The Gastro & Hep Report Fall 2014

Presentations in IBD

 

A Study of Fecal Microbiota Therapy in Patients With Ulcerative Colitis

 

Although fecal microbiota therapy (FMT) can successfully treat Clostridium difficile colitis, its efficacy in treating active ulcerative colitis (UC) has yet to be clearly established. At Digestive Disease Week 2014, Dr Paul Moayyedi of the Department of Medicine at McMaster University in Hamilton, Ontario presented results of the first randomized, placebo-controlled trial investigating the efficacy of FMT in treating UC (Abstract 929c). Eligible patients were ambulatory, with active UC defined by a Mayo score of at least 4 and an endoscopic Mayo score of at least 1 with no upper limit. All enrolled patients tested negative for the C difficile toxin gene by polymerase chain reaction. Patients were allowed to continue taking any UC medications they had been consistently using during the previous weeks. No antibiotics were allowed within 30 days before enrollment. Examination with a flexible sigmoidoscopy was performed at baseline.

Patients were randomized to receive FMT given by a 50-mL retention enema from an anonymous donor, or placebo consisting of a 50-mL water enema, once per week for 6 weeks. The primary outcome was remission of UC, defined as a Mayo score of 2 or less with an endoscopic Mayo score of 0 at week 7. Secondary outcomes included health measures assessed by the Mayo score, the IBD questionnaire, and the EQ-5D.

At the time of the presentation, 53 patients had completed the study, and 10 patients were continuing on study treatment. Baseline characteristics were well balanced between the 2 treatment arms. Forty-five percent of patients had pancolitis and 42% were taking steroid medication, with 19% on immunomodulators and 9% on biologic therapy.

No differences were observed for the study treatment vs placebo in primary outcome or in any of the secondary outcomes. Although no patients were in remission at week 7, 16 patients in the FMT arm subjectively reported symptom improvement. After these patients continued on unblinded, weekly FMT for a further 6 to 12 weeks, 9 of 27 patients achieved remission (33%; 95% CI, 15%-51%). No major AEs were attributable to FMT. Dr Moayyedi and colleagues concluded that FMT could be effective over time periods longer than 6 weeks, but the therapy should be evaluated only in the context of clinical studies.

 

Persistent Antibodies to Infliximab Predict Loss of Response in IBD

 

IBD patients treated with infliximab may develop antibodies to infliximab (ATIs), which can reduce the efficacy of this therapy. In a poster presented at Digestive Disease Week 2014, Dr Manon Leclerc of Saint-Etienne University Hospital in Saint-Priest-en-Jarez, France provided results from a prospective study investigating the ability of ATIs to predict loss of response in patients with IBD (Abstract Sa1257). Loss of clinical response was defined as an increase in clinical symptoms requiring a therapeutic adjustment (eg, infliximab dose intensification, initiation of another medication, and surgery). CRP was followed as a marker of disease activity. The study included 481 blood samples from 93 consecutive IBD patients, including 59 with Crohn’s disease, who were treated with infliximab. The patients’ mean age was 30 years. The mean duration of follow-up was 17.2 months.

Loss of clinical response was demonstrated in 32 patients (34.4%). Thirty-four patients (38%) had normal levels of CRP, and 27 patients (29%) had detectable ATIs. Among the patients with ATIs, a slight majority (51.9%) had only 1 ATI-positive sample; in 48.1%, more than 50% of the samples were ATI-positive. The presence of ATIs was significantly associated with loss of response (P=.011) and positive CRP (P=.0003). ATIs exceeding 20 ng/mL in the first sample predicted loss of response with 94% specificity and 22% sensitivity (likelihood ratio, 3.39; area under the receiver operating characteristic curve, 0.59). An increasing number of consecutive samples positive for ATIs correlated to an increasing likelihood of loss of response. The presence of positive ATIs in more than 50% of a patient’s samples was associated with more than 50% loss of response to infliximab during follow-up, and permanent ATIs were associated with systematic clinical relapse (P=.0044). However, transient ATIs (present in only 1 sample) were not associated with loss of response (P=.01; Figure 3). Other factors not associated with loss of response included concomitant thiopurines, and infliximab duration and dose. A univariate analysis showed that loss of response corresponded to an ATI level greater than 20 ng/mL (P=.0071), CRP levels greater than 5 mg/L (P=.0046), and clinical activity (P=.0026). A multivariate analysis was used to calculate the relative risks associated with maintaining clinical remission; they were 0.64 (95% CI, 0.46-0.90) for an ATI level greater than 20 ng/mL and 0.65 (95% CI, 0.43-0.90) for a CRP level greater than 5 mg/L. These 2 factors combined produced an increased relative risk of 0.21 (95% CI, 0.08-0.55).

 

C-Reactive Protein Predicts Histologic Activity in Patients With Ulcerative Colitis

 

Although mucosal healing has emerged as a new therapeutic goal for patients with UC, it may not accurately indicate histologic remission. Persistent microscopic inflammation is associated with an increased risk of colorectal neoplasia and may predict an increased risk of colectomy and hospitalization. Therefore, histologic remission is gaining use as a clinical trial endpoint in studies of UC. At Digestive Disease Week 2014, Dr Britt Christensen of the University of Chicago Department of Medicine in Chicago, Illinois described results of a study that examined the prevalence and possible predictors of mucosal healing, histologic remission, and mucosal healing without histologic remission in patients with confirmed UC (Abstract 964). The study identified patients who had undergone a baseline colonoscopy with biopsies taken in each colon segment (rectum, left side, and right side) and a follow-up colonoscopy more than 1 year later. Each colonic segment of the final colonoscopy was retrospectively evaluated for mucosal healing, defined as normal mucosa or quiescent inflammation by endoscopy, and for histologic remission, defined by the absence of residual histologic inflammation. The primary outcomes were mucosal healing, histologic remission, and mucosal healing without histologic remission.

