David A. Johnson, MD


June 08, 2015

This feature requires the newest version of Flash. You can download it here.

Lessons Learned

Hello. I'm Dr David Johnson, professor of medicine and chief of gastroenterology at Eastern Virginia Medical School in Norfolk, Virginia. I just got back from Washington, DC, and another somewhat overwhelming Digestive Disease Week. For those who were there, I'm sure you saw lots of interesting science presented. For those who weren't, well, you missed a lot of good things.

For both those present and absent, let me give you my perspective on the meeting. I think there were many interesting presentations. I will confess to you that I've gone through the entire abstract book, which is about 4 inches thick. I've looked at the poster and plenary presentations and also attended a number of plenary and scientific session presentations. Let's begin.

Eosinophilic Esophagitis

The first one that caught my eye was a presentation[1] from the University of North Carolina at Chapel Hill looking at eosinophilic esophagitis. The investigators compared the effectiveness of the six-food elimination diet vs topical corticosteroids, allowing either fluticasone or budesonide. The time horizon was 5 years. They allowed seven endoscopies to reassess the participants as foods were reintroduced. Lo and behold, from the perspective of a payer, it was far more advantageous to go through the six-food elimination diet, even with all these endoscopies after each reintroduction.

So if you get any pushback from your payer, this would be a good study to pull out, because going with the six-food elimination diet is shown to be more cost-effective. It is a whole lot less expensive from a food standpoint, but the endoscopies do still drive the cost.

Celiac Disease

The next one got a lot of press. If you haven't seen this, you probably will, or you'll hear about it.

A group from Columbia University[2] looked at the widespread contamination of probiotics with gluten. How many of your patients take probiotics? Mine come in all the time and tell me that they do. How many of you recommend probiotics? Be careful what you recommend, because this was a very interesting study.

Dr Green and his colleagues at Columbia looked at the gluten component using liquid chromatography. They looked at a composite of 22 probiotics, and a little more than two thirds of these were listed as gluten-free. When they actually studied them, however, they found that 55% contained gluten and two of them contained an incredibly high amount of gluten—amounts that they would view as a very alarming amount.

Of the 15 probiotics that were labeled gluten-free, eight (53%) tested positive for gluten, including two that contained gluten levels more than 20 parts per million—suggesting that gluten is in a lot of these probiotics, even when they say it isn't. A variety of gluten-containing products were in some of these probiotics. Two of them actually tested positive for barley and rye.

This may be something to question your patients about when they may not be responding to treatment. Certainly, be cautious when you recommend probiotics across the board, especially within the population with celiac disease.


Dr Stavropoulos,[3] the only gastroenterologist with a world-related population report of peroral endoscopy myotomy (POEM) therapy for achalasia, presented 5-year data on his remarkable success in providing treatment to a high-risk population.

He reported on 174 patients, of whom:

  • Four were older than 90 years;

  • 26% had severe comorbidities;

  • 49% had previous treatments (eg, Botox®, failed Heller myotomies) that made POEM challenging; and

  • 26% had advanced or end-stage achalasia with an esophageal diameter > 6 cm.

Remarkably, all these patients did better. There were no major adverse events, suggesting that POEM is a real deal that should be considered, even for elderly, end-stage patients with comorbidities, before you consider esophagectomy.

But it really comes down to the expertise of the locals, and although I don't know what we can extrapolate from this world-renowned experience, look at your institution referral rates for where these patients are best served. It is my bias that for type III achalasia, and particularly for chest pain, POEM is the way to go. And it may become one of the first interventions, as we look at these patients in the long term.

Crohn Disease and Therapy Withdrawal

The next couple of presentations dealt with Crohn disease, particularly as it relates to withdrawal of anti-tumor necrosis factor (TNF) alpha therapy.

One study[4] looked at withdrawal therapy in patients (from a single institution) who had significant, deep remission. In 61 patients who were in deep remission, investigators withdrew the anti–TNF-alpha antagonist, be it infliximab or adalimumab. They found that the relapse rate was approximately one half at 2 years. There was really nothing that suggested a preselection bias. The highest relapse rate was in the first year. Ileal disease somewhat increased the risk for disease flare-up, but they found no other identifiable factors.

Another fascinating build on that study comes from the United Kingdom.[5] In the United Kingdom, regulators require a reassessment of disease after 12 months of therapy, and if the patient is in clinical remission, it is recommended that you withdraw therapy.

