Comprehensive Reports on the Latest Advances in Gastroenterology and Hepatology From:
• The 50th Annual Meeting of the European Association for the Study of the Liver
April 22-26, 2015 • Vienna, Austria
• Digestive Disease Week 2015
May 16-19, 2015 • Washington, DC
Presentations in Hepatology
Daclatasvir/Sofosbuvir Plus Ribavirin in Patients With Advanced Cirrhosis or Liver Transplantation
Combinations of ribavirin, pegylated interferon, and direct-acting antiviral agents have dramatically improved overall outcomes in patients with hepatitis C virus (HCV) infection. However, patients with advanced cirrhosis and those who have undergone liver transplant are still in need of oral therapies that have a shorter duration and are without drug interactions. At the 50th annual meeting of the European Association for the Study of the Liver (EASL), Fred Poordad, MD, of the University of Texas Health Science Center in San Antonio, Texas presented findings from the phase 3, open-label ALLY-1 study (Abstract LO8). The study evaluated 12 weeks of treatment with daclatasvir (60 mg once daily), sofosbuvir (400 mg once daily), and ribavirin in patients with advanced cirrhosis or recurrent HCV after liver transplant. The ribavirin dose was adjusted from 600 mg daily to 1000 mg daily based on hemoglobin levels and creatinine clearance.
The 60 advanced cirrhosis patients and 53 post-transplant patients had a median age of 59 years (range, 19-82 years), and approximately two-thirds were men. In the cirrhosis and posttransplant cohorts, respectively, 40% and 42% of patients were treatment-naive and 75% and 77% had HCV genotype 1 infection. The posttransplant patients had METAVIR fibrosis classifications of F0 to F2 (43%), F3 (25%), or F4 (30%). In the cirrhotic cohort, patients had cirrhosis of Child-Pugh class A (20%), B (53%), or C (27%), and the model for end-stage liver disease (MELD) score for the cohort ranged from 8 to 27.
After treatment with daclatasvir, sofosbuvir, and ribavirin, the rates of sustained virologic response at 12 weeks (SVR12) were 83% for the advanced cirrhosis cohort and 94% for the posttransplant cohort. Of the patients with HCV genotype 1 infection, 82% of the cirrhotic patients and 95% of posttransplant patients achieved SVR12. Among the HCV genotype 3 cohort, SVR12 rates were 83% in the 6 (10%) patients with advanced cirrhosis and 91% in the 11 (21%) patients who were posttransplant. All 4 patients with advanced cirrhosis whose treatment was interrupted for a liver transplant achieved SVR12 following posttransplant treatment, including 1 patient who received no therapy after transplant. The most common adverse events (AEs) of any grade were headache, fatigue, anemia, diarrhea, and nausea.
Longer Retreatment With Ledipasvir/Sofosbuvir Is Feasible in Patients Who Failed 8 or 12 Weeks of Treatment
Patients who fail shorter treatment with direct-acting antiviral agents may benefit from retreatment with the same drugs given for a longer time. At the EASL meeting, Eric Lawitz, MD, of the University of Texas Health Science Center in San Antonio, Texas presented efficacy and safety results from an open-label, single-arm, phase 2 study investigating retreatment with the once-daily fixed-dose combination of ledipasvir (90 mg) and sofosbuvir (400 mg) for 24 weeks in patients who had failed the same drug combination given for 8 or 12 weeks in phase 2 or 3 trials (Abstract O005).
The 41 patients had a median age of 58 years (range, 35-71 years), and 83% were men. Eighty-three percent of patients had HCV genotype 1a infection, and the mean HCV RNA level was 6.2 log10 IU/mL (range, 4.5-7.4 log10 IU/mL). Of the 19 (46%) patients with cirrhosis, 79% had baseline NS5A resistance-associated variants (RAVs). Prior treatment duration was 8 or 12 weeks for 30 and 11 patients, respectively, and NS5A RAVs were present in 63% and 100% of these patients, respectively.
After 4 or 8 weeks of treatment, HCV RNA was undetectable in 95% and 100% of patients, respectively. One virologic breakthrough occurred at week 16, and several patients experienced relapse after cessation of treatment. The SVR4 and SVR12 rates were 73% and 71%, respectively. SVR12 rates for noncirrhotic vs cirrhotic patients were similar (68% vs 74%; Figure 1). However, a greater difference in SVR12 rates emerged for prior treatment duration of 8 weeks vs 12 weeks (80% vs 46%, respectively).
Baseline NS5A RAVs were associated with a SVR12 rate of 60% vs a rate of 100% in the absence of these variants. Moreover, baseline NS5A RAVs were more common in patients whose prior treatment lasted 12 weeks. Patients with 0, 1, or at least 2 baseline NS5A RAVs yielded SVR12 rates of 100%, 69%, and 50%, respectively, and specific baseline RAVs were associated with SVR12 rates ranging from 33% for baseline Y93H/N mutations (n=6) to 100% for baseline Q30R (n=4) or M28T (n=1). The majority of AEs were of mild or moderate severity. Two serious AEs occurred and were considered unrelated to study treatment. No treatment discontinuations were attributed to an AE, and no deaths occurred during the study.
Sofosbuvir/Ledipasvir and Ribavirin Treatment Demonstrates Efficacy in Patients Who Have Hepatitis C Virus Infection and Advanced Liver Disease
Effective and safe treatments are needed for patients with chronic HCV infection and advanced cirrhosis, particularly those with decompensated cirrhosis. At the EASL meeting, Michael Manns, MD, of the Hannover Medical School in Hannover, Germany presented preliminary safety and efficacy findings from the international SOLAR-2 (GS-US-337-0124) study, which investigated a single-tablet, 2-drug regimen of sofosbuvir (400 mg) and ledipasvir (90 mg) plus ribavirin taken for 12 or 24 weeks (Abstract G02). SOLAR-2 enrolled treatment-naive or -experienced patients with HCV genotype 1 or 4 infection at 34 sites in 12 countries. All patients had either decompensated liver disease or recurrent HCV infection following a liver transplant. The study included 168 posttransplant patients who had no cirrhosis, based on METAVIR stage F0 to F3, or had compensated cirrhosis, based on Child-Turcotte-Pugh class A. The study also included 160 patients with decompensated cirrhosis, based on Child-Turcotte-Pugh class B or C, of whom 53 were posttransplant. Patients in each cohort were randomized 1:1 to 12 or 24 weeks of treatment. Patients with a Child-Turcotte-Pugh score of 13 or higher were excluded.
The median age of the entire study population was approximately 59 years (range, 27-79 years), approximately three-fourths were men, and approximately 11% of patients had HCV genotype 4 infection. For the cohort of F0 to F3 and Child-Turcotte-Pugh class A patients, the SVR12 rates were 95% and 98% for 12 vs 24 weeks of treatment, respectively. For the patients with decompensated cirrhosis, SVR12 rates were 85% and 88% for 12 vs 24 weeks of treatment, respectively. For the patients with HCV genotype 1 infection, 12 vs 24 weeks of treatment yielded SVR12 rates of 96% vs 98% for the noncirrhotic or compensated cirrhosis patients and 88% vs 89% for patients with decompensated cirrhosis. The analysis included only 36 patients with HCV genotype 4 infection, thus precluding meaningful conclusions for this group. In the patients with decompensated cirrhosis, virologic response was accompanied by improvements in MELD and Child-Turcotte-Pugh scores and by improved levels of bilirubin and albumin. Treatment-related serious AEs were reported in 0% to 5% of patients in each of the 4 arms, and no deaths were considered treatment-related.
Favorable Rates of Sustained Virologic Response at 12 Weeks Observed After 8 Weeks of Treatment With Grazoprevir, Elbasvir, and Sofosbuvir
A shortened duration of treatment is desirable for patients with HCV infection. At the EASL meeting, Fred Poordad, MD, of the University of Texas Health Science Center in San Antonio, Texas presented results from the open-label, phase 2 C-SWIFT trial, which investigated a shortened duration of treatment with a daily fixed-dose tablet of grazoprevir (100 mg) and elbasvir (50 mg), plus daily sofosbuvir (400 mg) in patients with HCV genotype 1 or 3 infection, with or without cirrhosis (Abstract O006).