The study included 646 patients with a median age of 29 years (range, 5-82 years), of whom 60.0% had mucosal healing and 50.3% had histologic remission. Mucosal healing was strongly associated with histologic remission (OR, 13.41; 95% CI, 7.137-26.128; P<.001). Multivariate analysis revealed several predictors of mucosal healing, including longer duration of disease (P=.004), current therapy with an anti-TNF treatment (P=.019), current therapy with an antimetabolite (P=.037), and a CRP level of less than 3.0 mg/L (P=.002). Among the 385 patients who had mucosal healing, 80 (20.8%) had ongoing histologic activity. Multivariate analysis revealed that a CRP level above 2.9 mg/L was the only variable predictive factor of ongoing histologic activity in patients with mucosal healing (adjusted OR, 3.56; 95% CI, 1.55-8.16; P=.003). In summary, although the presence of mucosal healing was strongly associated with histologic remission, biopsy revealed histologic activity in approximately 21% of patients with mucosal healing, and a CRP level greater than 2.9 mg/L predicted ongoing histologic activity in patients with complete mucosal healing.

 

A Higher Dose of Mesalamine Reduces Fecal Calprotectin Levels in Ulcerative Colitis Patients

 

In patients with quiescent UC, lower concentrations of fecal calprotectin are associated with lower rates of relapse. At Digestive Disease Week 2014, Dr Mark T. Osterman of the Hospital of the University of Pennsylvania in Philadelphia, Pennsylvania presented results from a study examining mesalamine and fecal calprotectin levels (Abstract 862). The randomized controlled trial evaluated whether higher doses of mesalamine can reduce the concentration of fecal calprotectin in patients with quiescent UC. Enrolled patients had UC in remission (based on the Simple Clinical Colitis Activity Index), were taking no more than 3 g/day of mesalamine, and had a fecal calprotectin concentration exceeding 50 µg/g. As needed, the patients were switched to multimatrix mesalamine at 2.4 g/day for 6 weeks prior to randomization. Patients were evenly randomized to continue their current mesalamine dose or to increase the dose by 2.4 g/day for 6 weeks. The primary outcome was continued remission with fecal calprotectin concentration of less than 50 µg/g at 6 weeks. Secondary outcomes included continued remission with a fecal calprotectin concentration below 100 µg/g; this level was increased to less than 200 µg/g among patients who presented with 200 µg/g at baseline.

The study randomized 52 patients from 119 screened. The primary outcome was met by 7 of 26 patients (26.9%) randomized to dose escalation and by 1 of 26 control patients (3.8%; P=.0496). A greater proportion of patients in the dose-escalation group achieved a reduction in fecal calprotectin concentration to below 100 µg/g (52.6% vs 15.8%; P=.04)—or to below 200 µg/g for patients with fecal calprotectin concentrations above this level prior to randomization (76.9% vs 16.7%; P=.005)—than patients receiving placebo. At 6 weeks after randomization, the median change in fecal calprotectin concentration was -11 µg/g (interquartile range, -231 to 101) in the control group and -70 (interquartile range, -378 to 4) in the dose-escalation group (P=.28). After adjusting for baseline fecal calprotectin concentration, the dose-escalation group showed a trend toward a greater reduction at 6 weeks (P=.06). Among patients who were still in remission after the 6-week randomization phase, clinical relapse occurred more quickly among those with a fecal calprotectin concentration above 200 µg/g compared with patients showing a lower fecal calprotectin concentration (P=.01). Time to relapse was shorter for patients with a fecal calprotectin concentration below 100 µg/g (P=.09).

 

Preoperative Serologic Markers May Predict Postoperative Recurrence of Crohn’s Disease 

 

In a poster presented at Digestive Disease Week 2014, Dr Marc Ferrante of the Leuven University Hospitals in Leuven, Belgium described results from a study evaluating the use of preoperative serologic markers to predict postoperative clinical and endoscopic recurrence of Crohn’s disease (Abstract Su1349). The study enrolled 100 consecutive patients; their median age was 41.7 years, and 41 were men. All patients were undergoing ileal resection with ileocolonic anastomosis for refractory Crohn’s disease. A serum sample was collected from each patient within 1 week before surgery. All patients were followed prospectively with postoperative endoscopic evaluation at 6 months. The primary endpoint was endoscopic recurrence, and the secondary endpoint was time to clinical relapse. Blinded sera were analyzed by the enzyme-linked immunosorbent assay at an independent laboratory for the expression of the anti–Saccharomyces cerevisiae antibodies immunoglobulin A and immunoglobulin G and the atypical antibody perinuclear antineutrophil cytoplasmic antibody (pANCA), and for antibodies against 3 flagellar markers (Fla2, FlaX, and CBir1) and the outer membrane porin C. Thirty-four patients had undergone prior resection, 26 had familial IBD, and 27 were active smokers. Baseline CRP was greater than 5 mg/L in 55 patients.

Endoscopic recurrence was observed in 25 patients, and clinical relapse within 24 months of follow-up was observed in 29 patients. Based on multivariate analysis, factors independently associated with endoscopic recurrence included an anti-Fla2 antibody level greater than 66 ELISA units (EU; OR, 3.0 [95% CI, 1.1-8.7]; P=.037) and active smoking (OR, 3.1 [95% CI, 1.1-8.8]; P=.029). Several factors were independently associated with clinical recurrence: an anti-Fla2 antibody level greater than 66 EU (OR, 2.2 [95% CI, 1.0-4.6]; P=.041), pANCA positivity (OR, 2.5 [95% CI, 1.2-5.4]; P=.016), and active smoking (OR, 2.6 [95% CI, 1.2-5.5]; P=.011; Table 1). A cumulative risk score was developed by combining the 3 risk factors associated with endoscopic or clinical recurrence. It successfully predicted the likelihood of both endoscopic relapse (P<.001) and clinical relapse (P<.001), and showed that an increasing number of risk factors corresponded to a gradual increase in the relapse rate.

