Radiofrequency Ablation for Nondysplastic Barrett’s Esophagus
One of the big controversies in gastroenterology currently is whether to ablate non-dysplastic Barrett’s esophagus. In the past, the number of treatments needed and side effects of some of the ablative therapies kept ablation of nondysplastic BE from being used for most patients. For example, photodynamic therapy was developed and used to treat patients with dysplasia and early cancer with good success. However, the side effects (photosensitivity, chest pain, strictures) kept it from being a viable option for treatment of patients with no dysplasia. Other types of ablation, such as cautery and argon plasma coagulation, have also been used, but typically for patients with dysplasia. There’s no short answer to this question, so this post will be longer than usual.
Now that RFA (Barrx) is available, there has been renewed interest in ablating nondysplastic BE. Here’s some of the research that’s been published on ablation of BE with no dysplasia and BE with low grade dysplasia (LGD):
AIM-1 and AIM-2 Studies
This set of studies looked at 102 patients with Barrett’s esophagus, but no dysplasia. This is the largest study looking at RFA ablation in patients with BE and no dysplasia. All patients were treated with circumferential RFA.
AIM-1: 32 patients were treated with RFA. The purpose of this study was to choose the right dose of energy needed to ablate the Barrett’s esophagus. All the patients were treated with RFA, but at different doses (from 6 J/cm2 to 12 J/cm2). Repeat EGD with biopsy was performed at 1 month and 3 months after the initial treatment.
Results: 26 of 32 patients had some BE left after the first RFA treatment and were retreated with an RFA dose of 10 J/cm2.
At 12 month follow up, EGD results were available for 31 of the 32 patients.
- 19 patients (59% of the whole group) had no BE left.
- 8 patients (25%) had a partial response, so had some BE left.
- 5 patients (16%) had no response, so their BE was still present after treatment.
Adverse events/complications: 1 patient with scarring that resolved in a few months, 3 patients with chest pain, and 1 patient with a shallow mucosal tear that resolved. For the patients who were treated with a second ablation, 3 had fever, 1 had scarring that resolved, 1 had abdominal pain, and 2 had problems with sedation.
AIM-2: This study was done to see how effective RFA was in completely ablating nondysplastic BE. Using the dose of RFA that worked best from AIM-1 (10 J/cm2), 70 patients were treated with RFA.
Results: Of the 70 patients, 36 had a second RFA procedure to treat residual BE. 12 month EGD results were available for 69 of the 70 patients.
- 48 patients (69%) had a complete response to treatment, so had no Barrett’s esophagus left.
- 17 patients (24%) had a partial response, so had some BE left after treatment.
- 5 patients (7%) had no response to treatment.
Adverse events/complications: 24 adverse events were reported (out of 106 treatments). 2 patients had fever and 9 had sore throat/chest pain. There was also 1 superficial mucosal injury, 1 minor bleed, 1 scarred area that resolved in time, 2 sedation problems, and 8 episodes of nausea. There were no strictures or buried Barrett’s glands.
Summary of AIM-1 and AIM-2: About 2/3 of patients treated with RFA had complete eradication of their BE. Some patients (7-15%) didn’t have any improvement at all. The rest of the patients had partial improvement of their BE, with some BE left after treatment with RFA. Patients tolerated the procedures well.
Prospective Pilot Trial of HALO in BE with LGD
This small study (10 patients) looked at patients with 2-6 cm of BE with LGD who were treated with circumferential and focal RFA.
Results: 10 patients were treated with circumferential RFA. EGD with biopsy was repeated 1 and 3 months after the procedure to look for BE.
- 6 of the patients had some BE left after the 1st treatment. Of these 6, 2 still had LGD and 1 had HGD. These 6 patients had a second circumferential RFA procedure.
- At 1 year, EGD with biopsy showed that 9 of the patients had a small amount of BE left. 9 of 10 patients (90%) had no dysplasia, 1 patient had a nodule removed by EMR that showed HGD and a tiny intramucosal cancer.
- At 1 year, 7 of 10 patients (70%) had a complete response to treatment, meaning they had no BE left. 3/10 still had some BE.
- The 9 patients were treated with focal RFA after their 1 year EGD with biopsy.
- At 2 years, no patients had dysplasia on biopsy (100% response).
- At 2 years, 90% had complete eradication of BE.
Adverse events: 1 patient taking aspirin had bleeding that resolved. There were no strictures and no buried Barrett’s.
Summary of the pilot trial: At 2 years, none of the 10 patients had LGD. 9 of the 10 patients were BE-free. 1 patient was diagnosed with HGD and intramucosal cancer during the study (in a nodule) and was treated. Patients tolerated the procedures well. The authors point out that patients undergoing ablation need frequent biopsies to make sure that the dysplasia and BE are completely treated. They also suggest that surveillance biopsies should be continued after ablation is complete.
