To the Editor: Ang et al. (July 1 issue)1 developed a risk model for death from oropharyngeal squamous-cell carcinoma on the basis of human papillomavirus (HPV) status, pack-years of tobacco smoking, tumor stage, and nodal stage. Their results showed that the risk of death significantly increased with each additional pack-year of tobacco smoking. It would be interesting to include in the statistical Cox regression analysis the pattern of alcohol consumption. Recent studies2 have indicated that alcohol consumption has a role in the genesis of head and neck cancer in HPV-positive patients, showing consistency with the observation and hypothesis (reinforced by the study by Ang et al.) that HPV infection is not a sufficient cause of cancer and that it requires the accumulation of additional cellular changes.