 

Reduced Serum and Fecal Infliximab Levels in Nonresponding Ulcerative Colitis Patients

 

Some patients with UC exhibit a primary nonresponse to infliximab, an antibody that binds to TNF-α. The mechanism that causes this reaction is not fully understood. One possible cause may be a high inflammatory load, which presents a large concentration of TNF-α for neutralization and therefore reduces the efficacy of the antibody, leading to insufficient serum antibody concentrations. At Digestive Disease Week 2014, Dr Johannan F. Brandse of the Academic Medical Center in Amsterdam, the Netherlands presented results of a pharmacokinetic study addressing the relationship between infliximab serum and fecal concentrations and efficacy (Abstract 786). This multicenter, prospective, observational study enrolled patients with moderate-to-severe UC. Patients had no prior exposure to anti–TNF-α therapy. After baseline endoscopy, patients began infliximab treatment. Serum infliximab concentrations, ATIs, and fecal samples were collected at 10 serial time points during the first 6 weeks of induction therapy. Response was defined as endoscopic improvement at weeks 6 to 8. Pharmacokinetic analysis was performed by nonlinear mixed-effects modeling and described using a 2-compartment pharmacokinetic model.

Among the 15 enrolled patients, 7 (46.7%) were receiving treatment with concomitant thiopurines. Eight patients (53.3%) showed no endoscopic improvement. Median trough levels before the third infliximab infusion during week 6 were
2.5 µg/mL for endoscopic nonresponders and 8.2 µg/mL for responders (P=.03). Serum infliximab levels of 7 µg/mL or less at day 42, defined as a cutoff point, predicted endoscopic nonresponse, with an OR of 36 (P=.03). At week 6, 4 patients developed ATIs; their antibody clearance increased by 4.4-fold and their peripheral volume of distribution decreased
by 5%. Nonresponders showed a significantly higher
concentration of fecal infliximab on day 1 (P=.02). Average post hoc area under the infliximab concentration vs the time curve was 1204±507 mg/L/day in nonresponders vs 1417±444 mg/L/day for responders (P=.42).

 

AJM300 Shows Efficacy in Ulcerative Colitis

 

Selective inhibition of lymphocyte trafficking is a promising treatment for patients with UC; however, virtually all drug candidates are monoclonal antibodies. AJM300 is an orally available small-molecule antagonist of the α4 integrin. In cells that express the α4β1/α4β7 integrin, AJM300 inhibits cellular binding to VCAM-1/MAdCAM-1. At Digestive Disease Week 2014, Dr Mamoru Watanabe of the Department of Gastroenterology and Hepatology at Tokyo Medical and Dental University in Tokyo, Japan presented findings from a randomized, placebo-controlled, phase 2a study of the α4 integrin inhibitor in UC patients (Abstract 370). The study enrolled 102 Japanese patients with moderately active UC as evidenced by a Mayo Clinic score of 6 to 10, an endoscopic subscore of 2 or less, and a rectal bleeding score of 1 or less. Patients were required to have documentation of inadequate response or intolerance to mesalamine or corticosteroids. They were randomized to receive AJM300 (960 mg) or placebo 3 times daily for 8 weeks. The primary endpoint was a clinical response at week 8, defined as a reduction in the Mayo Clinic score by 3 or more points and a decrease of at least 30% from baseline, with a concomitant decrease in the rectal bleeding subscore of at least 1 point or an absolute rectal bleeding subscore of 0 or 1. Key secondary endpoints included clinical remission and mucosal healing at week 8.

The study met its primary endpoint by showing that a greater proportion of patients in the AJM300 treatment arm achieved a clinical response at week 8 compared with placebo (62.7% vs 25.5%; OR, 5.25; 95% CI, 2.23-12.82; P=.0002). Secondary endpoints were also met, as demonstrated by clinical remission rates at week 8 of 25.3% with AJM300 vs 3.9% with placebo (OR, 7.81; 95% CI, 1.64-37.24; P=.0099) and mucosal healing at week 8 of 58.8% with AJM300 vs 29.4% with placebo (OR, 4.65; 95% CI, 1.81-11.90; P=.0014). Significant increases in peripheral lymphocyte counts at weeks 2 and 4 were observed in the AJM300 arm only, without any change in neutrophil counts. AJM300 was well tolerated, with no serious AEs or serious infections observed. This study is the first to demonstrate the clinical benefit of an oral α4 integrin antagonist for patients with IBD.

 

Third Trimester Exposure to Biologic Therapy Does Not Increase Infant Infection Rates

 

The administration of infliximab or adalimumab in pregnant women is of interest because active placental transfer of the antibodies may occur, with the drugs persisting in the newborn for up to 6 months. Certolizumab pegol is another therapeutic anti–TNF-α antibody, but it has passive placental transfer, resulting in low infant serum drug concentrations. Use of anti–TNF-α antibodies during the third trimester of pregnancy is controversial, given concerns of increased infant infections and possible alterations to the developing immune system. At Digestive Disease Week 2014, Dr Uma Mahadevan of the University of California San Francisco School of Medicine in San Francisco, California presented results from a study showing that women with IBD can safely continue anti–TNF-α therapeutic antibody monotherapy throughout pregnancy (Abstract 960).

The analysis included a prospective cohort of pregnant women who were part of the PIANO (A 1000 Patient Prospective Registry of Pregnancy Outcomes) registry at 30 US IBD centers. Patients were followed through pregnancy and the first 4 years of the child’s life. This study analyzed use of biologics at 2 different time points: in the third trimester and in the first and second trimesters together. Women who received biologic therapy in their third trimester were compared with those who had not received it during this time, and women who received biologic therapy in their first and second trimesters, but not the third, were compared with those who had not received it during the first and second trimesters. Among the 1289 women enrolled, 1097 had completed their pregnancy, resulting in 1039 live births. Biologics were used by 501 women during pregnancy. Exposure in trimester 3 was reported in 422 women; they were compared to 597 women who had not been exposed in the third trimester. Seventy patients had received biologics in trimester 1 or 2.

