Matthew Meyerson discusses TCGA lung squamous cell cancer study

Color head shot of Matthew Meyerson, TCGA investigator, wearing glasses.

Matthew Meyerson, M.D., Ph.D.

Matthew Meyerson, M.D., Ph.D., a professor of pathology from the Harvard Medical School in Cambridge, Mass., and associate professor of pathology at the Dana-Farber Cancer Institute in Boston, led the writing committee for The Cancer Genome Atlas project on lung squamous cell cancer. He recently discussed his perspective on the findings.

NCI: Were there genes identified in the TCGA study that were unexpected, or was it more a matter that nobody has really interrogated the squamous cell genome that carefully heretofore?

Meyerson: I’d say the most unexpected finding was of some loss-of-function mutations in the HLA-A gene, which encodes a major histocompatability complex that plays an important immune regulatory role on the surface of cells. HLA-A codes for a protein on the surface of most cells that presents antigens to the immune system. So there’s been a concept that cancers evade the immune system, but this is the first evidence of somatic cancer genome alterations evading the immune system by changing their surface antigens. There were some papers published on models last year that showed this same concept in mouth cancers, but this is the first evidence in lung cancer.

NCI: Were there any other genes identified in the TCGA study that you think are noteworthy?

Meyerson: The HLA finding is new. Most of the other major genes that we’re identifying have previously been found, but this is the first comprehensive study of squamous cell lung carcinoma, and because of its comprehensive nature, what we’ve found is that there’s a gain-of-function mutation in many of the previously identified genes. This is quite important. A gain-of-function mutation in a targetable gene that codes for either a protein kinase or other type of kinase was found in more than 60 percent of the cases, and this might even be an underestimate. So this means that most cases of squamous cell lung cancer are now potentially treatable with targeted therapies—at least with investigational agents. I really think that this is going to catalyze a revolution in the treatment of squamous cell lung cancer. With lung adenocarcinoma, over the past decade or so, we’ve had a number of specifically targeted treatments introduced, and they are really advanced treatments for that subset of lung cancers. I think we’re going to see the same kind of development in squamous cell lung cancer.

NCI: There is an active cancer vaccine program at NCI, particularly in prostate cancer.  Does this finding suggest that we may also want to pursue vaccines in this form of lung cancer?

Meyerson: Absolutely. In fact, there were several clinical trials recently involving molecules that are known as immune system modulators. They’re basically antibodies that down-regulate the immune system. They are being used for exceedingly difficult cancers, and they’ve shown great promise in clinical trials for squamous cell lung cancer in particular. One of the trials is headed up by Julie Brahmer, M.D., at Johns Hopkins University.  It is one of the largest trials in this area.

NCI: Since we really didn’t know about HLA, were trials targeting surface antigens potentially the best approach, or should we rethink our vaccine approach?

Meyerson: No, those trials target antibodies that basically up-regulate the whole immune system. So there is no real specific targeting, but this TCGA finding suggests that a variety of targeting approaches might be suitable for squamous cell lung cancer.

NCI: How prevalent is squamous cell as a subset of lung cancers?

Meyerson: The mortality numbers for squamous cell lung cancer and lung adenocarcinoma aren’t exactly clear, but approximately 30 percent of lung cancer deaths are due to squamous cell lung cancer. There are approximately 150,000 to 160,000 lung cancer deaths per year in the United States. If you do the math on those two numbers, it gets you somewhere in the 40,000-plus range for deaths per year for squamous cell lung cancer in the United States alone, and that’s more than breast cancer, more than prostate cancer. It’s more than colorectal cancer; it’s more than pancreatic cancer. If you actually look at squamous cell lung cancer on its own, it would be the second-leading cause of cancer death after lung adenocarcinoma.

NCI: In terms of cancers related to smoking, squamous lung cancers congregate more in the center of the lung, whereas adenocarcinomas are more peripheral. Does this finding tell us anything about the components in cigarette smoke that may cause the subtypes of lung or how they develop?

Meyerson: Not yet. We haven’t learned a lot about the smoking signature from our study. There’s a distinct pattern of mutations that are associated with cigarette smoking, and we see that pattern of mutations in squamous cell lung carcinoma. Perhaps other researchers will take the information that we’ve generated, to gain more insight into the mechanisms by which cigarette smoking affects squamous cell lung cancer.

Group of scientists comprised of 28 men and four women, pose for a photo.

Some of the members of TCGA lung cancer disease working group

NCI: We have done a number of Genome-wide Association Studies, or GWAS, to get at some potential mechanisms or associations by which smoking initiates or promotes lung cancer.

