In 1990 in the United States, it is estimated that 30,200 people will develop oral cancer and that 9,050 will die of the disease (Silverberg and Lubera, 1988). This population of oral cancer patients represents approximately 5% of the total cancer population in the United States. Based on statistics collected between 1979 and 1984, the five-year survival rate for oral cancer patients is 54% for caucasians and 31% for blacks (Silverberg and Lubera, 1988). Oral cancer is an important problem not only because of the significant mortality associated with the disease, but also because of the often disfiguring and functional defects associated with the disease. Substantial epidemiologic research has investigated major factors in the development of oral cancer. Tobacco and alcohol use are the two major risk factors. Studies have proposed a direct dose- related carcinogenic effect of tobacco, a close correlation between alcohol intake and cancer, and the possibility of a synergistic effect of tobacco and alcohol when used together (Rothman and Keller, 1972, Wynder and Hoffman, 1976, Bross and Coombs, 1976, Schottenfeld, 1979, Spitz et al., 1988). Nonetheless, oral cancer develops in a significant proportion of individuals who do not use alcohol or tobacco, suggesting the importance of other factors. Herpes simplex virus type 1 (HSV-1) has been implicated in the etiology of oral cancer. Patients with oral leukoplakia have increased cell-mediated immune responses to HSV-1, smokers have higher levels of neutralizing antibody to HSV-1 than do non- smokers, and patients with oral cancer have higher levels of IgA and IgM antibody to cells infected with HSV-1 (Smith et al., 1976, Shillitoe et al., 1984). During the last few years, evidence has been provided that human papillomavirus (HPV) can infect the oral mucosa and is associated with several different benign epithelial tumors and hyperplasias of the oral cavity. There are now in excess of 50 genotypes of HPV, a family of viruses that infect only keratinocytes and are generally associated with benign epithelial proliferations. In addition, substantial evidence supports the hypothesis that HPV may be responsible for some of the changes leading to malignant growth. Although HPV cannot be cultured in vitro, the genomes of most of the types have been cloned allowing for studies of genome structure and gene function. The general organization of the papillomavirus genes is conserved among the types and all putative protein-coding sequences are on one DNA strand. The early (E) segment of the DNA contains up to eight open reading frames (ORF) and the late segment (L) contains two ORFs (Danos et al., 1984). Studies of the functions of the HPV genes have been modeled after those done with bovine papillomavirus (BPV). The E1 ORF is necessary for maintenance of HPV as episomal DNA (Saver et al, 1984, Lusky and Botchan, 1985). E2 is the most conserved HPV gene and has a regulatory function. The full-length E2 protein serves to transactivate early gene expression by binding to the transcriptional enhancer in the non- coding region (Phelps and Howley, 1987). A smaller protein transcribed from the 3' end of E2 acts as a repressor and competes with the transactivating protein for binding in the non- coding region (Lambert et al., 1987). The E4 gene appears to be involved in virus maturation (Doorbar et al., 1986). Both E5 and E6 gene products have transforming ability (Schiller et al., 1986, Yang et al., 1985). The E7 gene is important in the establishment of a high copy number of viral DNA in infected cells and may have a role in maintaining the transformed state (Lusky and Botchan, 1985). The L1 and L2 ORFs code for the capsid proteins of the virus (Giri and Danos, 1986). HPV DNA has been demonstrated in oral verruca vulgaris, condyloma acuminatum, squamous cell papilloma, leukoplakia, and focal epithelial hyperplasia (reviewed by Syrjanen, 1987). A possible role of HPV infection in the etiology of oral cancer was suggested in 1983 when cytopathic changes characteristic of HPV infection were described in 14 cases (Syrjanen et al., 1983). This same study found HPV capsid antigens in eight oral squamous cell carcinomas. With the availability of DNA cloning and hybridization techniques, more direct evidence for the presence of HPV in oral cancer has been provided. Loning et al. found HPV DNA in three cases of carcinoma of the buccal mucosa; one lesion had HPV 16 DNA, one HPV 11, and one was not typed (Loning et al., 1985). A follow-up study showed HPV 16-related DNA in four of seven oral cancers (Milde and Loning, 1986). A similar study by de Villiers et al. found HPV 16 DNA in two cases of squamous cell carcinoma of the tongue and HPV 2 DNA in one case (de Villiers et al., 1985). There are other isolated reports of HPV 16 in a single buccal mucosa carcinoma and a tongue carcinoma (Syrjanen et al., 1986, Lookingbill et al., 1987) and