Analysis Finds Early Endometrial Cancer Mutations Detectable in Uterine Lavage Fluid

John A. Martignetti, MD, PhD, discusses The Cancer Genome Atlas' comprehensive, genomics-based analysis of the endometrial cancer.

John A. Martignetti, MD, PhD

According to John A. Martignetti, MD, PhD, it is projected that, by 2030, endometrial cancer will become the third leading cancer for women. As mortality rates continue to climb, there remains a pressing need for effective screening and diagnostics.

The Cancer Genome Atlas (TCGA) carried out a comprehensive, genomics-based analysis of the disease, in which they revealed many of the molecular defects that define this cancer. Based on those findings, Martignetti and his colleagues hypothesized that ultra-deep, targeted gene sequencing may be able to detect somatic mutations in uterine lavage fluid obtained from patients undergoing hysteroscopy as a means of molecular screening and diagnosis.

In an interview withTargeted Oncology, Martignetti, an associate professor of Genetics and Genomic Sciences at the Icahn School of Medicine at Mount Sinai, and network director for the Laboratory for Translational Research at the Western Connecticut Health Network, discussed this novel genomic analysis and what impact the results could have in the detection, screening, and diagnosis of endometrial cancer.

TARGETED ONCOLOGY:What was the rationale for conducting this genomic analysis?


If you think about endometrial cancer, it’s a cancer where there are 55,000 cases a year. And the incidence is actually increasing, so it’s projected by 2030, endometrial cancer will probably have the same number of cases as colorectal cancer for US women. It’ll become the third leading cancer site in women by 2030.

What really has not kept pace with those increasing numbers is any way to diagnosis it. Symptoms can include nonmenstrual bleeding, postmenopausal bleeding, pain, weight loss, back pain, things like that. But really, there’s no clinically useful molecular markers or diagnostics that are currently available. And really, the only way to make a diagnosis, the gold standard way, is to bring a patient into the OR, and do a hysteroscopy, and that really has not changed.

Clearly, to bring the patient into the OR, it’s a big deal because you’re talking about the pain, the cost, the time, the risks of general anesthesia, and then you’ve got issues of tissue sampling, depending on the age of the patient, whether there’s any structural abnormalities, and ultimately, the final interpretation of whether that patient has endometrial cancer or not. It’s really a subjective test in the sense that it’s a section of tissue sent for histology. There aren’t any special stains, no special markers; it’s really a subjective call. And depending on where you are, you may be blessed with a great pathologist, or maybe that’s something you don’t have adequately available.

But there are 2 other interesting forces here. One force is essentially, if you think about a pap smear, they were devised about 70 years ago, and the concept is to touch the cervix and look for cytologic changes that would suggest someone has cervical cancer. Since that test was devised, people have been trying to do a similar thing for endometrial cancer. For a number of reasons, it’s just never really worked out well.

A couple of years ago, some people came up with a clever test of using a tampon-like device to look for genetic changes that could be associated with endometrial cancer. It was fairly good, but it wasn’t great. Earlier this year, another group actually took patients with endometrial cancer and did this lavage procedure that we’ve done in our own study and showed that patients with endometrial cancer did have genetic changes that were discoverable in this lavage.

The other force that made this study timely was, a couple years ago, TCGA sequenced endometrial cancers, and from their sequencing of more than 400 patient samples, they showed that genetic changes within the tumors could predict outcome, and there were specific genes that were mutated in these cancers.

TARGETED ONCOLOGY:And what specifically did you do in your study?


What we did was different from what’s been done before. We took 102 women who were being diagnosed—in other words, their physicians were suspicious that the patients might have endometrial cancer because the women may have had postmenopausal bleeding, pain, they might have had ultrasound findings of a mass, but couldn’t sort out what it was exactly, but there was a suspicion that there could be cancer there, and they needed to get a hysteroscopy.

A hysteroscope is a long steel tube that has a light source and a camera at the end, and also some ports that you flush saline out of, and then on the other end is the operator to see what’s in there, allowing the surgeon on the outside to go into the uterine cavity and see what that cavity looks like. When you do that, the uterus is kind of like a collapsed sac, and the surgeon will blow saline in there just to open the sac up, and what we asked for was essentially that saline that’s just dripping out. We asked for a collection of that because we figure, that saline is coating the entire cavity. Can we look at that, bring it back to the laboratory, isolate the DNA in there, and sequence it?