The investigators looked at 160 cases: 130 with infliximab and 30 with adalimumab. When they withdrew the therapy, the relapse rate was quite striking—44% at 1 year and 50% at 2 years. Younger age at diagnosis and an elevated white blood cell count were predictive of relapse. No other predictors were found, including such factors as smoking. Again, there was a very high relapse rate, even in the deep remission population. We clearly need better indicators of whom these agents can be withdrawn from.

Next is a very eye-opening study from Regueiro and colleagues,[6] which is postoperative intervention with an anti–TNF-alpha antagonist in the postsurgical patient with Crohn disease. This was a 5-year, randomized, placebo-controlled, multicenter study comparing the withdrawal of all therapy with infliximab (5 mg/kg). Dose escalation was allowed, the time horizon was 5 years, and 297 patients were followed. The primary endpoint was a composite of clinical symptoms combined with endoscopic findings.

At the end of the evaluation at 76 weeks, no statistical difference was found between the placebo and control populations, although there was a numerical difference: 13% recurrence in the active treatment group vs 20% in the placebo group. The primary endpoint was not met. However, if you look at endoscopic recurrence, there was nearly a twofold difference (30% with infliximab vs 60% with placebo). Because the primary endpoint was not met, however, it was not statistically significant, so the sponsor decided to close the study.

Nonetheless, I don't think all the answers are in here. The mucosal relapse rate is strikingly in favor of continued therapy. I'm not sure that we have grounds to say that we should withdraw therapy in patients who had enough complications to warrant surgery.

Endoscopy and Gastrointestinal Hemoclips

A couple of other very interesting studies related to prophylactic clip therapy following endoscopic intervention in the colon.

First was a systematic review and meta-analysis that looked at whether prophylactic clip placement after polypectomy was warranted.[7] The authors identified about 1200 studies, of which 38 met entry criteria, allowing for approximately 12,000 patients with more than 8300 polypectomies. The results showed that there was no significant beneficial effect from prophylactic clip placement. Clearly, this is something that catches our eye, because many times, these are placed by us. However, there is a sizeable cost for clip placement, and this is something that should be looked at in an individual institution. Nonetheless, this is something that did not routinely meet criteria.

To build on that was a study[8] from two tertiary referral centers in the San Francisco Bay area: one from the University of California, San Francisco, and the other from Stanford University. The researchers looked at risk factors for postpolypectomy bleeding, particularly in patients receiving anticoagulation or antiplatelet medications. They found that something was in favor of an increased risk for bleeding, and it makes sense: They found that large polyp size, restarting antithrombotics or anticoagulants, and right-sided cautery increased the risk for rebleeding. But similar to the meta-analysis above, no benefit was found from the use of cautery or clips for decreasing the risk for recurrent bleeding.

Treatment of Gastroesophageal Reflux Disease

The next study relates to the ongoing evaluation of the LINX® Reflux Management System (Torax Medical, Inc.; Shoreview, Minnesota). LINX is a magnetic device used for the enhancement of the lower esophageal sphincter. Ganz[9] led this coalition and has now presented the 5-year data from 100 patients. All patients were on proton pump inhibitors (PPIs) at study entry, and the use of PPIs at 5 years was 15%.

What I found very interesting was the reduction in regurgitation, which is the most problematic symptom that we deal with. Regurgitation was reduced from 57% at baseline to 1.2% at 5 years. It's a very powerful indication that this is a durable operation, with no significant gas or bloating effect. Retching and the inability to vomit seem to be very much overcome with this device.

Again, I think this is something that you should be looking at with your local surgeons. I would deal with your thoracic and foregut surgeons, particularly those who have an interest in this area. Many of these operations with LINX seem to require some crural diaphragm repair, especially if there is a hiatal hernia, so it may be something that you want to deal with an expert for.

Take a look at this study. I think the safety is overwhelmingly in favor of the device, and the efficacy is quite striking.

Pancreatic Insufficiency

Another interesting study[10] was related to pancreatic insufficiency in celiac disease.

We frequently start with a diarrhea evaluation and do a celiac disease comprehensive panel. Perhaps we'll do pancreatic elastase and stool elastase test, and look for pancreatic insufficiency. But lo and behold, there is a significant overlap of patients with celiac disease and patients who have fecal elastase levels that are quite low.

This study looked at a group of 36 patients with celiac disease, of whom seven had low fecal elastase levels. (One patient had had acute pancreatitis.) The seven patients were studied with endoscopic ultrasonography of the pancreas; no structural abnormalities were found. At the end of 6 months on a gluten-free diet, six of the seven patients had normalized their fecal elastase levels. Again, this was without intervention for pancreatic supplementation.