of HPV 11 in a lymph node metastasis of an oral squamous cell carcinoma (Dekmezian et al., 1987). Studies of larger numbers of oral cancers found HPV 2 DNA in three of nine cases of verrucous carcinoma (Adler-Storthz et al., 1986), Maitland et al. (1987) found HPV 16-related DNA in seven of fifteen carcinomas, and Syrjanen et al. (S. Syrjanen, personal communication) found HPV 16 in three cases and HPV 18 in three cases of carcinoma. The most substantial data implicating HPV in the etiology of epithelial cancers comes from studies of cervical cancer where evidence of HPV infection has been found in the majority (up to 90%) of cases examined (reviewed by Syrjanen, 1987). The most severe lesions are associated with HPV types 16 and 18, whereas the dysplastic, less severe lesions are associated with HPV types 6 and 11. In most cases, the HPV 16 and 18 DNA is integrated into the host cell genome, whereas the genomes of HPV 6 and 11 appear to persist in an episomal state. The progression from precancer to cancer in the uterine cervix is well established and, in HPV-positive cases is thought to reflect the ability of the HPV to integrate into the host cell genome (Syrjanen, 1987). Integration may not be required for transformation in all papillomavirus species. In some tumors and cell lines that contain HPV DNA in an integrated state, the E6 and E7 genes appear to be selectively maintained and expressed and the E2 gene is interrupted, partially deleted or absent (Takebe et al., 1987, Shirasawa et al., 1986, Schwarz et al., 1985, Choo et al., 1987). Several studies have shown that HPV 16 can transform established mouse and rat fibroblast lines as well as primary human fibroblasts and keratinocytes (Kanda et al., 1987, Yasumoto et al., 1986, Prisi et al., 1987). More recent in vitro culture models have shown the transforming functions to reside in the E6 and/or E7 genes (Bedell et al, 1987, Kanda et al., 1988). It can be postulated that as a result of integration and loss of the E2 repressor function, the expression of the E6 and E7 proteins in an unregulated manner leads to the development of the malignant phenotype. Other co-factors probably act synergistically with HPV in tumor induction. For example, in BPV-4-infected cattle that feed on bracken fern, which contains a protent mutagen, alimentary tract carcinomas develop from benign papillomas (Jarrett et al., 1978). HSV is a co-carcinogen that could interact with HPV (zur Hausen, 1982). In humans with epidermodysplasia veruciformis, exposure of the skin to sunlight is associated with the development of skin carcinomas in which HPV types 5, 8, and 14 are found (Pfister et al., 1981, Claudy et al., 1982). Immunosupression may also be a cofactor as evidenced by the finding of increased numbers of benign and malignant HPV-positive tumors in renal transplant recipients (Rudlinger et al., 1986, Van der Leest et al., 1987). The mucosal epithelium in the oral cavity is histologically similar to that of the female genital tract and like the female genital tract, it is continuously exposed to various environmental factors such as irritants and microorganisms. Thus there may be parallels in the oral cavity with respect to the capacity of HSV and HPV to participate in malignant transformation most likely in association with other cofactors. Clearly, there is a need to examine a large number of oral carcinomas to ascertain the frequency of infection with the more oncogenic HPV types and to show any correlation between HPV infection and progression from premalignancy to malignancy. It is also very important to study the relationship between the presence or absence of human papillomavirus in oral tumors as it relates to the prognosis for the patient. oral cancer. In addition, if HPV is found in histologically normal mucosa from oral cancer patients It will be most useful to know if the presence of the virus effects the tumor-free interval following treatment, the incidence of recurrent tumors, the incidence of second primary tumors, and the five-year survival rate. The virological data can be analyzed in the context of other risk factors such as alcohol and/or tobacco use to look for possible associations. By examining an appropriate number of samples of histologically normal oral mucosa, it will be possible to determine if virus infection is a potential risk factor for development of following treatment, it will be important to document if these patients have more recurrences of the disease. At the molecular level, information is needed regarding the physical state of the viral DNA in oral carcinomas since integration of the viral DNA into the host cell genome is thought to be an essential step in the transformation process. Information regarding which viral genes are retained in oral carcinomas and expressed as mRNA and as protein would be most useful in defining the role HPV may play in events leading to oral cancer.