top of page

Oral cancer poses a serious public health burden

Squamous cell carcinomas (SCCs) are among the most common cancers in the world. While non-melanoma skin cancers, including basal cell carcinoma and SCC, will affect nearly half of all Americans at some point in our lives, these common cancers can be easily managed with surgery, and rarely become invasive or metastasize. On the other hand, many SCCs, including those that affect the lung and esophagus, have much poorer prognoses. The less commonly studied oral squamous cell carcinomas (OSCCs) are the 6th most common cancer worldwide. Patient survival has not improved much in 30 years, and half of all patients succumb to the disease within 5 years. Few personalized therapies exist, and surgery followed by chemo-radiation remains the standard of care.

Genome sequencing reveals common alterations in differentiation genes in oral cancer

Efforts such as The Cancer Genome Atlas (TCGA, cBioPortal) have sequenced tens of thousands of tumor genomes, providing insights into the common and unique pathways that may promote cancer initiation and progression in specific patients and diseases. The TCGA study of oral cancers revealed that defects in differentiation genes such as NOTCH1 are surprisingly common, affecting more than half of all sequenced tumors (TCGA, Nature 2015). This demonstrates that the balance between self-renewal and differentiation is a frequently disrupted in oral cancers. Moreover, we have found that patients with alterations in differentiation genes show poorer prognosis than those without, and alterations in spindle orientation genes, though rare, correlate with extremely poor patient survival.


Developing mouse models of oral cancer

Over the past several years, we have been developing genetically-engineered mouse models (GEMMs) of oral cancer. Importantly, these models have induced tumor formation in the native site, the oral cavity, rather than in a surrogate tissue such as the epidermis. Our GEMM models include tissue-specific Cre-inducible models of human papilloma virus (HPV)-driven and p53-driven oral cancers. While high-risk HPV infection is most commonly associated with cervical cancers, it is also associated with a growing proportion of oral cancers, and as many as half of all patients admitted to UNC hospitals are HPV+. To model this disease subtype, we are collaborating with Antonio Amelio (Amelio lab @ UNC) on a new mouse model driven by the HPV oncogenes E7 and E6.

On the other hand, as many as 80% of oral cancer cases show alterations in the so-called “guardian of the genome,” p53, encoded in humans by the gene TP53. In oral cancer, as in many other cancers, TP53 is frequently mutated but rarely lost. Missense mutations are common, and many of them occur at “hot-spots” in the DNA binding domain that mediate interactions between p53 and its target genes. It has long been known that these hotspot mutations are in fact “gain-of-function,” and recently evidence suggests that they may induce the acquisition of “neomorphic” p53 targets such as epigenetic regulators, rather than just the simple loss of normal p53 targets. Thus, we hypothesize that hotspot p53 mutations confer unique changes to the stem cells in oral epithelia, which may contribute to their chemo-resistance and poor patient prognosis.


Ongoing questions being addressed in current studies include:

Which genes’ expression is altered by p53 mutation, and how does this compare to p53 loss?

How is division orientation impacted by p53 loss and HPV infection?

Does compromising the spindle orientation pathway alter tumor progression?

Why do women with oral cancer show much worse prognosis than men, in contrast to every other non-reproductive cancer?

bottom of page