No increased risk was observed for women exposed to biologics vs those who were not exposed, based on preterm birth (OR, 1.3), disease activity in term 3 (OR, 0.8), postpartum disease activity (OR, 1.2), and risk of infection at month 4, 9, or 12 (ORs, 0.9, 1.1, and 1.1, respectively). However, infants who were exposed in utero to an anti–TNF-α antibody plus azathioprine/6-mercaptopurine showed higher rates of preterm birth, after adjusting for IBD activity (OR, 2.6; P<.05). In-utero exposure to combination therapy also increased the rates of preterm birth (OR, 4.9), low birth weight (OR, 6.1), and admission to neonatal intensive care (OR, 3.9) in mothers with UC.

Presentations in GERD

 

Evidence of a Significant Association Between GERD and Myocardial Infarction

 

Gastroesophageal reflux disease (GERD) affects approximately 40% of the adult population in the United States. It is known that immune cells promote the formation of atherosclerotic lesions in the coronary arteries, and numerous proinflammatory molecules are commonly elevated in patients with GERD. At Digestive Disease Week 2014, Dr Ravi K. Prakash of the MetroHealth System in Cleveland, Ohio presented results of a study investigating whether the incidence of myocardial infarction (MI) is higher in patients with GERD (Abstract 722). The case-control study collected data from the Explorys database, a private, cloud-based health data aggregator that contains health data for more than 35 million patients in the United States. The study identified an aggregated cohort of eligible GERD patients who underwent an esophagogastroduodenoscopy (EGD). A control group consisted of all patients ages 18 to 85 years without GERD who had never undergone EGD. Patients with known prior coronary or peripheral arterial disease were excluded.

There were 316,390 patients in the GERD arm and 13.6 million in the control arm. Throughout the next 5 years, a first MI occurred in 5.96% of the GERD arm (18,860 cases) vs 1.05% in the control arm (144,140 cases). Therefore, in patients without known coronary or peripheral arterial disease at baseline, GERD was associated with a 57% increased risk of having a first acute MI within the next 5 years. The incidence of new MI was higher in GERD patients who were male, obese, or had hypertension (P<.05 for all). Multivariate logistic regression analysis that adjusted for 6 major cardiovascular risk factors showed a 49% increase in relative risk (RR) of MI associated with obesity, a 90% increase associated with smoking, and a 10% increase associated with hypertension. The incidence of a new MI was highest in obese male patients with GERD and hypertension compared with the control patients (P<.0001).

 

Transoral Incisionless Fundoplication Reduces GERD Symptoms

 

Proton pump inhibitor (PPI) therapy can prevent GERD symptoms, allow healing of the esophagus, and prevent the return of esophagitis. However, a known limitation of these drugs is the incomplete control of regurgitation and extraesophageal symptoms. Transoral incisionless fundoplication (TIF) surgery is a variant of traditional fundoplication that is conducted transorally rather than laparoscopically and may serve as an alternative to the use of PPIs. At Digestive Disease Week 2014, Dr Karim S. Trad of the George Washington University School of Medicine and Health Sciences in Washington, DC presented results of a study comparing PPIs to surgery (Abstract 724). The TEMPO (Transoral Incisionless Fundoplication EsophyX vs Medical PPI Open Label) trial randomized 63 patients with small hiatal hernias at 7 community hospitals in the United States. Patients in the PPI group received the maximum standard dose. Patients in the surgery group underwent esophagogastric fundoplication. The primary
outcome was elimination of daily regurgitation or extraesophageal symptoms, which were evaluated with the GERD HRQL Scale, the Reflux Symptom Index, and the Reflux Disease Questionnaire. Secondary outcomes included normalization of distal esophageal acid exposure, PPI usage, and healing of esophagitis.

The analysis included 39 TIF patients and 21 PPI patients. All patients had objective evidence of GERD with abnormal esophageal acid exposure. They had experienced symptoms for a median of 10 years, and they had been receiving PPIs for a median of 8 years.

Use of PPIs was dramatically reduced after the TIF procedure (Figure 4). After 6 months of follow-up, regurgitation was eliminated in 97% of TIF patients vs 50% of PPI patients (RR, 1.9; 95% CI, 1.2-3.1; P=.006). Elimination of regurgitation and extraesophageal symptoms occurred in 62% of TIF patients and 5% of PPI patients (RR, 12.9; 95% CI, 1.9-88.9; P=.009). Although only 5% of patients in the PPI arm reported elimination of regurgitation and extraesophageal symptoms after the initial 6-month period, this proportion increased to 67% (14 of 21 patients) 6 months after these patients crossed over and underwent TIF (P<.001). In addition, 71% of the crossover patients were completely off PPIs 6 months after undergoing TIF. Esophageal acid exposure was normalized in 54% of TIF patients vs 52% of PPI patients (RR, 1.0; 95% CI, 0.6-1.7; P=.914). Heartburn was eliminated in 90% of TIF patients vs 13% of PPI patients (P<.001).

 

An Endoscopic Stapling Device Reduces PPI Use in GERD Patients

 

Although PPIs provide adequate control of GERD for many patients, surgery is often necessary for patients with refractory disease and is an option for patients wishing to avoid the AEs and costs associated with long-term PPI therapy. An endoscopic stapling device employs a flexible surgical stapler plus an endoscopic video camera and an endoscopic range finder that can be used to create an anterior fundoplication. At Digestive Disease Week 2014, Dr William R. Kessler of the Witham Memorial Hospital in Lebanon, Indiana presented 3-year results from a prospective trial of patients treated with the stapling device (Abstract Tu1396). Patients had GERD documented by a pH probe and had responded well to PPI therapy. Patients were placed under general anesthesia and received 2 to 3 staple quintuplets to create anterior fundoplication. Phone interviews were used to collect follow-up data, which included a minimum of 3 years’ postprocedure GERD-HRQL scores, patients’ overall satisfaction, and information on use of PPIs, histamine-2 receptor antagonists, and antacids.

