How does human papillomavirus contribute to head and neck cancer development?

Human papillomavirus (HPV) is the major etiologic factor in the development of cervical cancer (1) and has been the target for detection and prevention of the disease (2,3). Results of recent molecular and epidemiologic studies suggest that HPV may also be an etiologic factor in a subset of head and neck squamous-cell carcinomas (HNSCCs), particularly those that develop at oropharyngeal sites (4–6). In this issue of the Journal, Braakhuis et al. (7) report the striking finding that HNSCCs with active HPV type 16 DNA (i.e., HPV16 DNA that expressed the viral E6 and E7 genes) had substantially lower rates of loss of heterozygosity (LOH) at chromosomal regions 3p, 9p, and 17p than tumors that contained inactive HPV DNA (i.e., HPV DNA that did not express E6 and E7). This finding did not extend to LOH at chromosomes 13q, 18q, 8p, and 6q (7). Not surprisingly, Braakhuis et al. (7) also observed a negative association between active HPV DNA and mutations in TP53, the gene encoding the p53 tumor suppressor, as has been reported by other groups (5,8). Because the chromosomal regions analyzed in the Braakhuis et al. study contain tumor suppressor genes and are commonly deleted in HNSCC, it is believed that LOH at these regions plays a role in head and neck tumorigenesis. The striking finding, therefore, suggests that HPV16 may play a role similar to those of the tumor suppressor genes located at the three chromosomal regions. In cervical cancer, the retention of HPV16 DNA and continued expression of viral E6 and E7 genes are required to maintain the malignant phenotype of the tumor cells (9). The finding by Braakhuis et al.—that only HNSCCs that expressed the HPV16 E6 and E7 genes showed distinct patterns of LOH but not the tumors that carried inactive HPV16 DNA (7)—supports the notion that active HPV16 is required in a subset of HNSCC to maintain the malignant phenotype. The lack of TP53 gene mutations in any of the tumors with active HPV16 in this study, and the observation that tumors with inactive HPV16 DNA had a similar TP53 mutation frequency as tumors that lacked HPV DNA, further underscore the importance of continued activation of HPV16 in maintaining a malignant phenotype. It is interesting that deletions at 3p and 9p, two of the three chromosomal regions that lacked LOH in tumors with active HPV16 E6 and E7, are among the earliest abnormalities detectable in head and neck tumorigenesis (10,11). It is possible that a lack of LOH at these regions reflects the absence of exposure to etiologic factors that cause DNA damage and genomic instability. The fact that many HNSCC patients whose tumors have active HPV DNA do not smoke or drink alcohol (5) supports this notion. Alternatively, HPV infection could occur before the deletion events. In this case, if HPV contributes to tumorigenesis through pathways that are similar to those resulting from the loss of tumor suppressor genes located at 3p and 9p, deletion of these chromosomal regions would no longer be required for tumorigenesis. The 9p region analyzed by Braakhuis et al. (7) harbors