The 102 patients with HCV genotype 1 infection received 4, 6, or 8 weeks of treatment, with treatment durations of 4 or 6 weeks for noncirrhotic patients and 6 or 8 weeks for cirrhotic patients (Table 1). The 41 patients with HCV genotype 3 infection received 8 or 12 weeks of treatment, with treatment durations of 8 or 12 weeks for noncirrhotic patients and 12 weeks for cirrhotic patients. The median age across the 7 treatment arms ranged from 42 to 57 years, with a median age of approximately 52 years for the entire study group. Cirrhosis was reported in 37% of the patients.
For HCV genotype 1 patients, SVR12 rates for noncirrhotic patients after 4 vs 6 weeks of treatment were 33% vs 87%. For cirrhotic patients after 6 vs 8 weeks of treatment, SVR12 rates were 80% vs 94%. For HCV genotype 3 patients, SVR12 rates for noncirrhotic patients after 8 vs 12 weeks of treatment were 93% vs 100%, and the cirrhotic patients achieved a SVR12 rate of 91% after 12 weeks of treatment. All virologic failures were due to relapse, and the majority of relapsing patients had either wild-type HCV infection at relapse or RAVs that were present at baseline. The novel drug combination was generally well tolerated. The most common AEs were headache, fatigue, and nausea. One cirrhotic patient discontinued treatment due to an AE, and 2 cirrhotic patients experienced a serious AE.
Systematic Review and Meta-Analysis of Hepatitis B Virus Reactivation in Patients Receiving Chemotherapy for Solid Tumors
For patients receiving chemotherapy for solid tumors, the risk of reactivation of hepatitis B virus (HBV) has not been established. At the Digestive Disease Week (DDW) 2015 meeting, Sonali Paul, MD, of the Tufts Medical Center in Boston, Massachusetts presented results of a systematic review and meta-analysis investigating HBV reactivation rates with or without prophylaxis in patients with solid malignancies and in chronic vs resolved HBV infection (Abstract 730).
The analysis included studies through March 31, 2015 available from MEDLINE, PubMed, Web of Science, Cochrane Central Register of Controlled Trials, TOXNET, and Scopus. The researchers included 22 studies of chronic HBV infection, representing 1578 patients, and 3 studies of resolved HBV infection, representing 322 patients. Twenty-four studies were observational, 1 was a randomized controlled trial, and 21 studies were from Asian countries. Patients had a median age of 47 years (range, 20-80 years) and had cancer of the breast (n=844), head and neck (n=335), gastrointestinal tract (n=274), or another area (n=125).
In patients with chronic HBV infection, the rate of reactivation without antiviral prophylaxis was 22% (95% CI, 17%-28%), with substantial heterogeneity among the included studies (Cochrane I2, 81%). The rate of reactivation in patients given antiviral prophylaxis was 4.6% (95% CI, 2.2%-7.9%), and heterogeneity was lower (I2, 61%). Forest plot analysis demonstrated that prophylaxis reduced the risk of reinfection in HBV patients receiving chemotherapy (odds ratio [OR], 0.14; 95% CI, 0.08-0.24; I2, 5%). The risk of reinfection ranged from 14% to 24% based on tumor type and from 3.3% to 32% based on type of chemotherapy. Prophylaxis decreased the rates of hepatitis (OR, 0.19; 95% CI, 0.11-0.33) and delayed chemotherapy (OR, 0.13; 95% CI, 0.05-0.37). None of the patients with resolved HBV infection received prophylaxis; their rate of reinfection was 2.2% (95% CI, 0%-8.6%), and heterogeneity was low (I2, 15%).
Infection-Related Acute-on-Chronic Liver Failure as a Predictor of Mortality in Patients With Cirrhosis
In patients with chronic cirrhosis, infection reduces survival. Infection-related acute-on-chronic liver failure (I-ACLF) status is defined as 2 or more extrahepatic organ failures and is used to classify patients with cirrhosis and sepsis to enable more effective intervention. At the DDW meeting, Jacqueline O’Leary, MD, of the Baylor University Medical Center in Dallas, Texas presented results of a study whose aim was to validate the ability of I-ACLF status to predict 30-day mortality in cirrhotic patients with infection (Abstract 591). The multicenter study examined data from prospectively enrolled patients in the North American Consortium for the Study of End-Stage Liver Disease database of tertiary care hepatology centers. The study included 336 infected patients and 449 uninfected patients who were hospitalized with a complication of cirrhosis. Assessments were performed for organ failure based on the presence of shock, grade 3/4 hepatic encephalopathy, need for dialysis, or need for mechanical ventilation.
Patients had a mean age of 57 years, 64% were men, and 43% had alcohol-induced cirrhosis. Mean baseline levels of serum creatinine, sodium, MELD scores, and Child-Pugh scores were higher in the infected patients, while mean baseline serum albumin levels and blood pressure were lower. Thirty-day survival decreased with an increasing number of organ failures, regardless of the presence of infection. Any single organ failure was correlated with a lower rate of survival for infected patients compared with uninfected patients. Based on multivariable modeling, the area under the curve for mortality was 0.81 for the entire group, 0.82 for infected patients, and 0.73 for uninfected patients. I-ACLF was the strongest predictor of 30-day survival (OR, 0.10; 95% CI, 0.06-0.22) for infected as well as uninfected patients after controlling for baseline MELD score, white blood cell count, serum albumin level, and infection status.
Obeticholic Acid for the Treatment of Primary Biliary Cirrhosis
Primary biliary cirrhosis (PBC) is a chronic, autoimmune, cholestatic liver disease that destroys the bile ducts and can lead to liver transplant or death. Disease progression is marked by increasing levels of plasma alkaline phosphatase and bilirubin. Obeticholic acid is a modified bile acid and agonist of the farnesoid X receptor that has demonstrated efficacy and tolerability in patients with PBC. Its mechanism of action differs from that of ursodeoxycholic acid (UDCA), which yields an inadequate response or is intolerable in as many as half of patients but is the only drug currently approved to treat PBC. At the DDW meeting, Kris Kowdley, MD, of the Swedish Medical Center in Seattle, Washington presented findings of an integrated analysis of results from 2 placebo-controlled phase 2 studies (747-201 and 747-202) and the phase 3 POISE study that evaluated obeticholic acid with or without UDCA in patients with PBC (Abstract 657).
The analysis included a total of 335 patients who had an inadequate response to or were unable to tolerate UDCA. The POISE study had a composite endpoint of alkaline phosphatase level of less than 1.67 times the upper limit of normal, with a minimum 15% reduction and a normal bilirubin level. Significantly more patients achieved the composite endpoint after treatment with obeticholic acid compared with placebo or UDCA monotherapy in each of the 3 studies (P<.05 for each). In the pooled analysis, 8% of the 124 placebo-treated patients and 45% of the 306 patients treated with obeticholic acid achieved the composite endpoint (P<.0001). Obeticholic acid treatment in the 3 studies was associated with significantly improved levels of alkaline phosphatase, bilirubin, gamma-glutamyl transpeptidase, alanine transaminase, and aspartate transaminase from baseline to the end of treatment. Treatment was generally well tolerated. The most common AE was pruritus, which occurred in 63% of patients treated with obeticholic acid (10 mg daily).
Presentations in IBD
Withdrawal of Anti–Tumor Necrosis Factor α Therapy in Patients Who Have Inflammatory Bowel Disease
In the United Kingdom, reassessment of inflammatory bowel disease (IBD) activity after 12 months of therapy is mandated by The National Institute for Health and Care Excellence and the Scottish Medicines Consortium. The guidelines mandate discontinuation of therapy in most patients who demonstrate clinical remission and mucosal healing. At the DDW meeting, Nicholas Kennedy, MD, of the Western General Hospital in Edinburgh, United Kingdom presented results of a retrospective study of relapse rates in IBD patients following withdrawal of anti–tumor necrosis factor α (anti-TNFα) therapy and results of a systematic review of related literature (Abstract 288).