TCGA did 2 very intense and very expensive studies, called whole-genome sequencing, where they sequenced every gene that was there in the tumor, and they did whole-exome sequencing, where it’s a fraction of the whole genome, but it’s still a very expensive look at essentially 23,000 genes in the body.

Instead of doing that, because we have the TCGA data, we asked our biostatisticians, “Hey guys, we can’t afford to do $10,000 or $20,000 worth of sequencing on 100 women. But what we could do is, if you could tell us that this is possible, could you find a subset of genes that we can sequence very deeply, with very high coverage, that would give us the same information as doing an entire genome?”

And that’s what they did. We actually have 2 panels here: 1 panel with 12 genes, and 1 larger panel with 56 genes. So instead of doing 23,000 genes, we did 12 or 56, and we used that to sequence the material we pulled out from these lavages. And the study was set up so that, a patient would come into the OR, we would collect the sample, go to the laboratory, the surgeon who is trying to determine the diagnosis does everything he or she would normally do, collect tissue, send it up to pathology, and then they get a result. At the end, when we finished all of our molecular analyses, we went back to the pathologist and said, “Hey, how do our results compare with yours?” In other words, we have mutations we identified; are these the patients who have endometrial cancer?

So from those 102 women, it turned out that 7 women had endometrial cancer. Six of those women had stage Ia disease—the earliest microscopic cancers. You’re talking essentially about tumors that are about 2 or 3 millimeters in size. If you look at your ballpoint pen, it’s around the size of that. If you think about it, that was found in the uterus, and we were able to find mutations in that lavage just coming from that tip of the tumor. We could show that the genetic approach was very sensitive to identify these mutations, and that had never been done before, to prospectively diagnose cancer in these patients using a genetic test. That was really exciting.

But if that were the whole story, what we would be talking about is a screening test, meaning taking women at a certain stage in their lives, and potentially using something like this to see if we could find endometrial cancer. But when we looked at all of our results, and we looked at the other women who were not diagnosed with endometrial cancer, it turns out that half the women in our study also had these genetic mutations. And that was really surprising because, again, these genes that we’ve put on our 2 panels, these are really well established cancer genes, such that when you look at a cancer, having these mutations are very strict identifiers of cancer.

But in these women, we found these mutations, and they did not have a clinical diagnosis of cancer. So we thought, “Wait a second, are these women maybe developing early cancer? We’ve picked them up before they developed cancer? Or is there something else going on in the biology that we need to reevaluate, and say, maybe, even though these are cancer mutations, it’s not the same as having cancer?”

You would think that because, historically, if you take 100 women to the OR for a suspicion of endometrial cancer, 5% to 10% of those women, wherever you go in the world, will have cancer. It’s not 50%; that would be really high and very unexpected. Again, based on the historical record, maybe in 1 or 2 of these women, we might have picked up a premalignancy.

TARGETED ONCOLOGY:Are there any next steps planned following the results of this analysis?


We’re monitoring these women from the original study to really see what evolves in them. This is just research right now, and the results were kind of unexpected. So we’re closely following these women who didn’t have the mutations, and the 50 who did, we want to see what happens next.

We’re also doing a larger study now with 1000 women, and we’re going to be looking across multiple institutions to make sure there’s no bias related to location. And we’re going deeper with the sequencing to add more tests to the original one to see if we can still develop the screening or diagnostic test and really learn about the biology.

TARGETED ONCOLOGY:What effect do you think a novel test like this could have on the treatment landscape down the road?

We have to be careful in applying all of these really powerful genomic technologies to screening and diagnostics. For example, here in endometrial cancer, we identified this landscape of previously unknown mutations. So if you were to just kind of go off and do this sequencing, you’d be telling half the patients who had this test that they have cancer. And that’s obviously not true. There’s something else going on there.

In general, when we think about screening and diagnostics using these powerful technologies, we’ve got to understand the underlying disease better, first off. Second, if, indeed, 50 women have these cancer-driving mutations, why aren’t they developing cancer? Is there something that we don’t know or don’t appreciate yet about what drives women to develop endometrial cancer? Is there something, a push that hasn’t happened, or is there some type of a block that’s occurred? If we’re seeing this in endometrial cancer, might this be going on in other cancers that we just haven’t explored yet?

Reference:Nair N, Camacho-Vanegas O, Rykunov D, et al. Genomic analysis of uterine lavage fluid detects early endometrial cancers and reveals a prevalent landscape of driver mutations in women without histopathologic evidence of cancer: a prospective cross-sectional study [published online December 27, 2016].PLOS Medicine.