A well-known mechanism for the development of leukaemias is that the chromosomes in a cell break apart and then recombine in an incorrect way. At the points of fissure, gene fragments are exposed that can recombine with so-called fusion genes, yielding fusion proteins.
In contrast, solid tumours have been seen as developing as a result of cells losing the inhibiting mechanism of tumour suppressor genes.
The researchers, from Lund University in Sweden, now believe this model works for hereditary cancers, but does not apply in the case of spontaneous malignancies, which make up 90-95 per cent of all cancer cases.
"In non-hereditary cancer forms it is the occurrence of fusion genes and not the lack of tumour suppressor genes that is essential," said Professor Felix Mitelman, the leader of the research team.
The study drew on the enormous database of cytogenetic abnormalities held at the Cancer Genome Anatomy Project at the US National Cancer Institute, called the Mitelman Database of Chromosome Aberrations in Cancer.
In leukaemia cells it is rather easy to find fusion genes and fusion proteins. But for technical reasons, this is much more difficult in solid tumours.
"And if you haven't seen them, you assume that they're not there. But what has been lacking is appropriate methods of examination," claims Mitelman.
The research team found that the number of fusion genes in solid tumours stood in the same proportion to the number of patient cases examined with leukaemias. This suggests that the same mechanisms are involved, they conclude.
The finding lends further weight to the idea that fusion proteins are a valid target in cancer. However, the downside is that there are probably a very great number of different fusion genes behind the major forms of cancer, with each transformation of genes found in just a few patients.
"Small groups of patients are not of interest to pharmaceutical companies. On the other hand, it may be that several fusion proteins have common traits that make it possible to use the same drug to combat them," suggests Mitelman.
The research is published in the journal Nature Genetics (30 March).