Bradley Monk, MD, FACOG, FACS: Those are the 2 validated biomarkers in cervical cancer. Let’s transition now to endometrial cancer. It’s my opportunity to discuss this. On May 23, 2017, the world changed. The world changed because we got the first FDA approval based on a biomarker. It was agnostic: 149 patients, 14 of whom had endometrial cancer. We were using drug along with it, and that allowed us to treat our patients with mismatch repair deficiency with pembrolizumab. Tell us about your experience with using testing for mismatch repair deficiency and using pembrolizumab in recurrent endometrial cancer in that setting.

Michael J. Birrer, MD, PhD: The testing for microsatellite instability [MSI] has been around for quite a while. It’s evolved over time, for good reasons. But it does present a certain amount of complexity. In that particular study you quoted, most patients—not all, but most patients—were subjected to the PCR [polymerase chain reaction] assay, which is looking actually at microsatellites and their variability. I think part of that was because the study had a predominance of colorectal cancer patients. The alternative assay, which I think a lot of use now and is in most PATH [pathology] departments, is actually staining for the 4 genes, right? And these are the mismatch repair genes. The absence of any 1 of them then leads to a diagnosis of mismatch repair deficiency.

It turns out that if you do a systematic analysis, if you just do PCR, you can miss mismatch repair. Because if there’s no tissue in there, no tumor, you miss about 5% to 6%. It also goes the other way, although I’ve not experienced that, that IHC [immunohistochemistry] tumor sits on the desktop, it’s not processed right, and you get bad staining. In this study most got PCR; some got immunostaining. I think throughout the country immunostaining dominates.

Bradley Monk, MD, FACOG, FACS: Are you doing IHC for the mismatched repair deficiency genes on endometrial cancer?

Thomas Herzog, MD: Standardly in the cross population.

Bradley Monk, MD, FACOG, FACS: Standardly. Krish, standardly?

Krishnansu S. Tewari, MD: Yup.

Bradley Monk, MD, FACOG, FACS: And then we’ve learned that the most common abnormality is MLH1. And when the MLH1 is deficient, you have to look for promoter hypermethylation. It’s an epigenetic silencing. And anything else probably should be referred for Lynch syndrome testing. So you have this opportunity to create a biomarker for therapy but also a biomarker for Lynch syndrome testing, unlike PD-L1 [programmed death-ligand 1] testing in cervical cancer in which you wait until they get metastatic disease. Here, we test early on because we’re testing also for Lynch syndrome. And you’ve had good experience with that?

Jubilee Brown, MD: Absolutely, and we’ve detected about 3% of our patients with Lynch syndrome.

Bradley Monk, MD, FACOG, FACS: That’s a good lesson. That 30% have mismatch repair deficiency, but probably only a tenth of those have lynch syndrome.

Thomas Herzog, MD: It can’t be emphasized enough: the importance of testing both in ovarian cancer and in the setting for uterine cancer.

Bradley Monk, MD, FACOG, FACS: Thank you.

Thomas Herzog, MD: Because it’s an opportunity to help guide treatment for this patient and inform on prognosis, but the cascade testing also saves lives. Finding out that these mutations exist in relatives where you can do risk-reduction strategies, including surgery, makes all the difference in the world. And Karen Lu showed a long time ago, but 50% of Lynch syndromes are actually diagnosed with the endometrial pathway. Everyone thinks of them as just being related to colon cancer.

Krishnansu S. Tewari, MD: Yeah. Testing is prognostic, helps guide therapy, and informs the risk for relatives but also informs the risk of other cancers—like colon cancer, breast cancer.

Bradley Monk, MD, FACOG, FACS: Sounds like BRCA, right?

Michael J. Birrer, MD, PhD: Yeah, and I think 1 other great unknown, which we don’t have enough data for, is what happens somatically with recurrence. We need more biopsies out there. Can you get a mismatch-repair abnormality later on in history of the disease?

Transcript Edited for Clarity