Something to remember when you're looking at gluten-sensitive enteropathy is that these patients may actually have low fecal elastase levels.

Irritable Bowel Syndrome

Two presentations[11,12] related to therapy with peppermint oil, which was reintroduced as a viable agent in the most recent irritable bowel syndrome (IBS) guidelines by the American College of Gastroenterology. These studieslooked at a new formulation.

You recall that L-menthol, which is the main component of peppermint oil, has antispasmodic and antinausea or anticarminative effects, as well as topical analgesic effects and a 5-HT3 receptor antagonist property. This was a prospective randomized study in IBS-D (diarrhea) and IBS-M (mixed). Seventy-two participants were enrolled, and the endpoint was 4 weeks.

There was nearly a twofold incremental gain in the active therapy arm with peppermint oil. IBgard is an over-the-counter, commercially available product, and it seems to have a remarkable effect. Virtually all patients who had moderate to severe symptoms improved, particularly related to abdominal pain or discomfort, bloating, or distention. There was no improvement in the constipation.For the passage of gas, sensation of incomplete evacuation, or obstipation, all seemingly improved in a very statistically significant fashion. There is a very low threshold to use this medication as it relates to a favorable benefit.

Barrett Esophagus

Let's move on to Barrett esophagus. A couple of presentations caught my eye.

One was progression as it relates to low-grade dysplasia.[13] This was the Mayo experience in a prospective registry. There is recognized intraobserver variability, but it is highly sensitized to the Mayo expertise. In looking at 337 patients, progression of disease was strikingly low. The incidence of progression from low-grade dysplasia to either high-grade dysplasia or esophageal cancer was approximately 0.8% per year. This is very low and very different from what we have seen from a lot of European studies.

With that idea, let's move forward to indefinite dysplasia.

The study[14] from Dr Falk's group at the University of Pennsylvania looked at 354 patients with indefinite dysplasia. Reactive atypia is sometimes the mirror of confusion, at least in a lot of institutions. But they in fact were looking at patients after they excluded prevalent disease, so patients had no dysplasia. They found that indefinite dysplasia was strikingly low. An incidence of any dysplasia was about 1.4 cases per 100 person-years of high-grade dysplasia or esophageal adenocarcinoma. This suggests that indefinite dysplasia shouldn't be dismissed.

There also are two recent publications that I want to mention. One is in the American Journal of Gastroenterology and is the BOB CAT guidelines,[15] which is a Delphi consensus guideline that registers recommendations for no dysplasia, low-grade dysplasia, and indefinite for dysplasia, as well as surveillance recommendations. This is something that I think is worth looking at. The other is a study[16] from the University of Washington in Seattle that looked at indefinite dysplasia, and is published in Clinical and Translational Gastroenterology.

Gastrointestinal Bleeding and NSAIDs

Next is a randomized trial[17] by Chan, who is world-renowned for some of the seminal studies in nonsteroidal anti-inflammatory drugs (NSAIDs) and bleeding, that looked at prevention of recurrent ulcer bleeding in arthritis patients with high cardiovascular and gastrointestinal risks. Patients were randomly assigned to receive either celecoxib or naproxen. Cardiac aspirin was allowed for these patients, and it was balanced at about 70% of patients in both groups. The endpoint was clinical rebleeding.

The hazard ratio was 0.45 for the cyclooxygenase 2 (COX-2) inhibitor, even while cardiac aspirin was allowed. Therefore, it was a 55% risk reduction as it relates to the selective COX inhibitor vs a nonselective inhibitor. If you're facing these high-risk cardiovascular or gastrointestinal patients, it seems that celecoxib, and allowing cardiac aspirin if you feel that it is necessary, is the way to go.

PPI Safety and Bone Density

This next study is near and dear to my heart, because I get these questions all the time.

This is Targownick's study[18] from Canada. Targownick has been a leader in the evaluation of bone density changes as they relate to PPI use, which is something that continues to plague us. There are actually lawsuits out there because of PPIs and fractures.

This is a 5-year study that was a little different, but very good. The investigators looked at it in an interesting way: Rather than standard bone density, they looked at three-dimensional CT, which provides volumetric bone density as well as differentiation between the cortical and trabecular compartments. It is a very significant, highly advanced way to evaluate patients. They looked at 104 participants: 52 PPI users and 52 controls. The bottom line is that there was no difference.

Please, don't dismiss this in your conversations with patients. I always tell them that this is what the latest information is. The data continue to resound that the retrospective databases are not correct as they relate to PPI use and bone density loss in hip, cervical spine, and radial fractures. If patients need a PPI, they should take it. If they need calcium and vitamin D, they should take them, but not because they are on a PPI.

