Capsule-Sponge Testing Could Help Close the Barrett’s Screening Gap: An Expert Q&A

Esophageal adenocarcinoma (EAC) incidence in the US has risen approximately 5-fold over the past 4 decades, even as rates of other solid tumors such as colorectal and breast cancer have declined. Screening for Barrett’s esophagus (BE), the precursor lesion, is the main strategy for early detection. Yet an estimated 70% to 90% of patients who meet American College of Gastroenterology (ACG) criteria for screening never undergo it. Because screening depends on sedated upper endoscopy, uptake has stayed low. Non-endoscopic cell collection devices offer a possible way to move screening into primary care.

The DETECT-ME study was a prospective, multicenter cohort study presented as a late-breaking abstract at the ACG 2026 Annual Scientific Meeting. It evaluated a novel DNA methylation biomarker panel applied to cells collected with the EndoSign capsule-sponge device. Nicholas J. Shaheen, MD, MPH, MACG, of the University of North Carolina at Chapel Hill School of Medicine, is the study’s senior author. Investigators at 18 sites enrolled 2 groups: patients at high risk for BE/EAC undergoing screening endoscopy, and patients with known BE presenting for surveillance or endoscopic eradication therapy. All participants underwent capsule-sponge sampling before esophagogastroduodenoscopy (EGD).

Among 324 patients in the primary analysis, the test achieved a sensitivity of 90% (95% CI, 84-94) and a specificity of 86% (95% CI, 79-91). Sensitivity was 99% (95% CI, 94-100) for long-segment BE and 100% (46/46) for patients with any dysplasia or EAC, compared with 82% (95% CI, 73-89) for short-segment BE. Sample collection succeeded in 95% of participants, with no device-related adverse events, and 81% said they were willing to repeat the procedure.

Below, Shaheen discusses the gaps behind low screening rates, how the capsule-sponge test works, the tolerability data, and how the approach could fit into primary care workflows.

Patient Care Online: Why do so many at-risk patients for Barrett’s esophagus and esophageal adenocarcinoma go unscreened today, and what gap are non-endoscopic approaches like capsule-sponge testing trying to fill?

Shaheen: Our attempts at stopping esophageal cancer by screening have largely failed. In the US, somewhere in the neighborhood of 70% to 90% of those who are eligible for screening never undergo screening. By ACG criteria, around 13% of the US population is eligible for screening, and the fact that we’re missing so many of them explains why we haven’t changed the trends for this cancer. Other solid cancers, like colon cancer and breast cancer, are trending down in the US. Esophageal adenocarcinoma, the primary kind of esophageal cancer we see here, has gone up over the last 40 years, and not just a little bit. It’s a 5-fold increase. So we’re clearly failing in our mission to stop cancer, and we’re failing our patients because we’re not getting them screened.

Most of that is because we haven’t done a good job of telling our primary care colleagues how important screening is and exactly who needs it. So with the patients they see every day who should be screened, the topic never even comes up. Colorectal cancer screening is on the PCP’s list, and breast cancer screening is on the list if they’re seeing a woman. We just don’t make that list. Part of it is an education gap, and that’s on us. We haven’t communicated the urgency and the right information to put this at the front of the minds of the PCPs who will ultimately make that decision.

PCO: Can you walk us through how this test works and what the DETECT-ME study set out to evaluate?

Shaheen: The basic idea is this. There’s a big gap between the people who should be screened and the people who are screened. People don’t want to come in for a sedated procedure. Right now, if you’re a patient with chronic heartburn, which is the group we’re most interested in, getting screened for esophageal adenocarcinoma is a pretty involved process. Can we turn it into a point-of-care process that fits into a primary care office or workflow? Can we democratize screening so it isn’t tethered to an endoscopy unit? That’s the bigger question we’re asking with DETECT-ME.

DETECT-ME uses delightfully simple technology. It’s a sponge on a string. It doesn’t get much simpler than that. The sponge is encapsulated. You swallow the capsule, and I’ve done this myself. The capsule dissolves in your stomach and the sponge pops out. It’s about the size of a quarter. Then the string attached to the sponge is pulled back, and the sponge collects cells from the esophagus on its way up.

This technology has been around for years. What’s new in DETECT-ME is that it’s paired with a modern molecular test. In the old days, we would stain those cells by immunohistochemistry for a marker found in Barrett’s, and that was actually pretty good. We wanted to make it better. Now that we can do genetic sequencing on the DNA, could we build a highly sensitive and specific test? So in DETECT-ME, we took the molecular test we had developed and paired it with this very simple technology. The idea is that a PCP office could give the test quickly and cheaply, we run the genetics, and we screen patients for Barrett’s that way. We could tell about 90% of them they have nothing to worry about, and send only the roughly 10% who test positive on to endoscopy.

PCO: What were the key findings on how well the test detected Barrett’s esophagus, and what stood out to you about its performance in the patients at highest risk for cancer?