Included patients had a confirmed diagnosis of IBD and at least 12 months of prior anti-TNFα therapy. The trial’s primary endpoint was moderate-to-severe relapse at 12 months. The study included 146 patients with Crohn’s disease (CD) and 20 with ulcerative colitis (UC) or unclassified IBD. Approximately 85% of patients were being treated with infliximab, and the remainder were being treated with adalimumab. The median time on anti-TNFα therapy was 28 months (range, 14-45 months), and the median follow-up was approximately 2 years. All patients were in clinical remission at the time of therapy cessation, and 62.5% had normal levels of C-reactive protein (CRP), fecal calprotectin, and other laboratory parameters. Complete mucosal healing was observed in 89 of the 102 patients, with endoscopy performed prior to treatment cessation.
Relapse rates at 1 and 2 years were 36% and 56% for patients with CD and 42% and 47% for patients with UC or unspecified IBD. Of the patients who relapsed within 12 months, 24 patients were given systemic corticosteroids, 7 were admitted to the hospital, and 2 underwent resectional surgery. For the CD patients, multivariable analysis modeled without fecal calprotectin identified the risk factors of age at diagnosis of less than 22 years (hazard ratio [HR], 2.29; 95% CI, 1.35-3.88) and white blood cell count greater than 5.25 × 109/L (HR, 2.06; 95% CI, 1.11-3.80 × 109/L). For the 42 CD patients with available fecal calprotectin data, multivariable analysis identified levels of fecal calprotectin greater than 50 μg/g as a risk factor (HR, 3.04; 95% CI, 1.26-7.37). A systematic review and meta-analysis yielded relapse rates of 39% for patients with CD (HR, 0.39; 95% CI, 0.35-0.43) and 37% for patients with UC (HR, 0.37; 95% CI, 0.28-0.46).
Relapse After Withdrawal of Anti–Tumor Necrosis Factor α Therapy in Patients With Crohn’s Disease
Patients in remission from CD could benefit from cessation of anti-TNFα therapy; however, as many as 50% of patients relapse. At the DDW meeting, Martin Bortlik, MD, of the Charles University in Prague, Czech Republic presented results from a study designed to elucidate risk factors for relapse, including the importance of deep remission (Abstract 283).
The prospective, observational study included consecutive patients from a single center who discontinued anti-TNFα therapy at the time of corticosteroid-free, clinical and endoscopic remission. Deep remission was defined as the presence of endoscopic and clinical remission, with a calprotectin level of less than 150 mg/kg and a CRP level of 5 mg/L or lower. Relapse was defined as clinical worsening confirmed either by imaging, laboratory parameters, or the new onset of perianal disease.
The 61 enrolled patients had a median age of 31 years (range, 15-65 years), and 59% were women. The median duration of anti-TNFα therapy at the time of discontinuation was 23 months (range, 4-73 months), and 77% of patients continued immunosuppressive therapy after stopping anti-TNFα treatment. Median follow-up after cessation of anti-TNFα therapy was 28 months (range, 7-47 months).
Thirty-two (53%) patients relapsed, with a median time to relapse of 8 months (range, 1-25 months). Remission was maintained for 6, 12, or 24 months in 82%, 59%, and 51% of patients, respectively. The relapse rates in 27 patients with deep remission vs 23 patients without deep remission were similar (P=.84). In multivariate analysis, patients with colonic disease were less likely to relapse compared with patients who had ileal or ileocolonic disease at 1 year (OR, 0.15; 95% CI, 0.03-0.72; P=.02) and at the end of follow-up (HR, 0.30; 95% CI, 0.10-0.87; P=.03). Factors that did not influence the risk of relapse included the type of anti-TNFα therapy, the type of immunosuppressive therapy, corticosteroid use, and levels of calprotectin or CRP. Sixteen of the 32 patients were re-treated with anti-TNFα agents, and 87.5% exhibited a response. Of the remaining patients, 4 underwent surgery and 12 received a different type of treatment for their CD.
Infliximab Vs Placebo For Patients With Crohn’s Disease Following Ileocolonic Resection
At the DDW meeting, Miguel Regueiro, MD, of the University of Pittsburgh Medical Center in Pittsburgh, Pennsylvania presented results of the PREVENT trial, which examined the efficacy of infliximab vs placebo for preventing clinical recurrence in patients with CD following intestinal resection with ileocolonic anastomosis (Abstract 749). The primary endpoint of the randomized, double-blind, placebo-controlled trial was the prevention of clinical recurrence of CD by week 76 of postsurgical treatment.
Key patient inclusion criteria included documented diagnosis of CD, CD activity index of less than 200, and ileocolonic resection within 45 days prior to enrollment. Clinical recurrence was defined as a CD activity index of greater than 200 with a greater-than-70-point increase from baseline, endoscopic recurrence, and a negative stool test result for C difficile infection; development of a new fistula or redraining of an existing fistula or abscess also constituted clinical recurrence. Two hundred ninety-seven patients were randomized to receive infliximab (5 mg/kg) or placebo treatment. Patients had a median CD activity index of 106. Medications in use at baseline included corticosteroids (4.7%), immunomodulators (17.5%), and aminosalicylates (18.5%), and 22.6% of patients had a history of anti-TNFα use. Patients in the 2 treatment arms had similar risk factors for recurrence.
After 76 weeks of treatment, the placebo and infliximab treatment arms showed clinical recurrence rates of 20.0% and 12.9%, respectively (P=.097). The time to clinical recurrence was not significantly different between the 2 treatment arms (P=.141). For the secondary endpoint of endoscopic recurrence, with data handling rules applied, the infliximab and placebo arms demonstrated recurrence in 60.6% and 30.6% of patients, respectively (P<.001).
In the placebo vs infliximab arms, 8.9% vs 29.4% of patients discontinued study treatment due to an AE, and at least 1 infusion reaction was reported in 8.2% vs 19.4% of patients. Noted study limitations included a clinical recurrence rate that was lower than expected, enrollment of a low-risk study population, and subjectivity of the CD activity index.
Serum Neutrophil Gelatinase-Associated Lipocalin and Matrix Metalloproteinase 9 as a Marker of Mucosal Healing in Patients With Crohn’s Disease
Examination of serum markers of mucosal healing could provide a less costly alternative to endoscopy for patients with CD. At the DDW meeting, Magali de Bruyn, PhD, of the Rega Institute for Medical Research in Leuven, Belgium described results of a study investigating serum levels of neutrophil gelatinase-associated lipocalin and matrix metalloproteinase 9 (NGAL-MMP-9) as a surrogate marker for mucosal healing in patients with CD (Abstract Su1219).
The study included 108 patients with active CD and 43 healthy controls. All participants received an infliximab infusion, and NGAL-MMP-9 levels were determined by sandwich enzyme-linked immunosorbent assay (ELISA) before the first infliximab infusion and up to 5 years after. Endoscopic healing was defined as the complete absence of ulcerations; partial healing was defined as significant endoscopic improvement with ulcerations present. Histologic healing was defined as the absence of epithelial damage.
Of the 108 patients with active CD, complete or partial endoscopic healing was observed in 38 (35%) and 34 (31%) patients, respectively, and no endoscopic healing was observed in 36 (33%) patients. At baseline, the median level of NGAL-MMP-9 was 77.6 ng/mL (interquartile range [IQR], 36.9-141.0 ng/mL) in patients with CD vs 25.5 ng/mL (IQR, 17.8-42.8) in healthy controls (P<.001). After infliximab treatment, NGAL-MMP-9 levels decreased significantly relative to baseline in the 72 patients with healing (69.0 ng/mL; IQR, 32.6-135.5 ng/mL vs 35.2 ng/mL; IQR, 9.4-56.1 ng/mL; P<.001). Serum levels of the marker also decreased in the 36 patients with CD who did not exhibit endoscopic healing, from 100.9 ng/mL (IQR, 43.4-152.6 ng/mL) at baseline to 43.8 ng/mL (IQR, 27.0-96.8 ng/mL) after treatment (P<.002). However, the decrease was significantly more profound in patients with complete healing (P=.020). Levels of NGAL-MMP-9 correlated with the level of neutrophils (P<.001) and CRP (P<.001), endoscopic activity (P<.001), and histologic activity (P<.001).
