Transcript:Robert L. Ferris, MD, PhD: Head and neck cancer is a disease comprised of several different types. One is due to environmental carcinogens, such as tobacco smoke and heavy alcohol use, and an increasing subset is driven by human papillomavirus, or HPV. And so, like lung cancer, HPV-negative head and neck cancer is caused by carcinogens that induce DNA damage, tumor suppressor gene alterations, and a large number of genetic alterations. The other way to get head and neck cancer is infection by human papillomavirus, and that’s a different subset that was discovered 30 or 40 years ago. It has been increasing in frequency, and has different biological and clinical characteristics.

Head and neck cancer is a disease with a large number of genetic abnormalities within the tumor cell. When we think of tumor drivers or driver mutations, we think of oncogenes that have activating mutations. Although those occur to some degree in head and neck cancer, the majority of the genetic alterations in head and neck cancer cells are to tumor suppressors. P53 is the most commonly altered genetic alteration, but also we see some activating mutations in the PI3 kinase and other tumor suppressors such as the Notch pathway, p16, and others.

Additional genetic drivers in head and neck cancer come from the virus oncogenes, from human papillomavirus virus (HPV). HPV E6 and HPV E7 are other genetic drivers, but they come from outside the cell as a virus infects the cell and turns it into a head and neck cancer cell.

Ezra Cohen, MD: When we think about head and neck cancer and its molecular biology, we’ve come to realize many different aspects related to this disease. One of them is mutational burden. And, of course, we’ve now come to recognize that mutational burden has an impact and implications for the application of immunotherapy. We’ll talk about that in just a few minutes. But, let’s talk about what mutational burden means in the context of head and neck cancer. We normally define it by some measure of the number of mutations per base pair—usually it’s per megabase—and we get a number, depending on the cancer, of somewhere in the hundreds. For high-mutational burden cancers, we usually talk about numbers in the several hundred. For melanoma, we’re even talking about possibly 1000. For cancers that we know have a high mutational rate, like microsatellite in stable cancers, again we’re talking about many hundreds or 1000. For head and neck cancer, we’re really dealing with a moderately mutated cancer.

What’s interesting is that we always thought that for HPV-negative cancers—typically tobacco related, alcohol related, traditional risk factors—the mutational load for those cancers would be significantly higher than for HPV-positive cancers. Well, it turns out that that is only somewhat true. There are some HPV-negative cancers that have very high mutational loads—again, in the several hundred to even close to 1000. What we do see, though, is that HPV-positive cancers are also moderately mutated with mutational loads usually in close to 100, or a little bit over 100. That is not unusual for HPV-positive cancers.

So, we used to think that the HPV-negatives would have a much higher mutational burden than HPV-positives. That turns out, on the average, not to be true. There are some HPV-negative cancers with a very high mutational burden that we don’t see in the HPV-positives, but on average the 2 appear to be about the same. And the reason that the HPV-positives have a higher mutational burden than we thought was because some of the alterations in HPV-positives actually disrupt the DNA repair or the translational machinery, resulting in a cycle that causes more mutations.

Transcript Edited for Clarity