Meyerson: There’s one interesting result of a genome-wide association study that recently came to light. The investigators found several loci on the HLA-A gene associated with squamous lung cancer.

NCI: Does this finding tell us that we may need a multiple number of drugs to hit a multiple number of targets in squamous lung cancer?

Meyerson: Absolutely. For squamous cell lung cancer, we will need multiple drugs for multiple targets, and I think the picture is similar to lung adenocarcinoma, although the distribution of targets is very different. For example, for the epidermal growth factor receptor, or EGFR , which is mutated in about 10 or 15 percent of lung adenocarcinomas for patients in the U.S., and nearly half of lung adenocarcinomas in Asia, it’s treatable with particular drugs or drug combinations. In squamous cell lung cancer, it was actually thought that wasn’t a good target, but actually one surprise is we did actually find EGFR mutations in about 1 percent of the cases. So while it’s a lot lower frequency, it is a different spectrum of EFGR mutations, but they are there and that should actually give us a target to hit in the future.

NCI: In terms of treating this particular form of lung cancer, how many drugs do you get a sense that will have to be created de novo versus those that we could take off the shelf and find some real use for?

Meyerson: I think that a lot of the drugs that could be useful are currently in clinical trials. These include fibroblast growth factor receptors, or FGFR inhibitors, and PI3-kinase inhibitors. That means, of course, we don’t know yet how they’re going to perform. I think there are going to be other drugs that will include many of the classes of enzymes that we’ve found to be mutated in adenocarcinomas. Most cases of squamous cell lung cancer demonstrate a loss of a type of tumor suppressor gene called CDK, or cyclin dependent kinase genes.

There is a loss of this tumor suppressor, so you can imagine that if you actually used cyclin dependent kinase inhibitors that you might be able to treat tumors with this loss.  This would be an argument that we don’t know what drugs in clinical trials will be effective, but that we should certainly test these drugs in squamous cell lung cancers.

Because there are drugs now in clinical trials, I would view our results as not leading to an immediate change in the practice of squamous cell lung cancer, but rather leading to a change in what clinical trials we can do to improve the care of squamous cell lung cancer.

NCI:  This was the first TCGA study to incorporate whole genome sequencing, although I think just in 19 samples.

Meyerson: This is the first one to incorporate whole genome sequencing events in cases. I actually think full genome sequencing is a challenge right now. I’m expecting that the whole genome sequencing data that we’ve provided are going to be incredibly useful for discovery. I don’t think we’ve actually captured all of the important observations in our paper, and my expectation is that cancer genome projects around the world will be working at the whole genome sequence data generated by TCGA, and that a lot of the most important discoveries may be made by re-analyses by other scientists over the next couple of years.

I think we’re going to learn more and more. I think we can understand a lot about genome alterations in the coding sequences of genes, but we really don’t understand much at all about the genome alterations of the non-coding sequences.

I think this has been a really strong collaborative effort across many, many, many institutions, and I think that one impact of that is that the participation of many groups in many institutes are in a way stimulating lung cancer research throughout the U.S., along with our partners from Canada, throughout North America, even throughout the world. The other thing I should mention is there are several other systematic one-character genome papers coming out over the next few weeks that will add to our knowledge base.

Finally, TCGA has also started a group that’s called the pan-cancer analysis group, and I think it is now possible to start looking at what we’re learning in all the different types of cancers and putting that together.

NCI: So it’s no longer a physiological examination? It’s looking for genetic markers that may have commonalities in a lot of different cancers?

Meyerson: Right. Commonalities, differences; just being able to analyze across cancers. I think that’s mostly commonalities to give us kind of an understanding so we’re amassing the dataset that makes that possible.

NCI: The final question is really a big picture question in terms of TCGA and the pace of discovery. As this is the fourth comprehensive report from TCGA, what has this effort shown us so far about how we can work globally with other partners in terms of really accelerating genomic research?

Meyerson: What I would say is after ramping up in the pilot phase, I think TCGA is really proving to be a very effective project, and as you mentioned, this is the fourth comprehensive report with colorectal, squamous cell lung cancer, and breast all coming out over the past or next few months. Within the next year I think we can expect kidney cancer, endometrial cancer, lung adenocarcinoma, head and neck cancer, maybe melanoma, and, I think, bladder cancer. So probably another six or seven reports in the next year, and these are providing both the discovery set and a reference data set that cancer researchers will be able to use to advance cancer discovery over the next few years.

Print This Post Print This Post

Comments are closed.