Based on receiver operating characteristic analysis, a serum NGAL-MMP-9 cutoff level of 26.4 ng/mL corresponded with complete endoscopic healing and histologic healing. Sensitivity and specificity were 58% and 85%, respectively, for endoscopic healing and 63% and 84%, respectively, for histologic healing. The results compared favorably to those associated with CRP, for which a cutoff of 5 mg/mL showed specificities of 58% and 52% for endoscopic healing or histologic healing, respectively.
Statins Provide a Protective Effect Against New-Onset Inflammatory Bowel Disease
Exposure to medication has been identified as a possible trigger for the onset of IBD. At the DDW meeting, Ryan Ungaro, MD, of the Icahn School of Medicine at Mount Sinai in New York, New York described results of a study that investigated the relationship between exposure to specific classes of medication and the risk of new-onset IBD (Abstract 7).
The matched case-control study identified patients from a national medical claims and pharmacy database. The study included adult patients with international classification of diseases (ICD)-9 codes of 555.x for CD and 556.X for UC between January 2008 and December 2012. Using propensity scoring, each IBD patient had up to 10 matched controls based on age, sex, race, and state. Controls had no ICD-9 codes for CD, UC, or IBD-associated diseases and no prescriptions for IBD medications. New-onset IBD was defined as having at least 3 separate CD or UC ICD-9 codes and at least 1 year with no IBD-related codes.
Consistent with prior publications, the analysis found that the new onset of IBD was more likely after exposure to trichomonacides (OR, 1.89; 95% CI, 1.83-1.95), other antibiotics (OR, 2.45; 95% CI, 2.18-2.76), injectable estrogens (OR, 1.66; 95% CI, 1.09-2.51), and transdermal estrogens (OR, 1.28; 95% CI, 1.18-1.39), as well as after exposure to aminoglycosides and quinolones. However, combinations of progesterone plus estrogen showed a protective effect (OR, 0.86; 95% CI, 0.83-0.90), as did oral contraceptives without estrogen (OR, 0.88; 95% CI, 0.78-0.98). Notably, the analysis identified HMG-CoA reductase inhibitors, including simvastatin, atorvastatin, and pravastatin, as having a protective effect against the new onset of IBD (OR, 0.68; 95% CI, 0.66-0.69). The protective effect was observed for CD (OR, 0.63; 95% CI, 0.61-0.65) and for UC (OR, 0.72; 95% CI, 0.69-0.74).
Recurrent Clostridium difficile Infection in Patients With Inflammatory Bowel Disease
Clostridium difficile infection (CDI) is more common in IBD patients than in the general population. At the DDW meeting, Roshan Razik, MD, of Mount Sinai Hospital in Toronto, Ontario described results from the RECIDIVISM study, which examined the incidence of CDI and recurrent infection in IBD (Abstract 594).
The investigators retrospectively reviewed records from the Division of Infectious Diseases CDI Database between 2010 and 2013 to identify IBD patients who had recurrent CDI (2 or more incidents of CDI), 1 CDI incident, and non-IBD patients. The review identified 54 (11%) patients with CD, 56 (11%) patients with UC, and 393 (78%) patients with no bowel disease. The entire study population had a mean age of 58.8 years, and 61.4% of the subjects were women.
Compared with the non-IBD population, IBD patients with CDI had a lower age (39 years vs 64 years; P<.001), used more corticosteroids (39.1% vs 12.0%; P<.001) and more immunosuppressive medications (42.7% vs 13.2%; P<.001), and were more likely to have had a prior bowel resection (28.2% vs 11.5%; P<.001). Recurrent CDI was noted in 32% of IBD patients vs 24% of non-IBD patients (P<.01), and the time to recurrence was longer for the IBD vs non-IBD patients (213.7 months vs 124.0 months; P<.001). IBD patients were more likely to undergo colectomy due to infection with C difficile (6.4% vs 0.3%; P<.001). CDI-attributable mortality occurred in 12.2% of the non-IBD population vs 0% of the IBD population (P<.001).
Based on multivariate analysis, statistically significant predictors of recurrent CDI in the IBD population included recent hospitalization and recent use of antibiotics, 5-aminosalicylic acid, corticosteroids, biologic therapy, or immunosuppressants (P≤.001 for each). CDI was more likely to recur in patients with nonileal CD (P<.001), and IBD patients without a prior bowel resection were more likely to experience CDI recurrence (P=.020; Table 2).
Zinc Intake and the Risk of Inflammatory Bowel Disease
Zinc has effects on autophagy, the innate and adaptive immune response and maintenance of the intestinal barrier. Based on results in cell culture, the role of zinc in metalloproteinases and other enzymes, and the association between zinc deficiency and more severe colitis, dietary zinc may play a role in the development of IBD. At the DDW meeting, Ashwin Ananthakrishnan, MD, of Massachusetts General Hospital in Boston, Massachusetts described results of a prospective study that investigated the association between zinc intake and the risk of CD and UC within 2 cohorts of women (Abstract 5). Dietary zinc levels were determined by means of the validated semiquantitative food frequency questionnaire, which was administered every 4 years, and included food sources of zinc as well as zinc supplements and multivitamins.
The study population included 170,809 women from the Nurses Health Studies I and II, representing 3,317,550 person-years of follow-up. The investigators identified 269 incident cases of CD, yielding an incidence of 8 per 100,000 person-years, and 338 cases of UC, yielding an incidence of 10 per 100,000 person-years. Zinc intake ranged from 9 mg daily in the lowest quintile to 33 mg daily in the highest quintile. Increased zinc intake was associated with a reduced risk of CD. Compared with the lowest quintile of zinc intake, HRs for quintiles 2, 3, 4, and 5 (representing increasing amounts of zinc intake), respectively, were 0.92 (95% CI, 0.65-1.29), 0.60 (95% CI, 0.40-0.89), 0.57 (0.38-0.86), and 0.74 (95% CI, 0.50-1.10). In contrast, no significant relationships were observed between zinc intake and risk of UC.
Compared with patients with daily zinc intake of less than 8 mg daily, a reduced risk of CD was observed for those with daily zinc intake of 8 to 16 mg daily (HR, 0.69; 95% CI, 0.44-1.08) and for those with daily zinc intake of greater than 16 mg daily (HR, 0.52; 95% CI, 0.32-0.86). Results remained robust after adjusting for dietary vitamin C, phytates, or calcium.
Presentations in Endoscopy
Outcomes of a 5-Year Series of Peroral Endoscopic Myotomy
Peroral endoscopic myotomy (POEM) is an endoscopic procedure for the treatment of motility disorders of the esophagus. The procedure is related to natural orifice transluminal endoscopic surgery and was developed to deliver the long-term benefits of surgical myotomy by using flexible endoscopes, thus avoiding the costs and complications associated with surgery. At the DDW meeting, Stavros Stavropoulos, MD, of Winthrop University Hospital in Mineola, New York described results from a 5-year, prospective series of POEM performed in patients with achalasia at a single tertiary care center (Abstract 181).
The 204 patients in the series had a mean age of 53 years (range, 10-93 years), including 16 patients over 80 years. Prior achalasia treatment included pneumatic balloon dilation (17%), suboptimal balloon dilation (22%), botulinum toxin (20%), and Heller myotomy (15%). Seventy-three percent of patients had stage I or II achalasia, and 10% of patients had stage III achalasia. Of the 17% of patients with stage IV disease, 34 of 35 patients had sigmoid disease, with an esophageal diameter ranging from 6 to 13.5 cm.
After POEM treatment, the Eckardt score decreased from a mean of 7.7 (range, 4-12) at baseline to a mean of 0.2 (range, 0-3) at the 3-month assessment (P<.0001). Clinical success, defined as an Eckardt score of 3 or less, was observed in 97% of patients at the 3-, 6-, and 12-month follow-up assessments. Lower esophageal sphincter pressure decreased from a baseline mean of 42.9 mm Hg (range, 5.4-111 mm Hg) to a mean of 18.3 mm Hg (range, 0-50 mm Hg) at the 3-month follow-up assessment (P<.0001).
Of the 70 patients who were tested for barium emptying at 5 minutes, all exhibited greater than 50% emptying, and 40 (57%) patients exhibited 100% emptying. The mean hospital stay was 1.8 days (range, 1-6 days). Six POEM failures occurred in the series, and 5 of these patients were successfully treated with salvage therapy. No deaths occurred. Minor AEs were observed in 7 (3%) patients, and 2 (1%) patients required readmission within 30 days of POEM. Objective gastroesophageal reflux disease (GERD) assessment at 3 months showed 39 (40%) of 98 patients with reflux esophagitis and 36 (37%) of 97 patients with a positive pH test result.