Twenty-two patients underwent surgical stapling at a single institution. One patient experienced a gastrointestinal bleed 8 days after the procedure and required a blood transfusion, after which bleeding stopped spontaneously. Mean GERD-HRQL scores improved at 1, 2, and 3 years. For the 14 patients who were off daily PPIs, the mean GERD-HRQL score improved from 30.6 at baseline to 6.8 at year 3. For the 5 patients who were still receiving daily PPIs, the improvement was from 12.2 to 8.8. The proportions of patients who reported satisfaction with the procedure were 67% at 1 year, 78% at 2 years, and 63% at 3 years. Most patients remained off daily PPIs at all time points. Among the patients who continued on a PPI after the stapling procedure, mean omeprazole equivalent doses were decreased from a baseline level of 47.7 mg/day to 26.8 mg/day at 1 year, 19.7 mg/day at 2 years, and
15.9 mg/day at 3 years. Fewer than 5% of patients experienced complications, and 74% of patients remained off daily PPIs at 3 years after the procedure.

 

The Effect of Proton Pump Inhibitors on the Risk of Esophageal Adenocarcinoma in Patients With Barrett Esophagus

 

Patients with Barrett esophagus are at increased risk for esophageal adenocarcinoma (EAC). Studies presented at Digestive Disease Week 2014 offered conflicting findings regarding whether the use of PPIs decreases the risk of EAC.

Dr Srinivas Gaddam of the Washington University School of Medicine in Saint Louis, Missouri presented results from a prospective study evaluating the association between the use of PPIs and statins with the development of high-grade dysplasia and EAC (Abstract 710). The study, conducted by the Barrett’s Consortium Group, investigated a large cohort of Barrett esophagus patients undergoing screening and/or surveillance at 5 centers. The diagnosis of Barrett esophagus included intestinal metaplasia in the columnar epithelium in the distal esophagus. The primary endpoint of progression to high-grade dysplasia or EAC included both prevalent and incident cases, based on the assumption that clinical and demographic factors of patients in both categories should be similar.

The study analyzed 3635 patients, with a mean age of 60.9 years. Most patients (87.8%) were male. Patients had a mean body mass index of 28.4 kg/m² (standard deviation, 6.1 kg/m²) and a mean Barrett esophagus length of 3.5 cm (standard deviation, 3.3 cm). The study population included 506 cases with high-grade dysplasia/EAC. Univariate analysis using Chi-square and Mann-Whitney U tests revealed several significant risk factors for high-grade dysplasia/EAC: age, male sex, longer Barrett esophagus length, smoking (currently or previously), hiatal hernia, and presence of visible lesions within the Barrett esophagus segment. Protective factors included use of aspirin or nonsteroidal anti-inflammatory drugs (NSAIDs), PPIs, and statins (all P<.05). PPI use emerged as a significant protective factor (OR, 0.41; 95% CI, 0.22-0.75; P=.004) based on a statistically significant logistic regression model that adjusted for male sex, white race, Barrett esophagus length, use of H2 receptor antagonists, past or present smoking history, and aspirin or NSAID use. However, statin use did not provide an independent protective effect (OR, 0.73; 95% CI, 0.5-1.7; P=.10). Consistent with a protective effect conferred by PPI use, patients who did not develop high-grade dysplasia/EAC had a longer duration of PPI use compared with patients who developed high-grade dysplasia/EAC (4.2 years vs 3.2 years; P=.02).

Contrasting data were presented by Dr Gwen Masclee of the Maastricht University Medical Center in Maastricht, the Netherlands. Dr Masclee described a case-control study evaluating the influence of NSAIDs, statins, and PPIs on the risk of EAC in Barrett esophagus patients (Abstract Sa1837). The study cohort consisted of Barrett esophagus patients from 2 primary care databases, one in the United Kingdom and the other in the Netherlands, who were treated from 1996 through 2013. Adult patients with a diagnosis of EAC at least 1 year after the Barrett esophagus diagnosis were compared with an age-matched control group of patients without EAC. Year of Barrett esophagus diagnosis, body mass index, smoking history, and alcohol use were examined as confounding factors.

The study identified 15,134 patients, including 45 with EAC and 12 with high-grade dysplasia. Risk of EAC did not correlate to the presence of hiatal hernia or the presence of esophagitis or gastritis upon diagnosis of Barrett esophagus. Risk of EAC was increased by current smoking (OR, 2.63; 95% CI, 1.1-6.1) and current excessive alcohol use (OR, 1.9; 95% CI, 1.0-3.7). Statin use spanning 2 to 3 years yielded an OR of 0.66 (95% CI, 0.08-5.23) and an adjusted OR of 0.65 (95% CI, 0.08-5.11). For statin use of at least 3 years, the OR was 0.51 (95% CI, 0.14-1.86), and the adjusted OR was 0.48 (95% CI, 0.13-1.74). Increasing cumulative doses of statins yielded an OR of 0.58 (95% CI, 0.19-1.79) and an adjusted OR of 0.52 (95% CI, 0.17-1.63). Inclusion of the high-grade dysplasia endpoint yielded an OR for NSAID use of 0.98 (95% CI, 0.50-1.94) and an adjusted OR of 0.95 (95% CI, 0.48-1.89). For statin use of 2 to 3 years or greater than 3 years, ORs were 0.69 (95% CI, 0.09-5.43) and 0.54 (95% CI, 0.15-1.96), respectively, and adjusted ORs were 0.68 (95% CI, 0.09-5.36) and 0.54 (95% CI, 0.15-1.97), respectively. NSAID use was associated with a nonsignificant risk reduction of 2%. PPI use did not influence risk of EAC or the combined outcome of EAC and high-grade dysplasia.

 

A Novel Medical Device Reduces Symptoms of Extraesophageal Reflux

 

Extraesophageal reflux (EER) is caused by reflux of gastroduodenal contents into the laryngopharynx and results in symptoms such as chronic cough (CC), hoarseness, and globus pharyngeus. Although PPIs are often not effective in the treatment of EER, they are frequently prescribed. A novel medical device was developed to address the need for effective treatment in these patients. At Digestive Disease Week 2014, Dr Michael F. Vaezi of the Vanderbilt University Medical Center in Nashville, Tennessee presented results from a prospective study investigating the efficacy and safety of the upper esophageal sphincter (UES) assist device for patients with suspected EER (Abstract 725). The multicenter study included patients with extraesophageal symptoms and a reflux symptom index (RSI) score of greater than 13. Predominant symptoms at presentation included CC, choking, aspiration, chronic postnasal drip, globus pharyngeus, sore throat, and throat clearing. Each patient was individually fitted with a UES assist device that was adjusted to apply pressure above 20 mmHg to the external cricoid area, and external monitors ensured that this level was maintained throughout the study. The study’s primary efficacy endpoint was reduction in RSI score at 4 weeks from baseline, with safety analysis based on AEs.

