Cancer—A Biological Approach*

There is still an active body of opinion that would regard cancer as a parasitic disease due to infection of cells by extrinsic viruses. Such an attitude has the great attraction that it suggests straightforward approaches to the prevention of cancer by immunological methods and holds out more hope of an eventual effective chemotherapy. As various writers have pointed out, the claim that some neoplastic processes are due to virus infection is not inconsistent with a recognition of the overriding importance of somatic mutation. Any claim that all cancers are of infective character is unjustified and frankly absurd in the present state of knowledge. It has never been easy to define a virus, and the recognition of provirus in bacteria, with the convincing evidence that it is intimately built into the host's genetic mechanism, has now made it even more difficult. Any definition must cover the fact that there are two very different systems from each of which by appropriate manipulations a continuing sequence of typical virus infections may be produced. These are respectively the infective particle and the infected host cell. A virus, in the conventionally understood form of a population of infective virus particles, can be defined as a self-replicating agent or organism smaller than bacteria and capable of multiplication only within living susceptible host cells. It may be desirable to add certain qualifications when we have to discuss demonstrable or hypothetical self-replicating subcellular entities. A virus particle must be shown to carry specific patterns in protein and nucleic acid which are different from those of the host cell and which are not derived genetically from the host species or any species related to it. At a certain stage of infection, host cells may contain no virus demonstrable by the classic method of breaking up the cell and inoculating the product into other susceptible cells. There is other evidence that at this stage the cells do not contain formed virus particles. Under certain circumstances this phase may be prolonged. Its existence can be recognized either by providing a situation where infective virus particles will develop and be identified by standard methods or by demonstrating specific virus antigen in or liberated from the cells. In general a virus (in the sense covering both the infective and vegetative phases) can be recognized only if it produces cytopathogenic effects in some type of host. It is by no means necessary, however, that intracellular