Cancer is a disease caused by the accumulation of genetic and epigenetic changes in two types of genes: tumor suppressor genes (TSGs) and proto-oncogenes. Extensive research has been conducted over the last few decades to elucidate the role of TSGs in cancer development. In cancer, loss of TSG function occurs via the deletion or inactivation of two alleles, according to Knudson’s two-hit model hypothesis. It has become clear that mutations in TSGs are recessive at the level of an individual cell; therefore, a single mutation in a TSG is not sufficient to cause carcinogenesis. However, many studies have identified candidate TSGs that do not conform with this standard definition, including genes inactivated by epigenetic silencing rather than by deletion. In addition, proteasomal degradation by ubiquitination, abnormal cellular localization, and transcriptional regulation are also involved in the inactivation of TSGs. This review incorporates these novel additional mechanisms of TSG inactivation into the existing two-hit model and proposes a revised multiple-hit model that will enable the identification of novel TSGs that can be used as prognostic and predictive biomarkers of cancer.