Forty-seven patients completed the study protocol at
4 centers. Patients had a mean age of 50 years and a
median body mass index of 25.8 kg/m² (interquartile range, 23-29 kg/m²). Most patients (72%) were female. The most common troublesome symptoms included CC (28%), excess mucus/postnasal drip (19%), throat clearing (11%), and hoarseness (11%).

The study met its primary efficacy outcome based on a reduction in the median RSI score from 27 (95% CI, 21-32) at baseline to 12 at week 4 (95% CI, 7-19). An improvement of greater than 25% was reported by 86% of patients, and 30% of patients reported an improvement of at least 75%. Mean improvement from baseline was 54%±28.1%, with 75% of patients and 92% of providers reporting satisfaction with the UES assist device. AEs were generally mild and included device technical issues in 14% of patients and mild skin irritation in 8%.

 

Pathologic Acid Exposure Time and Baseline Impedance Correlate With Probability of Response to Proton Pump Inhibitors

 

GERD may account for up to 90% of patients with CC. Patients with CC usually have normal results on multichannel intraluminal impedance–pH monitoring (MII-pH), and they often have an unsatisfactory response to PPIs. Measurement of the esophageal impedance baseline value in patients with typical symptoms appears to increase the sensitivity of MII-pH. At Digestive Disease Week 2014, Dr Mentore Ribolsi of the Campus Bio-Medico in Rome, Italy presented results from a prospective trial that evaluated characteristics of the reflux pattern and the role of MII-pH variables in predicting PPI response (Abstract Tu1850). Patients with CC persisting for at least 8 weeks and no evidence of esophagitis were selected from 3 hospitals in Italy. Patients without respiratory diseases underwent MII-pH and thus received a double dose of PPIs for 6 weeks or longer. The Fisman cough severity/frequency score was assessed before and after MII-pH. The control group included 60 patients with nonerosive esophageal reflux disease with typical symptoms. Pathologic impedance baseline value was defined as less than 1944.

Among the 156 CC patients, 68 (43.5%) had responded to PPIs. A pathologic acid exposure time was seen in 31 responders (46%) and 15 (17%) nonresponders. Patients with CC had a higher number of reflux episodes than patients with typical symptoms, regardless of whether they responded to treatment with PPIs (58.5 and 56.5, vs 41.0, respectively). Weakly acidic refluxes occurred in a larger proportion of nonresponder CC patients (53%) than in CC responder patients (37%) and patients with typical symptoms (36%). Among the nonresponders, 43 patients (49%) had pathologic acid exposure time or impedance baseline values; 15 patients (17%) presented with pathologic acid exposure time only (P<.001). The probability of a PPI response was increased 2-fold in CC patients with a pathologic acid exposure time, a pathologic impedance baseline value, or both compared with patients who exhibited a normal acid exposure time and impedance baseline value. Dr Ribolsi and colleagues concluded that an impedance baseline value is a promising variable that increases the diagnostic yield of MII-pH in patients with CC.

 

Presentations in Endoscopy

 

Video Capsule Endoscopy Is Safe and Effective in Pediatric Patients

 

Wireless video capsule endoscopy (VCE) is a relatively noninvasive imaging modality that provides advanced gastrointestinal imaging for patients with Crohn’s disease and other conditions. Few studies have examined pediatric use of this imaging technology. At Digestive Disease Week 2014, Dr Stephen Nanton of the Avera Children’s Clinic in Sioux Falls, South Dakota described results from VCE in the diagnosis and management of gastrointestinal disorders in children (Abstract Su1747). The study included 488 consecutive children and young adults (ages 2.5-21.5 years) who had been referred for VCE. The minimum weight requirement was 14.6 kg.

Among the 511 VCE procedures performed, indications included suspected Crohn’s disease, anemia, abdominal pain, polyposis syndrome, celiac disease, and gastrointestinal bleeding. A capsule delivery device successfully placed the VCE capsule in the majority of cases. In 25 patients (4.8%), the capsule was placed using laryngoscopic assistance, with 1 failure. In 91 children, the VCE capsule was ingested in the physician’s office. The capsule retention rate was 0%.

Twenty-nine children with suspected celiac disease based on elevated tissue transglutaminase antibody levels were examined. Proximal small bowel polyps were found unexpectedly in 7 of these patients (24%). One patient was subsequently diagnosed with Bannayan-Riley-Ruvalcaba syndrome. Classic scalloping of duodenal mucosa, loss of villi, and mosaic patterning were also observed by VCE imaging. Among the 76 patients diagnosed with IBD, 2 were found to have isolated small bowel disease.

In all of the 15 children with Crohn’s disease who underwent 2 or more VCE studies to monitor response to therapy, management was modified. In 6 of these patients, VCE observations led to step-up treatment from mesalamine to 6-mercaptopurine. In 8 of 15 patients, VCE results led to the initiation of biologic therapy. In 1 patient, budesonide therapy was initiated for persistent small-bowel Crohn’s disease. In 8 of the children with Crohn’s disease (53%), endoscopic improvement was documented by VCE, and 15 patients (100%) exhibited clinical improvement after receiving step-up therapy that was initiated in response to the VCE findings. Of the 17 patients found to have polyps in the small bowel, 2 (12%) had been previously diagnosed with familial adenomatous polyposis, 3 (18%) had IBD, and 8 (47%) had celiac disease. For 1 patient, after abdominal computed tomography suggested intussusception, VCE enabled the rare discovery of duodenal windsock diverticulum.

No complications were noted with this approach. The findings underscore the safety and clinical utility of VCE in the pediatric population.