Prophylactic Clip Placement After Snare Polypectomy
Delayed gastrointestinal bleeding is a known complication of snare polypectomy. Some reports suggest that gastrointestinal bleeding can be prevented by the use of prophylactic clips. At the DDW meeting, Tarun Rai, MD, of the Kansas City Veterans Administration Medical Center in Kansas City, Missouri presented results from a systematic review and meta-analysis that investigated the role of prophylactic clip placement after snare polypectomy to prevent delayed gastrointestinal bleeding (Abstract 329).
Two independent reviewers conducted a search of PubMed, EMBASE, the Cochrane Central Register of Controlled Trials and Database of Systematic Reviews, and recent abstracts from major conference proceedings, using the search terms polyp, bleeding, colonoscopy, and clips. Randomized and nonrandomized studies comparing prophylactic clip placement with no clip placement were included, and pooled data were generated for delayed postpolypectomy bleeding, overall complications, and perforation rates.
The analysis included 2 randomized and 6 case-control studies comprising 12,108 patients and 8354 polypectomies. In polypectomies with vs without prophylactic clip placement, postpolypectomy bleeding was observed in 1.1% vs 0.82% of patients, respectively. Analysis of pooled data demonstrated no significant differences between the 2 groups in terms of rates of delayed bleeding (OR, 1.14; 95% CI, 0.31-4.17; P=.84; I2, 79%), overall complications (OR, 0.64; 95% CI, 0.28-1.46; P=.29; I2, 45%), or perforation (OR, 1.41; 95% CI, 0.25-8.09; P=.70; I2, 0%).
Risk Factors for Bleeding After Polypectomy in Patients on Antithrombotic Agents
Few studies have investigated the relationship between the use of anticoagulants and bleeding after polypectomy, and there is no consensus regarding optimal timing for re-introducing anticoagulants after endoscopy. At the DDW meeting, David Lin, MD, of the Veterans Affairs Palo Alto Health Care System in Palo Alto, California presented results of a systematic review of bleeding complications following polypectomy performed by 24 endoscopists at 2 tertiary care centers between January 2004 and June 2012 (Abstract 331). The study’s aim was to investigate the incidence and risk factors of bleeding after polypectomy for patients on antithrombotic therapies.
The study included consecutive patients undergoing elective colonoscopy who were on antithrombotic agents and experienced hematochezia within 1 month after polypectomy. Three matched control cases were selected for each bleeding patient based on antithrombotic agent, polypectomy technique, procedure date, and study site. The overall incidence of postpolypectomy bleeding was 1.2% (95% CI, 0.91%-1.54%). The highest incidence of bleeding was observed in the 87 patients on heparin (14.9%; 95% CI, 8.20%-24.4%), followed by aspirin (0.92%; 95% CI, 0.64%-1.26%; n=3933), clopidogrel (0.84%; 95% CI, 0.27%-2.00%; n=595), and warfarin only (0.66%; 95% CI, 0.18%-1.69%; n=604).
The 59 patients with postpolypectomy bleeding and the matched controls had similar baseline characteristics, with a median age of approximately 65 years and similar rates of hypertension, heart disease, and kidney disease. Patients who experienced bleeding after polypectomy were more likely to have a polyp of 2 cm or larger (41% vs 10%; P<.001), and the mean size of the largest polyp was greater in patients who experienced bleeding after polypectomy (13.9 ± 14.6 vs 7.3 ± 4.3; P<.001).
The study identified several characteristics that were more common among patients who bled after polypectomy vs those who did not, such as restarting medication within 1 week of polypectomy (75% vs 34%; OR, 4.5; 95% CI, 2.2-9.4; P<.001), having at least 1 polyp measuring 2 cm or greater (41% vs 10%; OR, 5.9; 95% CI, 2.7-12.9; P<.001), or having multiple large polyps (20% vs 8%; OR, 2.9; 95% CI, 1.1-7.3; P=.001). Cautery in the right colon was also more common in patients who bled after polypectomy (76% vs 55%; OR, 2.6; 95% CI, 1.3-5.5; P=.004). The rate of prophylactic clip placement was similar between the 2 cohorts (P=.32).
Chromoendoscopy for Detecting Dysplasia in Patients With Long-Standing Ulcerative Colitis
Patients with UC have an increased risk of colorectal cancer compared with the general population. In comparison to white-light endoscopy, chromoendoscopy can provide images with improved resolution and better contrast through the concomitant use of dyes, such as indigo carmine and methylene blue. At the DDW meeting, Venkat Subramanian, MD, of the Saint James University Hospital in Leeds, United Kingdom presented results of a randomized, parallel-group study evaluating rates of dysplasia detection using high-definition white-light endoscopy (HDWLE) vs high-definition chromoendoscopy (HDCE) in patients with long-standing UC (Abstract 446).
High-definition equipment was used for all procedures. A 0.2% indigo carmine spray was applied on withdrawal using a dye spray catheter, and targeted and quadrantic random biopsies were taken from each colonic segment in all study participants. Included patients had UC of more than 8 years’ duration extending proximal to the splenic flexure and requiring surveillance colonoscopy. The study randomized 53 patients to HDWLE and 50 patients to HDCE. Patients in both arms had a median age of approximately 55 years and a mean duration of disease of approximately 22 years. Both arms were generally well matched in terms of family history of colorectal cancer, smoking status, and medication use.
The number of patients with dysplasia detected was significantly higher in the HDCE arm vs the HDWLE arm (22% vs 9.4%; P=.04). Six low-grade lesions were detected with HDWLE, whereas HDCE detected 1 high-grade and 13 low-grade lesions. The mean number of dysplastic lesions detected per patient was also higher with HDCE (0.26 ± 0.6 vs 0.12 ± 0.4; P=.04). Mean withdrawal time was significantly higher in the HDCE arm than in the HDWLE arm (21.2 ± 5.8 minutes vs 13.6 ± 3.3 minutes; P<.001). No dysplastic lesions were detected in any of the random biopsies.
Endoscopic Mucosal Resection for the Removal of Large and Flat Lesions of the Colon
Large and flat colonic lesions are at increased risks of bleeding, perforation, and incomplete removal with traditional snare resection. Although patients with these types of lesions are usually referred to surgery, advanced endoscopic techniques could provide a less invasive alternative. At the DDW meeting, Gottumukkala Raju, MD, of the MD Anderson Cancer Center in Houston, Texas presented results of a study that investigated endoscopic mucosal resection of large and flat lesions in patients referred to a single center between 2009 and 2012 (Abstract 1062).
The study included 205 patients with lesions of presumed benign pathology. Patients had a mean age of 64 years (range, 29-88 years), and approximately half of the patients were men. The majority of lesions (61%) were located in the right colon. Patients had lesions that were less than 10 mm (25%), 10 to 20 mm (28%), or greater than 20 mm (47%).
Of the 205 patients, 8 (3.8%) had an optical diagnosis of cancer, underwent biopsy confirmation, and were referred to surgery; 40 (19.3%) patients could not undergo the endoscopic procedure for various reasons, including 16 patients in whom access to the lesion was difficult; and 159 (76.8%) patients experienced a successful endoscopic procedure. Complications occurred in 6 (3.7%) patients and included delayed bleeding in 5 patients, of whom 3 required hospitalization, and 1 patient with a perforation that was successfully closed without surgery.
Surveillance colonoscopy was performed in 139 patients at 3 to 6 months after endoscopic mucosal resection and in 60 patients at 18 months after resection. In 6 patients, residual tumor was observed and removed by repeating the endoscopic procedure. Precancerous lesions observed during surveillance included adenomas (2.78%) and serrated polyps (0.84%).