Pilot Series of RFA for BE With or Without Neoplasia
40 patients with BE were studied and treated with circumferential and focal RFA. Of the 40 patients, 5 had dysplasia and the other 35 had nondysplastic BE. All patients had follow up EGD 3 months and 12 months after the first (circumferential) RFA treatment. Patients with residual BE were retreated with focal RFA.
Results: The paper presents the results of the 10 patients who completed 12 months of follow-up after treatment. 28 patients had repeat EGD 3 months after RFA treatment, but hadn’t completed 12 month follow-up at the time the paper was published.
- For the 10 patients who have completed the study and have had their 12 month EGD
- 7 (70%) have no BE left on EGD with biopsy (complete response)
- 3 (30%) had a partial response with only a few islands of BE left
- For the 28 patients who have completed 3 months of follow up after treatment
- 12 (43%) had a complete response (no residual BE)
- 14 (50%) had a partial response (some BE left)
- 2 (7%) have no improvement in their BE after RFA
Adverse events/complications: Mild throat and chest pain in some patients, there were no perforations or strictures. There was one case of ‘buried Barrett’s, or BE glands under new squamous epithelium, treated with repeat ablation.
Summary of the pilot series: 3 months after treatment with circumferential RFA, 43% of patients had no residual Barrett’s esophagus. For the 10 patients with 1 year follow up, 70% of patients had no BE left. The final results of the study are pending. The procedures were well tolerated.
So the big question – what do these studies mean for patients with nondysplastic BE? The RFA procedures seem fairly safe, with the main adverse events and complications being chest discomfort and sore throat. There were no perforations and the bleeding rate was low. For patients with nondysplastic BE and LGD, 60-70% of patients had complete eradication of their BE, but 1/3 still had some BE left. These studies aren’t particularly large, but larger studies are likely ongoing around the US and Europe.
Questions that are unanswered:
- What we don’t know yet is whether ‘a little bit’ of BE is as dangerous as longer BE. Cancer is more common in patients with long-segment BE, but cancers do occur in short lengths of BE. Are the few tongues and islands left after RFA still a risk for future cancer? No one knows. Long term studies would be needed.
- How many treatments is the right number? Most of these studies used 2 treatments, but RFA sessions can be repeated more if needed.
- After treatment, what is the risk of the BE re-growing? No one can answer this question yet because there isn’t data to answer the question. Several of the RFA studies are following patients long term, so more data on this topic will be available in a few years.
- Do patients with complete ablation still need surveillance EGDs? For now, the answer is yes because no one knows if the BE will return. EGD with biopsy should still be performed with biopsies of anything that looks like BE, plus biopsies of the areas that were treated with RFA to look for buried BE glands.
- Is RFA for ablation of nondysplastic BE cost-effective? Unclear. Treatments are relatively expensive. Because RFA is a new technology, not much has been done to study this yet.
- The risk of esophageal cancer in non-dysplastic BE is low (0.5%/yr). Do the potential risks of RFA outweigh the potential benefit? Most of the studies use 1-2 treatments for BE. What happens to the risk if a patient needs to have 3 or 4 treatments? The answer to this isn’t clear and is definitely worth discussing with your own gastroenterologist.
- 1/3 of patients with LGD will have regression to nondysplastic BE. Should all patients with LGD be treated with RFA? Unclear. The published studies of LGD are small and more data is needed.
In summary, there is still much debate in the GI community about ablation of nondysplastic BE. Some gastroenterologists do use RFA to ablate nondysplastic Barrett’s. But other gastroenterologists don’t and are waiting for more study results. Hopefully soon there will be more published research on the subject. At this point, there’s no a clear right answer, so definitely talk with your own GI physician about her/his opinion about treating nondysplastic BE with RFA. You can also consider seeing another gastroenterologist for a second opinion to get a second perspective on treatment.
References for the studies mentioned above:
Sharma VK, et al. Balloon-based, circumferential, endoscopic radiofrequency ablation of Barrett’s esophagus: 1-year follow-up of 100 patients. Gastrointestinal Endoscopy 2007, vol 65, number 2, pp.185-195.
Sharma VK, et al. A prospective pilot trial of ablation of Barrett’s esophagus with low-grade dysplasia using stepwise circumferential and focal ablation (HALO system). Endoscopy 2008; vol 40, pp. 380-387.
Hernandez JC, et al. Pilot series of radiofrequency ablation of Barrett’s esophagus with or without neoplasia. Endoscopy 2008l vol 40, pp 388-392.
posted by Kerry Dunbar, MD