 

Endoscopic Balloon Dilation Plus Corticosteroid Injection Is Feasible in Pediatric Patients With Crohn’s Disease

 

Crohn’s disease is a chronic unremitting immune-mediated inflammation that is complicated in approximately 30% of pediatric patients by the presence of strictures on long-term follow-up. Strictures tend to recur after resection and can require repeated surgical interventions. At Digestive Disease Week 2014, Dr Roberto Gugig of Children’s Hospital Central California in Madera, California described results of a prospective pilot study in pediatric patients evaluating the use of endoscopic balloon dilation plus intramural corticosteroid injection, which is currently considered an appropriate conservative therapy in adults (Abstract Su1762). The 12 study participants had a median age of 14.5 years (range, 9-17.5 years), and 8 were male. Strictures were less than 4 cm in length; were located in the esophagus (n=1), antrum (n=1), right colon (n=4), or ileum (n=7); and were inflammatory (n=10) or fibrotic (n=3). In addition to endoscopic balloon dilation, patients received 4-quadrant triamcinolone (40 mg/mL) in aliquots of 0.5 mL to 1 mL. Technical success was defined as the ability of the endoscope to traverse the stricture after dilation.

Twelve patients underwent 15 dilations, with 4 patients (33%) requiring repeat dilations. Technical success was achieved in all patients. Improvement of symptoms was also achieved in all patients. Long-term success, defined as freedom from symptoms for longer than 1 year, was achieved in 9 patients (75%). One patient with fibrotic strictures required surgery, and no endoscopy-related complications or perforations were observed. Dr Gugig concluded that, in select pediatric patients with Crohn’s disease and strictures, endoscopic balloon dilation plus corticosteroid injection reduces the need for surgical interventions.

Electroacupuncture Decreases Analgesic Use During Endoscopic Ultrasound

 

Although electroacupuncture is a part of traditional Chinese medicine, limited data exist on its risks and benefits as an analgesic during endoscopy. At Digestive Disease Week 2014, Dr Anthony Y. Teoh of the Department of Surgery at the Chinese University of Hong Kong in Shatin, Hong Kong described results from a study investigating the efficacy of electroacupuncture in reducing analgesic consumption and diminishing patients’ perceptions of procedure-related pain (Abstract Tu2036). The prospective, double-blind, sham-controlled, randomized study included consecutive patients ages 18 to 80 years. All patients were undergoing endoscopic ultrasound for the first time; they were grade 1 or 2 according to criteria from the American Society of Anesthesiologists. Patients received electroacupuncture for approximately 45 minutes prior to the surgical procedure as well as during the procedure. Electroacupuncture was applied to acupoints associated with treating upper abdominal pain and anxiety, including Zusanli (stomach meridian ST-36), Hegu (large intestine meridian LI-4), and Nei guan (pericardium meridian PC-6). Electric stimulation was delivered at a frequency of 2 Hz, with a pulse width of 200 µs, at a stimulation intensity below the level causing discomfort. The primary outcome was the patient-controlled dose of analgesia consisting of propofol and alfentanil. Patient-related secondary outcomes included overall pain score, patient satisfaction, and willingness to repeat the procedure.

The study randomized 64 patients evenly into 2 well-balanced arms with similar numbers of patients who underwent radial, linear, or mini-probe examinations. Patients who received active electroacupuncture required a significantly lower dose of propofol (P<.001) and requested analgesics significantly less often (P<.001; Table 2). The patient pain score was reduced from 6.2 for patients receiving the sham procedure to 2.1 for those receiving electroacupuncture (P<.001), and electroacupuncture increased the patient satisfaction score from 7.1 to 8.2 (P=.002). Patients in the electroacupuncture arm were also more willing to repeat the procedure (P=.05). The endoscopists’ satisfaction scores were similar between the 2 arms (P=.110). Based on the superior results achieved with active electroacupuncture, the study was terminated early.

 

L-Menthol Spray Reduces Colonic Peristalsis and Improves the Adenocarcinoma Detection Rate

 

Although colonoscopy is a preferred method of detecting colorectal neoplasms, colonic peristalsis during the procedure can obscure some neoplasms from view. Intravenous or intramuscular injection of antispasmodic agents can alleviate the problem, but unexpected AEs may result. L-menthol, the major constituent of peppermint oil, can suppress gastric peristalsis. In a topic forum at Digestive Disease Week 2014, Dr Ken Inoue of the OSF Medical Group in Bloomington, Illinois and colleagues presented results of a prospective, single-blind, randomized, placebo-controlled study examining the value of L-menthol during colonoscopy (Abstract 740).

Patients scheduled for colonoscopy were randomized to receive L-menthol (n=118) or placebo (n=108). Patients in the active arm received 20 mL of a 1.6% L-menthol solution sprayed directly on the cecum and through the working channel by means of an endoscope. The L-menthol treatment was repeated if peristalsis resumed. The primary outcome was the difference in the rate of adenoma detection, defined as the proportion of screened subjects in whom at least 1 adenomatous lesion was identified. The secondary outcome was the proportion of patients with no peristalsis. Peristalsis was graded on a scale from 0 (none) to 3 (severe).

Baseline data were similar between the 2 arms. The adenoma detection rate was significantly higher in the L-menthol arm compared with the placebo arm (60.2% vs 42.6%; P=.0083). The proportion of patients with no peristalsis was significantly higher after treatment with L-menthol compared with the placebo group (71.2% vs 30.9%; P<.0001). In the L-menthol treatment arm, peristalsis scores were significantly reduced after treatment (P<.0001); in contrast, in the placebo group, peristalsis scores did not differ significantly before vs after treatment. No AEs were observed in either arm.

 

A New Biopsy Technique Improves Diagnosis of Barrett Esophagus

 

Barrett esophagus is characterized by esophageal goblet cell metaplasia. It is typically diagnosed using the modified Seattle protocol technique, in which 4-quadrant forceps biopsies are taken at sites no further than 2 cm
apart throughout the Barrett esophagus segment. However, because this method leaves a large amount of unsampled tissue, areas of advanced dysplasia or neoplasia may remain unexamined. Studies with wide area transepithelial sampling with 3-dimensional analysis (WATS3D) have demonstrated improved detection of Barrett esophagus and dysplasia. The new technique collects a disaggregated tissue specimen from a wide area and surveys the entire thickness of the epithelial section being analyzed, thus including the submucosa.