A Novel, Edible Preparation for Effective Bowel Cleansing for Colonoscopy
Proper preparation for colonoscopy is essential for accurate results, yet as many as 25% of colonoscopy preparations are inadequate. Many patients fail to complete cleansing due to poor taste and the large volume of fluid that must be ingested. At the DDW meeting, Campbell Levy, MD, of the Dartmouth-Hitchcock Medical Center in Lebanon, New Hampshire presented findings from a study that evaluated a novel, nutritionally balanced, low-residue diet incorporating a purgative agent that was developed to replace fasting and liquid cleansing (Abstract Su1525). The study’s primary aim was to determine the feasibility of the novel bowel cleansing preparation in patients undergoing colonoscopy.
Adults scheduled for colonoscopy for colorectal cancer screening or surveillance of polyps at a single center were eligible. Meals, snacks, and drinks blended with polyethylene glycol 3350, sorbitol, and ascorbic acid were provided. The protocol specified stopping the study after enrollment of 30 patients or after 8 of 10 patients, enrolled in groups of 5, achieved a global colon preparation assessment of good or excellent.
The study was halted after all 10 of the first 10 enrolled patients had successful colon cleansing, including 1 patient with a global assessment of excellent and the remainder with an assessment of good. The 10 patients had a median age of 61 years (range, 46-73 years). Indications included screening for colorectal cancer (70%) and adenoma surveillance (30%). Eight patients completed 100% of the preparation, and 2 patients completed 95% of the preparation.
All patients gave the experimental preparation an overall tolerability rating of 1 on a 5-point scale, indicating no problems with the preparation. All 10 patients also indicated a willingness to take the preparation again for a subsequent procedure. No significant changes in electrolyte or creatinine levels were observed, and no AEs were reported.
Presentations in IBS
Validation of a Serologic Test for Diarrhea-Predominant Irritable Bowel Syndrome
Due to a lack of disease-specific biomarkers, diarrhea-predominant irritable bowel syndrome (IBS-D) is diagnosed by excluding other diagnoses, such as IBD and celiac disease. Animal models injected with Campylobacter jejuni produce host antibodies against cytolethal distending toxin B (CdtB) that can crossreact with vinculin in the host gut, inducing symptoms similar to those of IBS, and the crossreactivity has been replicated in human colon tissue samples. At the DDW meeting, Mark Pimentel, MD, of the Cedars-Sinai Medical Center in Los Angeles, California presented results of a large study that evaluated antibodies to CdtB and vinculin as biomarkers for IBS-D (Abstract 311).
The study recruited 2375 IBS-D patients who met the Rome III criteria from the multicenter TARGET 3 trial. The study also recruited 142 patients with IBD, 121 patients with celiac disease, and 43 healthy subjects. Plasma levels of anti-CdtB and antivinculin antibodies were determined by ELISA to evaluate differences in antibody levels among the study cohorts.
Levels of anti-CdtB antibodies and antivinculin antibodies were significantly higher in IBS patients compared with all other cohorts in aggregate and compared with each non-IBS cohort (P<.00001 for all comparisons). A key goal of the study was to optimize the ability of the antibody test to distinguish IBS-D from IBD. An optical density cutpoint of 2.80 was identified for the anti-CdtB ELISA, which conferred a specificity of 91.6%, sensitivity of 43.7%, and likelihood ratio of 5.2. For antivinculin, the optimal optical density cutpoint of 1.68 was identified, with a specificity of 83.8%, sensitivity of 32.6%, and likelihood ratio of 2.0.
Successful Retreatment With Rifaximin for Symptomatic Diarrhea-Predominant Irritable Bowel Syndrome
Abdominal pain, bloating, and urgency are common symptoms of patients with IBS-D. At the DDW meeting, William Chey, MD, of the University of Michigan in Ann Arbor, Michigan presented results of a multiphase, randomized, controlled trial that assessed the efficacy of rifaximin retreatment in patients who responded to an initial course of the same antibiotic (Abstract 313).
The study enrolled patients with IBS-D who met Rome III criteria and had mean severity scores of at least 3 for abdominal pain (on a scale of 0-10) and at least 3 for bloating (on a scale of 0-6). Patients also had at least 2 stools that were type 6 or 7 (ie, loose or watery) on the Bristol Stool Scale during the 7-day baseline evaluation. After this evaluation, eligible patients received open-label rifaximin (550 mg) 3 times daily.
Patients who experienced symptom recurrence within 18 weeks were randomized to the double-blind portion of the study, in which patients received treatment with rifaximin (550 mg) or placebo 3 times daily; patients received the same treatment during two 2-week periods separated by 10 weeks without treatment. Abdominal pain, bloating, stool consistency, and urgency were assessed by daily questionnaires. Efficacy was defined as patients with a positive response during at least 2 of 4 weeks following treatment, and recurrence was defined as a loss of response for at least 3 of 4 weeks following treatment. The primary endpoint was the proportion of patients experiencing a decrease of at least 30% in mean abdominal pain and a concomitant decrease of at least 50% in the number of days per week with stool consistency matching Bristol Stool Scale type 6 or 7.
For the first repeat treatment, 308 patients were randomized to placebo and 328 patients were randomized to rifaximin; for the second repeat treatment, 283 were randomized to placebo and 295 were randomized to rifaximin. Prior to repeat treatment, the 2 arms were well balanced in terms of demographics and IBS-D symptoms. The trial met its primary endpoint, with rifaximin yielding a significantly higher proportion of composite responders than placebo after the first retreatment (35% vs 25%; P=.005). Significant improvements were also observed for rifaximin compared with placebo for specific IBS-D symptoms after the first retreatment period, including abdominal pain, stool consistency, bloating, and urgency (P<.05 for each). Rifaximin also yielded a greater proportion of composite responders after the second retreatment (37% vs 29%; P=.0375). Rifaximin treatment was generally well tolerated.
Tenapanor for the Treatment of Constipation-Predominant Irritable Bowel Syndrome
Tenapanor is an investigational small molecule that reduces sodium and phosphate uptake from the gut. In a phase 2a trial, tenapanor improved symptoms of constipation-predominant IBS (IBS-C). At the DDW meeting, William Chey, MD, of the University of Michigan in Ann Arbor, Michigan presented results of a double-blind, randomized, placebo-controlled phase 2b trial that evaluated the efficacy and safety of tenapanor for the treatment of IBS-C (Abstract 1020).
Enrolled patients were adults up to 75 years of age who had IBS-C, as defined by Rome III criteria, and active disease during the screening period. Patients were randomized to receive 12 weeks of twice-daily treatment with tenapanor, dosed at 5 mg, 20 mg, or 50 mg, or placebo. The primary endpoint was the complete spontaneous bowel movement (CSBM) responder rate, defined as the proportion of patients with an increase of at least 1 CSBM per week from baseline for at least 6 of the 12 treatment weeks.
The 356 randomized patients had a mean age of 45.7 years, and 87% were women. The 4 treatment arms were well balanced for baseline disease characteristics assessed during screening. After treatment, the CSBM responder rates for placebo vs 5 mg, 20 mg, or 50 mg of tenapanor were 33.7%, 40.2%, 43.7%, and 60.7%, respectively, and reached significance for the highest dose of tenapanor vs placebo (P<.001; Figure 2). Overall responders experienced a decrease in abdominal pain of at least 30% and an increase of at least 1 CSBM per week for at least 6 of the 12 treatment weeks. Placebo vs 5 mg, 20 mg, or 50 mg of tenapanor yielded overall responder rates of 23.6%, 25.3%, 33.3%, and 50.0%, respectively, with the final result achieving significance relative to placebo (P<.001). Compared with placebo, tenapanor (50 mg twice daily) also significantly improved stool consistency (1.0 vs 2.2; P<.001) and yielded a greater proportion of patients with an abdominal pain response (65.5% vs 48.3%; P=.026).
Most AEs were mild to moderate in severity. Three serious AEs occurred, none of which were considered treatment-related. The most common AE was diarrhea, occurring in 11.2% of patients receiving tenapanor at the highest dose. Tenapanor concentration was below the lower limit of quantification in more than 97% of serum samples, and the highest concentration measured was 1.03 ng/mL.
Characterization of Irritable Bowel Syndrome Symptoms After Ingestion of Fructose or Inulin
Fermentable carbohydrates can induce IBS symptoms; however, few studies have shown a direct association between IBS symptoms and ingestion of specific carbohydrates. At the DDW meeting, Giles Major, MD, of the University of Nottingham in Nottingham, United Kingdom presented results of a double-blind crossover study that examined the relationship between ingestion of fructose or inulin and the development of clinically important symptoms (Abstract 246).
