Approaches to immunological control of virus-associated tumors in man: introductory remarks.

The rapidly accumulat ing evidence for the role of viruses in the et io logy of human cancer makes the topic of antiviral agents, such as vaccines, of great interest to tumor immunologists. Since direct proof of oncogenic i ty in humans is impossible, however, a discussion of the means of controlling human tumors that are potential ly virus induced must be speculative and draw primari ly on observations taken from apparent ly relevant animal models. Three groups of viruses, known to cause tumors in animals, have also been suspected of causing cancer in man: the DNA herpesviruses, the "Ctype" RNA viruses, and the "Btype" RNA mammary t umor viruses. Other DNA viruses, particularly the papovaviruses (47), are also demanding attention as oncogenic agents. Because of the relatively little information we have on the biology of the candidate tumor viruses, in v i t ro as well as in v ivo, the risks inherent in vaccinating normal individuals w i th live attenuated virus are currently considerable. Killed virus, a possibil i ty in spite of the problems of defining adequate bioassay systems for residual infectivity, is a possibi l i ty but shares with subviral fractions the potential of providing only short-term immunity. However, as immunoepidemio log ica l studies demonstrate "periods of high risk," wh ich are common in animal tumor systems and logically are present in tumors of chi ldhood, the util ity of such vaccines should not be completely ignored. In view of the concerns regarding vaccination of normal individuals w i th potential ly oncogenic materials, the use of antiviral vaccines at the present t ime would seem to be far more appropriate as part of an immunotherapy regimen in patients already affl icted with the disease. The accumulating evidence that persisting susceptibi l i ty, perhaps genetically related, maintains certain individuals at high risk to cancer indicates that "relapse" in a number of cases may indeed be re induct ion of disease. Three examples of human tumor site and is very responsive to retreatment with the BL, 2 acute lymphocyt ic leukemia, and breast cancer. In BL, Ziegler has described 2 types of relapse, early and late, and has provided clinical evidence suggesting that late relapse is actually disease reinduction (59). His studies demonstrate that late relapse usually presents in sites other than the 1st tumor site and is very responsible to retreatment with the drugs used at disease onset, whereas the early relapse is resistant to the drugs used initially and local disease recurrence is the rule. Presuming that the late relapses (particularly those fo l lowing unmaintained remissions of 6 months)