At Digestive Disease Week 2014, Dr Seth A. Gross of the New York University School of Medicine in New York, New York described a study of this new biopsy technique for Barrett esophagus. The retrospective study estimated the adjunctive yield of WATS3D compared with forceps biopsy across numerous community-based gastroenterology practices (Abstract 371). The study included patients with GERD and proven or possible Barrett esophagus. The upper endoscopies were performed by 28 gastroenterologists. WATS3D biopsies were obtained with the standard 2-brush technique, with additional forceps biopsies obtained afterward during the sample endoscopy. All biopsy samples were analyzed at a central laboratory.

The analysis included 2498 specimen sets. Most of the patients (60%) were female, and the mean age was 55 years (range, 15-97 years). GERD was the most common indication for endoscopy. In 80% of patients, the Barrett esophagus length was less than 3 cm. Forceps biopsy resulted in the identification of Barrett esophagus in 277 cases (15.1%). Dysplasia was seen in 17 patients (4.51% of Barrett esophagus cases and 0.68% of patients who underwent forceps biopsies). Adjunctive use of WATS3D led to the identification of an additional 258 Barrett esophagus cases, increasing the diagnostic sampling yield to 25.4% and representing a 68.4% increase in the identification of this condition. In addition, WATS3D enabled the detection of an additional 10 cases of dysplasia and 1 case of cancer, all of which were missed by forceps biopsy. In total, the use of WATS3D increased the detection of dysplasia and neoplasia by 64.7%, leading to an overall detection rate of 1.12%. Dr Gross and colleagues concluded that the use of WATS3D in community-based gastroenterology practices improves patient care by identifying dysplasia missed by standard forceps biopsy.

 

Peroral Endoscopic Myotomy and Laparoscopic Heller Myotomy Are Equivalent for Treating Idiopathic Achalasia

 

Achalasia is a rare primary esophageal disorder marked by the absence of peristalsis and an inability of the smooth muscle fibers of the esophagus to relax. Symptoms may include regurgitation, coughing, choking, and esophagitis. The gold standard surgical treatment for achalasia is laparoscopic Heller myotomy (LHM). Peroral endoscopic myotomy (POEM) is a newer approach that may offer an effective and less invasive alternative. However, no consensus exists regarding which of these 2 therapies provides the greatest safety and efficacy for achalasia patients, and cost comparisons have not been published. At Digestive Disease Week 2014, Dr Vivek Kumbhari of the Johns Hopkins University School of Medicine in Baltimore, Maryland presented results of a retrospective study examining the efficacy, safety, and costs of LHM and POEM in treating patients with idiopathic achalasia (Abstract Tu1354).

This single-center study evaluated consecutive patients who underwent LHM or POEM between 2008 and 2013. All LHMs included a Toupet fundoplication and were performed transabdominally. All POEMs were performed by a single gastroenterologist in the endoscopy unit. Clinical response was defined by symptom improvement and a decrease in Eckardt stage to 0 or 1.

Twenty-one POEM and 66 LHM cases met the inclusion criteria, and baseline demographics and preprocedure characteristics were similar between both groups, as was the myotomy length. Procedure time was significantly curtailed by use of POEM (136 minutes vs 277 minutes; P<.01). The 2 procedures yielded similar clinical responses (P=.45), at a median follow-up of 2.6 months for POEM and 1.9 months for LHM. Both procedures were associated with similar lengths of hospitalization following surgery (P=.08). Rates of mild and moderate AEs were also similar between the 2 groups, and no severe complications were observed. The 2 procedures incurred similar mean total charges for the duration of the hospital stay. Dr Kumbhari and colleagues concluded that POEM and LHM are associated with similar safety, efficacy, and costs.

Underwater Endoscopic Mucosal Resection Is Feasible for Scarred Polyp Removal

 

Resection of colon polyps often leads to the formation of scar tissue, complicating future surgeries. Recurrence rates exceeding 25% have been reported. Underwater endoscopic mucosal resection effectively removes large colorectal polyps and offers a promising approach for the removal of scarred polyps. At Digestive Disease Week 2014, Dr Chris M. Hamerski of the California Pacific Medical Center in San Francisco, California described results of a study of polyp removal by underwater endoscopic mucosal resection (Abstract Tu1454). The study evaluated the feasibility and efficacy of underwater endoscopic mucosal resection performed at a single center for the removal of polyps 10 mm or larger and characterized by scarring from previous resection attempts. Mucosal resection was performed under full water immersion using a snare. En bloc resection was performed when feasible, and hot biopsy forceps were used to remove any remaining tissue that was not removable by snare.

The study included 58 patients who underwent colonoscopy for endoscopic mucosal resection of a polyp with a prior resection attempt. The mean polyp size was
20 mm (range, 10-20 mm). The polyp locations included the right colon (72%), the rectum (16%), and the descending or sigmoid colon (12%). The median resection time was 5 minutes (range, 1-175 minutes). Forty-three percent of polyps were removed en bloc, with an additional 43% removed in pieces using the snare. Ten percent of cases required additional resection with biopsy forceps, and 1 polyp (2%) required use of argon plasma coagulation. For 1 polyp with significant scarring, complete resection was attempted, but the case was referred to surgery.

The majority of polyps (86%) were adenomas without high-grade dysplasia. Malignancy was observed in 4 polyps (7%); of these, 2 were removed en bloc, and pathology revealed negative margins. The other 2 polyps
were surgically examined, and no residual adenoma or cancer was found. Four polyps showed evidence of high-grade dysplasia. Follow-up data were available from
20 patients (34%), at a median follow-up of 26 weeks. One patient (5%) had residual adenoma on follow-up colonoscopy, and 1 patient developed delayed bleeding requiring hospitalization.