Enrolled patients were adults 65 years or younger with IBS, as defined by Rome III criteria, and bloating. On each of 3 study days, patients received a drink that included either 40 g of fructose, 40 g of inulin, or 40 g of glucose. All drinks were flavored with lime and looked identical, and study days were scheduled at least 1 week apart to avoid crossover. The primary endpoint was the composite score for the symptoms of flatulence, bloating, pain, and diarrhea, with each symptom scored on a scale of 0 to 3. An increase of at least 3 points in the composite score relative to baseline was considered symptomatic. Other outcomes included breath hydrogen, small bowel water content, and colonic volume and gas, with the latter 2 assessed by magnetic resonance imaging.
Twenty-nine patients, of whom 22 were women, completed all 3 study treatments. The median age was 34 years (range, 25-48.5 years). Symptom intensity was greater after ingestion of fructose or inulin compared with glucose (P<.05 for both). In the responding patients, the time to peak symptom intensity was approximately 1 hour after fructose ingestion vs approximately 4 hours with inulin (P<.005). Median breath hydrogen levels increased from 1 ppm (range, 0.5-3 ppm) after glucose ingestion to 18 ppm (range, 5-50 ppm; P<.005) after fructose ingestion and to 66 ppm (range, 31.5-125.5 ppm; P<.005) after inulin ingestion.
Median colonic volume increased from 56 mL ± 59 mL with glucose to 142 mL ± 140 mL (P<.05) with fructose and to 265 mL ± 191 mL with inulin (P<.005). Fructose induced a greater increase in small bowel water content compared with glucose (P<.005), and both test carbohydrates caused a greater increase in colonic gas compared with glucose (P<.05 for fructose; P<.005 for inulin). In symptomatic patients, peak IBS symptoms were most strongly correlated with small bowel water content after fructose ingestion (R=.51) and with colonic gas after inulin ingestion (R=.57).
Efficacy of Lubiprostone in Constipation-Predominant Irritable Bowel Syndrome Patients With Different Baseline Pain Scores
Lubiprostone is a selective chloride channel activator that has demonstrated efficacy in treating symptoms of IBS-C. At the DDW meeting, Lin Chang, MD, of the David Geffen School of Medicine at the University of California, Los Angeles in Los Angeles, California presented results of a post hoc analysis of data from 2 trials in IBS-C evaluating the efficacy of lubiprostone and the impact of baseline abdominal pain on response rates (Abstract Mo1262).
The post hoc analyses included data from 2 double-blind, randomized, placebo-controlled, phase 3 trials of lubiprostone in patients with IBS-C. Response rates were determined by a scale ranging from 0, indicating no symptoms, to 4, indicating very severe symptoms. Eligible patients had a baseline pain score of at least 1.36. Responders were defined as having a mean pain reduction of at least 30% compared with baseline and an increase of at least 1 SBM per week for 6 of 12 treatment weeks. The efficacy of lubiprostone on abdominal pain alone was also evaluated.
In study 1, composite response rates were 25.5% (39 of 153 patients) with lubiprostone vs 13.0% (10 of 77 patients) with placebo (P=.0338). In study 2, composite response rates were 27.2% (37 of 136 patients) with lubiprostone and 17.4% (15 of 86 patients) with placebo (P=.3275). Using a baseline pain score of at least 1.36, the pooled data showed that lubiprostone significantly improved response rates (26.3% vs 15.3%; P=.0079). The improvement in response rate with lubiprostone was observed for patients with a baseline pain score of at least 1.5 (P=.0180) or at least 2.0 (P=.0069). No significance was observed for baseline pain scores of at least 2.5 or at least 3.0, but data from a limited number of patients were available for these analyses.
Presentations in GERD
5-Year Results of a Magnetic Device for the Treatment of Gastroesophageal Reflux Disease
In March 2012, the US Food and Drug Administration approved the LINX magnetic device, which is surgically implanted at the lower esophageal sphincter for the treatment of GERD. The device is constructed of linked, expandable magnetic beads and mechanically restores antireflux competency to the esophagus. At the DDW meeting, Robert Ganz, MD, of the University of Minnesota in Minneapolis, Minnesota presented final results of a 5-year, prospective, multicenter study evaluating the efficacy and safety of the device (Abstract 688).
The study enrolled 100 patients at 13 centers in the United States and 1 center in the Netherlands. Included patients were age 18 to 85 years and had at least 6 months of typical heartburn, with or without regurgitation. Patients were on proton pump inhibitors (PPIs) and were partially responsive to daily PPI therapy. Patients had pathologic acid exposure and no esophageal motility disorders. Clinical efficacy and tolerability were assessed by comparing each patient’s symptoms and AEs before vs after implantation of the device.
Five-year data were available for 85 patients. GERD health-related quality-of-life (HRQL) scores improved over 5 years, with median scores decreasing from 27 without PPI therapy to 11 with PPI therapy at baseline and to 4 after magnetic sphincter augmentation without PPI therapy. Outcomes that improved at 5 years vs baseline with patients not taking PPI therapy included dissatisfaction related to reflux (from 95% to 7.1%), moderate or severe heartburn (from 89% to 11.9%), and moderate or severe regurgitation (from 57% to 1.2%). The proportion of patients taking daily PPIs decreased from 100% at baseline to 15.3% at 5 years, and the proportion of patients taking double doses of PPIs decreased from 36% to 2%. The incidence of esophagitis decreased from 40% at baseline to 16% at 5 years.
The device exhibited acceptable safety, with no mucosal breaks related to the device, no device erosion or migration, and no operative AEs. The most important AE was dysphagia, which was typically mild, and affected 68% after implantation vs 7% after 5 years; however, daily, bothersome dysphagia was reported by 5% of patients at baseline and 6% of patients at 5 years. Device removal was performed in 7 patients due to dysphagia (n=4), ongoing reflux (n=2), or vomiting (n=1).
Cost-Effectiveness of Corticosteroid Therapy Vs the 6-Food Elimination Diet for the Treatment of Eosinophilic Esophagitis
First-line treatment for eosinophilic esophagitis can consist of swallowed topical corticosteroids or the 6-food elimination diet, which involves removing milk, egg, wheat, soy, nuts, and seafood from a patient’s diet. Both approaches involve significant costs: corticosteroids are inherently expensive, and the 6-food elimination diet requires multiple sequential endoscopies to assess response and identify triggers. At the DDW meeting, Cary Cotton, MD, of the University of North Carolina at Chapel Hill in Chapel Hill, North Carolina presented results of a study examining the cost-effectiveness of these 2 treatment options (Abstract 4).
The authors performed a decision analysis invoking a Markov model for the choice between corticosteroid therapy and the 6-food elimination diet as first-line therapy. Patients receiving corticosteroid therapy accumulated costs for 1 post-diagnosis endoscopy to assess response to treatment and 2 months of an induction dose of corticosteroids comprising either fluticasone (1760 μg daily) or budesonide (2 mg daily), plus ongoing treatment at half of the induction dose for responders. Patients with a response to dietary therapy underwent 7 endoscopies. The cohort was limited to 10,000 patients, and crossover to the alternative treatment was allowed.
In the base case analysis, at 1 year after initiating treatment, patients in the corticosteroid treatment arm had accumulated $4765.17 in costs vs $3035.66 for the patients with dietary restriction, and both arms had accumulated 0.98 quality-adjusted life-years (QALYs). The reduced cost and similar accumulation of QALYs continued through 5 years, at which point the accumulated costs were $13,231.43 for initial corticosteroid therapy, with 4.70 QALYs, and $8817.59 for initial diet therapy, with 4.71 QALYs. In the vast majority of probabilistic sensitivity analyses, the 6-food elimination diet was cost-effective over corticosteroid therapy. The study was limited by a 5-year time horizon, reflecting the available data.