Hepatitis B

I want to draw your attention to a systematic review and meta-analysis[19] for hepatitis B reactivation in patients who were undergoing chemotherapy for solid malignancies. Most of us would recognize that when using rituximab, we would screen patients for hepatitis B; oncologists recognize this. But this systematic review and meta-analysis suggested that even patients who are undergoing chemotherapy for solid tumors or hematologic-type cancers should all undergo hepatitis B screening.

I also would like to draw your attention to a marvelous review[20] in Gastroenterology. Reddy coordinated this, and it relates to hepatitis B reactivation in patients undergoing immunosuppression of any sort. It relates to immunosuppression for rheumatologic or immunologic suppression with chemotherapy. It's a must-read review. I sent this guideline to my oncology group. We have a group of about 100 oncologists in our area, and they said, "Wow, we don't do this." So read this guideline. Look at it. Share it with your oncologists.

Again, the systematic review just heightened my attention to this guideline.

Cirrhosis and Hepatitis C

There were a couple of papers related to liver disease that I thought were important.

Mohanty's study[21] looked at statin use in patients who were facing decompensation or death. Very frequently, we get gun-shy about using statins in patients who are cirrhotic. This study looked at the concomitant use of statins in cirrhotic patients. The time horizon was 10 years, and there was a risk reduction of 40% at 1, 5, and 10 years. Recognize that the risk reduction in decompensation and death in cirrhotic patients was favorably enhanced in patients who were taking statins.

Now what does this mean? It means that we need to study this further. I wouldn't rush out and put all your cirrhotic patients on statins; but if they need statins, don't withhold them. We know from evaluations of cirrhotic patients that they are vastly underserved as far as the concomitant use of statins for hyperlipidemia, and for obvious reasons. So, again, don't withhold these medications from this population. These data on decompensation and cirrhotic patients are striking.

A new index was proposed by O'Leary and colleagues.[22] I've never seen this before. This index is for patients with infection- or non–infection-related cirrhosis, and it relates to 1-month mortality. It is an infection-related index that is based on renal disease, hepatic encephalopathy, respiratory insufficiency, and shock. It was a new validated measure, and it gave a nice clinical bedside snapshot of the mortality of these people. It also underscores how important renal disease is in these patients. Look back 6 months, particularly from admission, to the progression of renal disease; you need to be looking at this.

There were a number of studies that spoke about hepatitis C. What can I tell you? Hepatitis C, we're all in there. We can cure the disease. I attended a very dynamic session led by Chung and Martin about profiling how evolutionary hepatitis C is. Martin led the American Association for the Study of Liver Diseases in a combined initiative with the Infectious Diseases Society of America to set forth a "living document" that allows for regular, routine updates of hepatitis C therapies. This document is available on HCVguidelines.org.

This is something that you have to keep on your desktop. Look at it. It's updated regularly. I think this will set a stage for how we start to do guidelines, because in rapidly evolving areas such as hepatitis C, we can't wait for something to be updated every 2 or 3 years.

Inflammatory Bowel Disease

There were a couple of inflammatory bowel disease (IBD) presentations.

One was on vedolizumab.[23] This study comes from Christensen and colleagues at the University of Chicago. Having the world's largest experience, they looked at the clinical remission rates in vedolizumab therapy. Remarkably, they found that their induction rates at 14 weeks were in the 35%-40% range.

If you go back to Sandborn and colleagues' paper[24] in the New England Journal of Medicine in 2013, the induction rates weren't quite so exciting. The range was around 14%, and the placebo rate was 6.8%. It didn't really work that much for Crohn disease. But when it got into those patients who responded, the maintenance of remission was much better—in the 35%-plus range. I think we're probably just saying that the induction window probably needs to be longer. Certainly the University of Chicago experience underscores that.

The second presentation[25] also was from the University of Chicago, again looking at low-grade dysplasia patients. When we see low-grade dysplasia, we would immediately say that you have to go to surgery. Well, these investigators looked retrospectively in their evaluation and found that when those patients went to surgery, none had any evidence of high-grade dysplasia. These patients had no other visible lesions. What we now can do with chromoendoscopy and local resection of visible lesions underscores that we've changed the paradigm of how we approach these patients with IBD and even low-grade dysplasia with resectable lesions.

So, lots of stuff—I've tried to just give you my perspective. These are good abstracts that I've cited for you, so delve into them, because there are some game changers. I'm Dr David Johnson. See you next time.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: