Commentary on: Beyond Seattle: WATS and Advanced Imaging in Barrett’s Esophagus

Abstract

The incidence of esophageal cancer continues to rise but countless data shows that it is highly curable when diagnosed and treated early with currently available advanced therapeutic techniques. Barrett’s esophagus (BE) has been identified as the leading risk factor, if not precursor, for the development of esophageal cancer. However, despite the fact that we are armed with the knowledge of the highest-risk group and have excellent therapeutics, esophageal cancer continues to still kill >16 000 Americans each year with a 5-year survival rate still a dismal 19%.1 The problem therefore is that we are failing at finding curable disease. First problem is there is no universally agreed upon protocol for screening patients to identify the BE. People reasonably argue that proposals to examine all 30 million Americans with GERD using current technology are not fiscally reasonable in light of lives saved. Second, once we have identified BE, we enroll them in surveillance which at best skips large areas of their disease and hope nothing grows until the next time, which in some recommendations is many years later. This is a problem. We need cheaper, quicker, better solutions for both screening and surveillance. This article reviews the current status of diagnostic capabilities and quite convincingly makes the case for “better”—we can definitely do better than the gold standard Seattle Protocol (especially given that adherence to the actual protocol is shamelessly low2). The ASGE Technology committee criteria provides a useful tool to ensure that we objectively embrace or discard new technologies in terms of “better” leaving issues of “cheaper” and “quicker” largely unmitigated. Payors want “cheaper” and proceduralists want “quicker.” The loss of volumetric laser endomicroscopy (VLE) is heartbreaking because it was the best technology to look for recurrence under Barrett’s ablation, so-called buried glands in this high-risk group. As with many of these technologies, the cost is not deemed justified to most payers so we find ourselves stuck between our desire to make an impact in this deadly disease with improved surveillance and early intervention and the need to get paid for our work and expertise. AI is fantastically exciting but I worry it will also be out of reach financially. Similarly, probe-based confocal laser endomicroscopy (pCLE) is not widely adopted due to concerns that it adds too much procedure time and due to proceduralist insecurity of the ability to read the images. However, I will point out that using a real-time live point sampling technique (rather than the record and review method) hardly adds any time at all and substantially increases the sampling yield compared to random biopsy. Despite the trepidation with image interpretation, the pattern recognition learning curve is really quite acceptable3 (you’re smart, you can do it!). Lastly, the pCLE probe can be used ex vivo to confirm that the lesion in question was removed. Are these enough to keep the technology available? I would like to say I use it for all Barrett’s surveillance but currently only about 1⁄2 of my patients can get it due to insurance. So, I hope so. Is the WATS brush the holy grail? At least better? The non-randomized data for WATS is exceptionally promising at detecting BE and dysplasia. In addition, we recently completed a multicenter randomized trial to compare the WATS brush to routine protocol forceps biopsies for both screening and surveillance endoscopy.4 In general, the results showed no difference in detection of BE or dysplasia between the two techniques suggesting that it was un-necessary to do both. This finding is especially useful in long segment BE in which the Seattle protocol, or even a more targeted approach with pCLE, is quite time consuming and likely more prone to false negatives than the wide epithelial sampling provided by the WATS brush. So better? Probably/possibly. Certainly not worse! Quicker? Definitely. Cheaper? Well, that depends on how you calculate it but likely yes (remember the pathologist charges by the jar). Overall, we need to figure this out. We need a cost effective, low risk screening modality, and a plan. We need to risk stratify and save some of the most frequent surveillance schedules and more advanced technologies for those who are highest risk. Not every short segment BE needs annual, expensive multimodality surveillance. But random biopsies every 3 to 5 years for a 65-year-old male with long segment BE? Makes me shutter. We can do better.