Risk Factors for Progression to Esophageal Adenocarcinoma for Patients With Barrett Esophagus and Low-Grade Dysplasia
Patients with Barrett esophagus and low-grade dysplasia are at risk of progression to esophageal adenocarcinoma. However, estimates of progression rates are limited by heterogeneity in pathologic assessments. At the DDW meeting, Rajesh Krishnamoorthi, MD, of the Mayo Clinic in Rochester, Minnesota presented results of a study that examined the rates and predictors of progression in patients with Barrett esophagus and low-grade dysplasia (Abstract Sa1067).
Included patients had Barrett esophagus with a histologic diagnosis of low-grade dysplasia and were part of a large, prospective registry in a tertiary care center. Progressors were defined as patients who developed high-grade dysplasia or esophageal adenocarcinoma more than 12 months after the index date. The study assessed several risk factors relating to patient demographics, endoscopic and histologic findings, and use of medications, with univariate and multivariate analyses performed to identify predictors of progression.
The mean age of the 337 included patients was 63 years, and 84% were men. Mean follow-up was 7 years. Twenty-one (6.2%) patients progressed to high-grade dysplasia/esophageal adenocarcinoma, and the annual incidence of progression to high-grade dysplasia/esophageal adenocarcinoma was 0.8%. Based on univariate analysis, factors that increased the risk of progression included increasing Barrett esophagus segment length (HR, 1.140; 95% CI, 1.004-1.294; P=.0434) and the presence of nodules (HR, 5.54; 95% CI, 1.61-19.03; P=.0065). Longer follow-up duration after the diagnosis of low-grade dysplasia, which likely reflected the persistence of low-grade dysplasia, was associated with a reduced risk of progression (HR, 0.780; 95% CI, 0.680-0.894; P=.0003).
Multivariate analysis confirmed increased duration of follow-up (HR, 0.763; 95% CI, 0.661-0.880; P=.0002) and increasing Barrett esophagus segment length (HR, 1.211; 95% CI, 1.049-1.397; P=.0088) as risk factors for progression. In addition, younger age at diagnosis of low-grade dysplasia (HR, 0.954; 95% CI, 0.919-0.989; P=.0110) was associated with the risk of progression. The results suggest that patients with a longer Barrett esophagus segment length and younger age when diagnosed with low-grade dysplasia may be candidates for intensive surveillance or endoscopic therapy.
Use of a Sleep-Positioning Device to Reduce Symptoms of Nocturnal Gastroesophageal Reflux Disease
Nocturnal acid reflux can damage the esophagus and has been associated with the development of esophagitis. The MedCline sleep-positioning device was designed to improve upper body elevation and lateral positioning of the left side of the body, allowing patients to sleep in an inclined position. At the DDW meeting, Sanath Allampati, MD, of the Cleveland Clinic in Cleveland, Ohio presented results from a prospective trial that examined the ability of the MedCline sleep-positioning device to improve nocturnal symptoms of GERD (Abstract Mo1137).
Enrolled patients were on antisecretory medications with continued frequent moderate-to-severe nocturnal heartburn and regurgitation. Upon enrollment and at the end of the study, patients completed the GERD-HRQL Questionnaire and the Nocturnal GERD Symptom Severity and Impact Questionnaire (N-GSSIQ), which queries nocturnal GERD symptoms, morning impact of GERD, and concerns about nocturnal GERD. Patients were instructed to continue taking antisecretory medication and to sleep on the sleep-positioning device for at least 6 hours each night for 2 weeks.
Of the 25 patients, 24 were taking PPIs and 1 patient was taking a histamine-2 receptor antagonist. At the end of the 2-week study period, mean N-GSSIQ scores had significantly improved relative to baseline (54.6 vs 17.0; P<.001). Improvement occurred in all 3 subsections of the questionnaire, including nocturnal GERD (33.6 vs 9.3; P<.001), morning impact of nocturnal GERD (6.2 vs 2.2; P<.001), and concern about nocturnal GERD (14.7 vs 5.5; P<.001). Improvement from baseline also occurred in the 16 patients who completed the GERD-HRQL Questionnaire (28.4 vs 16.3; P<.001). No AEs were reported.
Bone Structure and Metabolism in Users of Long-Term Proton Pump Inhibitor Therapy
Based on observational studies that have demonstrated an association between the use of PPIs and bone fractures resulting from osteoporosis, PPI prescribing information carries a black-box warning regarding possible increased risk of bone fracture with their use. At the DDW meeting, Laura Targownik, MD, of the University of Manitoba in Winnipeg, Manitoba presented results of a study that used quantitative 3-dimensional computed tomography (CT) to investigate bone structure in patients who had taken long-term PPI medication (Abstract 781).
The study initially identified all persons residing in Manitoba, Canada over the age of 50 years who had either taken PPIs continuously for at least 5 years or had not taken any PPIs during the previous 5 years. A random sample of patients from these groups was contacted for participation in the study. Bone mineral density of the hip and spine was assessed by standard 2-dimensional, dual-energy X-ray absorptiometry, and the left hip was scanned by 3-dimensional CT. Blood was collected to assess levels of bone metabolism markers.
The analysis included 52 patients with long-term PPI use and 52 controls. The 2 groups demonstrated similar demographic characteristics. The median age was 65 years, and approximately half of the patients were women. The mean T-score for the total hip bone, assessed by dual-energy X-ray absorptiometry, was -0.32 ± 1.01 for the PPI users and -0.29 ± 1.16 for the control group. T-scores for the femoral neck, intertrochanteric area, and L1-L4 vertebrae were similar for both cohorts. A T-score of less than -1 for the entire hip bone was observed in 30 (58%) of the PPI users and 29 (56%) of the control patients.
The 2 groups also showed similar levels of bone metabolism markers in the blood, although gastrin levels differed significantly between the PPI users vs nonusers (106 ± 70 ng/L vs 27.5 ± 19.6 ng/L; P<.001; Table 3). In the PPI users vs the controls, 3-dimensional CT yielded similar levels of bone mineral density for the total hip bone (341 ± 59 mg/cm3 vs 333 ± 59 mg/cm3), the cortical bone (1090 ± 95 mg/cm3 vs 1075 ± 107 mg/cm3), and trabecular bone (119 ± 22 mg/cm3 vs 112 ± 19 mg/cm3), and measures of bone strength were similar for both groups.
Sex Differences in Proton Pump Inhibitor Dose Requirements for the Treatment of Gastroesophageal Reflux Disease
Women on long-term PPI therapy have significantly higher fasting and postprandial gastrin levels than men (Helgadottir et al. Dig Liv Dis. 2014;46[2]:125-130). Therefore, men might require higher PPI doses than women for managing symptoms of GERD. At the DDW meeting, Holmfridur Helgadottir, MD, of the National University Hospital of Iceland in Reykjavik, Iceland presented results of a study investigating the relationship between PPI dose and gastrin levels in men vs women (Abstract Mo1153).
The study enrolled 49 women and 51 men with erosive esophagitis. Patients had a median age of 59 years. PPI therapy included omeprazole (n=28), rabeprazole (n=41), and esomeprazole (n=38). All patients received omeprazole for 2 months during the study and were randomized to continue at the same dose or to take half of their previous dose by weight for 2 months. Forty-nine patients were randomized to continue with the same PPI dose, 34 were randomized to reduce the daily PPI dose from 20 mg to 10 mg, and 17 were randomized to reduce the daily dose from 40 mg to 20 mg. Failures were defined as patients who left the study and resumed their prior medication. Two patients were excluded from the analysis.
Female patients had significantly higher baseline gastrin levels than male patients (76 pg/mL vs 50 pg/mL; P=.017). The failure rate was 2 times higher in the dose reduction group vs the control group (24% vs 13%). From baseline to the end of study treatment, median gastrin levels decreased from 71 pg/mL to 53 pg/mL (P=.002) for the entire dose reduction group and increased slightly in the control group (52 pg/mL vs 56 pg/mL; P=.04). Among the patients randomized to dose reduction, 3 (12%) of the female patients failed to complete the 2 months of reduced dose therapy vs 9 (36%) of the male patients (P=.09). Of the women taking 20 mg or 40 mg at baseline, 13% and 11%, respectively, failed to successfully complete 2 months of therapy at the lower dose. Median age, median baseline gastrin levels, proportion with Helicobacter pylori infection, and body mass index were similar for the patients with successful vs unsuccessful dose reduction.