that a small simple chemically inert compound, dimethylnitrosamine, was a potent carcinogen, was made by Barnes and Magee in 1956. A few years later I joined them in an attempt to find out how and why, in terms of metabolism of the carcinogen, and reactions with protein, RNA and DNA. From work carried out mainly in Heidelberg and in America, it was very soon found that many of the N-nitroso compounds were carcinogenic, they showed a surprising organospecificity, and that more than half of the 300 tested were carcinogenic for the esophagus. When it became apparent that human exposure to the compounds in the environment was widespread, and that no other chemically identified environmental compounds were potent esophageal carcinogens in animal experiments, the obvious possibility that they were a major cause of esophageal cancer was considered.

The problem in testing this concept was that human exposure was so widespread but difficult to quantify that epidemiological surveys could not associate the disease with exposure to nitrosamines. Instead incidence correlated with other risk factors, especially alcohol consumption and dietary deficiencies. A likely explanation seemed to me to be in the induction of cell replication. I had found that, while continuous feeding of dimethylnitrosamine induced liver cancer, one single injection of the compound was not carcinogenic unless given when the liver cells were dividing in response to partial hepatectomy. The secondary risk factors of alcohol, dietary deficiencies and mycotoxins were shown to stimulate basal cell replication in the esophagus.