Shaheen: The first big question was whether Americans would put up with this. Most of the earlier work had been done in Britain, and we knew patients there tolerated it well. Would Americans? Here the sponge was given on a new flat plastic platform that looks something like a tongue depressor, called the EndoSign device. We placed the sponge at the back of the throat in over 400 patients. The good news is that over 95% of them swallowed it and completed the test, and by and large it was very well tolerated. Only about 2.6% said they wouldn’t do it again. A few said they’d have to think about it. The second important thing is that in the vast majority of patients, we got a good enough sample to analyze.

This is being developed as a screening test, meaning it’s for people who’ve never had an endoscopy. For this study, we did include a group like that. But we also had a second group of people we already knew had Barrett’s. You might ask why we’d include people we know have Barrett’s in a study about finding Barrett’s. In most screening populations, only about 10% of the people you screen have it. So unless we screened thousands and thousands of patients, the only economical way to do this with a few hundred patients was to enrich the population with people who already had Barrett’s.

We ended up with over 300 patients we could analyze. Within 6 weeks of the sponge test, all of them had an upper endoscopy, which was the gold standard we compared the sponge results to. If the sponge says yes and the endoscopy says yes, that’s a true positive. If the sponge says yes and the endoscopy says no, that’s a false positive, and so on.

Long story short, the sponge plus the test worked great. The sensitivity was 90%. The specificity, meaning how often the sponge correctly came back negative in patients without Barrett’s, was also high at 86%. Even better, the test found the people who matter most. We know that people with the longest segments of Barrett’s are the most likely to develop cancer, and the test found those people. When the test made a mistake, it was in short-segment disease. Most of the misses were in people with about 1 cm of Barrett’s, who are at the lowest risk. You want your test to be perfect, but if it has to make a mistake, you want it to make that mistake in the people it’s least likely to hurt. When you think about it, that makes sense. The sponge passes through that area, and if it doesn’t pick up that small patch of cells in 1 cm, you’ve missed it. No one in the entire study with 3 cm or more of Barrett’s was missed.

The test also found every patient with precancer. There were 46 patients in the study with some degree of dysplasia or with cancer itself, and the test went 46 for 46. That’s terrific, because a screening test can’t miss the people most likely to die of cancer. All of that is encouraging, and we were very happy with how the test performed. Does that mean we can’t make it better? No. But it’s a terrific start.

PCO: Patients found the procedure well tolerated, and most said they’d be willing to repeat it. Why does that matter for a screening test?

Shaheen: You don’t want somebody going home and telling their neighbor, “I swallowed this sponge and it was absolutely miserable,” because then you won’t get a shot at the neighbor. One of the things that helped socialize colonoscopy was Katie Couric’s willingness to have one, because it showed people it wasn’t a miserable experience. You want patients to say, “This wasn’t so bad. And by the way, I have Barrett’s, and because you’re my family, you’re at risk for Barrett’s too. I want all of you to go.” Imagine trying to get that message across when the patient’s experience was miserable. It’s not going to go very far. It’s important that we build a test with high tolerability.

PCO: How do you see a test like this fitting into the screening pathway alongside upper endoscopy, and what would need to happen for it to reach the primary care setting, where many at-risk patients are seen?

Shaheen: That’s the million-dollar question. For it to work, it has to sit in front of endoscopy. It has to be the initial screening test that decides whether you go on to endoscopy. We’ve already talked about the fact that PCPs have a thousand things to do. They’re incredibly busy. I think the way this will ultimately work is that GIs will partner with PCPs and say, “Let’s offer this in your office 2 half days a week. Schedule your patients with risk factors into those 2 half days.” They come in for their general exam, and they’ve come in fasting because they’re getting blood work. While they’re there, we ask them to swallow the sponge. Some will say, “You’re crazy. I’m not doing that.” If we believe the data from England, where this sponge has been used for a while, a large proportion will say, “Sure, I’ll do it,” and we’ll give it to them right then and there. We need to make a big easy button for primary care doctors. We get the results back, schedule the positive patients for their EGD, and those patients come into our practices.

GIs sometimes worry that technology like this will decrease the number of EGDs they do. We’re not getting 90% of these patients now. If we can implement this, it will actually increase our business, and it will increase it with the highest-risk patients who have already tested positive. Instead of doing a test with a 10% chance of finding Barrett’s, you’ll be doing one with maybe a 50% or 75% chance. That’s much higher yield. We get to the right patients and reduce disease, and that’s what we’re after. What I’ve just described is 2 or 3 steps down the road, but that’s how we get out of this box. The current paradigm has failed, so we have to think differently. If we keep doing what we’ve always done, we’ll keep getting what we’ve always gotten.

PCO: Is there anything else you’d like to highlight?

Shaheen: We need to be the ambassadors and the change makers for this kind of idea. If we don’t embrace it and bring it to our PCP colleagues, it won’t happen. We really do need to be the opinion leaders in our communities when these technologies become available, or else they’ll sit on the shelf.

Editors’ note: Shaheen reports relevant disclosures with Cyted, Lucid, Phathom Pharmaceuticals, Medtronic, and others.

Reference: Shaheen NJ, et al. A prospective, multicenter cohort study of a novel methylation biomarker panel using non-endoscopic cell collection to detect Barrett’s esophagus: the DETECT-ME study. Presented at: American College of Gastroenterology 2026 Annual Scientific Meeting; October 9-14, 2026; Nashville, TN.