At this year's Digestive Disease Week, important announcements were made and clinical advances discussed in every gastroenterologic specialty. Here, Gastroenterology & Hepatology summarizes some of the most important presentations, with expert commentary from opinion leaders in each area.
Presentations in GERD
Abstract 753 The LOTUS Trial—Comparing Esomeprazole to Laparoscopic Anti-reflux Surgery for the Management of Chronic Gastroesophageal Reflux Disease: a 3-year Interim Analysis
Lundell and associates from Karolinska University Hospital Huddinge in Stockholm, Sweden, other centers throughout Europe, and AstraZeneca Research and Development in Mölndal, Sweden, note that the increasing efficacy and safety of medical therapy for gastroesophageal reflux disease (GERD) warrants a direct, long-term investigation with laparoscopic antireflux surgery. The authors announced the 3-year interim results of the LOTUS trial, an open, parallel-group, multicenter, randomized, controlled trial conducted in 11 European countries comparing the safety and efficacy of esomeprazole (Nexium, AstraZeneca) with laparoscopic antireflux surgery. The standardized protocol for laparoscopic antireflux surgery consisted of total fundoplication and posterior crural repair, and medical therapy included esomeprazole 20 mg once daily, which could be increased stepwise to 20 mg twice daily if there was inadequate GERD control. Time-to-treatment failure (life-table analysis) was the primary outcome variable in the trial. Inadequate control of GERD signified treatment failure, which was defined as the need for more than 40 mg of esomeprazole in the medical therapy arm or the need for additional operations or medical therapy beyond antacids in the laparoscopic antireflux surgical arm.
A total of 554 patients were randomized to treatment with esomeprazole (n=266) or laparoscopic antireflux surgery (n=288). Both treatment arms had similar baseline demographics and disease characteristics. In the esomeprazole arm, approximately 93% of patients remained in remission compared to approximately 90% in the laparoscopic antireflux surgical arm. No major postoperative complications were reported, and esomeprazole was well tolerated by patients. The researchers concluded that for the first 3 years of this long-term study, both esomeprazole and laparoscopic total fundoplication were similarly effective and well-tolerated treatments for controlling GERD.
This research was supported by a grant from AstraZeneca.
NS The question of whether medical or surgical therapy is superior for patients who have moderate-to-severe reflux disease remains unsettled. Several studies in the literature have suggested that patients treated surgically may have better control of their symptoms than those treated medically. However, advances have been made in both medical and surgical therapies since the time of those studies. Thus, these researchers compared state-of-the-art surgical management with state-of-the-art medical management for patients with chronic reflux disease. It is striking how high the remission rates were in both arms (93% in the medical arm and 90% in the surgical arm), suggesting that both therapies are quite good and that the rates are close to each other. Not much distinguishes the efficacies of these two therapies.
It is important to note that this is a European and not an American trial, so there may be differences among practices, making the generalizability of the trial results less definite. In addition, the primary outcome variable, inadequate control of symptoms, signified different future steps to patients in different arms of the study. Patients in the medical arm who failed therapy would be considering surgery next, whereas patients in the surgical arm who failed would be contemplating pill intake. This may have made medical patients a bit more reluctant to complain about GERD symptoms. However, even with those caveats, the two management strategies appear to be similar, which leaves the issue of patient preference. If two treatment strategies appear to be relatively similar, it is important that clinicians talk with their proton pump inhibitor (PPI)-dependent patients and inform them of their treatment options and elicit their desires. Of course, these are only 3-year interim results and the final results will need to be analyzed. Previous datasets suggest that there may be an erosion of benefit in the surgical antireflux arm and that more of these patients will end up back on medical therapy. We will see if that trend holds true in this study.
Abstract 694 Which Supraesophageal Reflux Symptoms Reliably Respond to PPI Therapy? Results From a Randomized, Double-Blind, Placebo-controlled Trial (RCT).
Iuga and colleagues at the Mayo Clinic in Rochester, Minn., observed that GERD has been cited as one of the causes of laryngopharyngeal irritation resulting in chronic upper airway symptoms and that conflicting evidence exists regarding the effects of acid suppression on these symptoms. The authors sought to determine whether PPI therapy resolved chronic upper airway symptoms, including dry cough, globus sensation, hoarseness, nocturnal cough, sore throat, or throat-clearing and whether a positive response to PPI therapy could be predicted from the features of a patient's history. The authors conducted a randomized, controlled trial of patients from the community, as well as primary and tertiary care clinics, who had at least 1 of 6 symptoms that were chronic (>1 mo) and recurrent (>2 episodes per wk). Individuals who used PPI therapy within 4 weeks of starting the study were excluded. Patients were randomized in a 2.33:1 ratio to either twice-daily esomeprazole 40 mg or placebo for 6 months, and completed monthly phone surveys. Complete response was the primary outcome of the trial and was defined as the resolution of a symptom present at baseline by the third month of therapy and the continuing absence of that symptom for the remainder of the trial.
During the trial, 429 patients were randomized to PPI therapy (n=302, 46% men, mean age 56) or placebo (n=127, 46% men, mean age 54). The overall frequency of nonexclusive symptoms consisted of 214 dry coughs, 148 globus sensations, 191 cases of hoarseness, 109 nocturnal coughs, 78 sore throats, 364 throat clearings, and 160 patients with weekly GERD symptoms. A difference was found in the proportion of throat-clearing cases that were resolved with PPI therapy (7%) versus placebo (<1%; odds ratio [OR] 9; 95% confidence interval [CI], 1.1, 66; P=.04); however, the majority of patients who reported throat-clearing did not achieve complete or partial resolution. As the authors expected, patients with GERD symptoms improved with PPI therapy (OR 15; 95% CI, 3.3, 67; P=.003). No significant differences between PPI and placebo were identified in the complete resolution of any other symptom. The presence of GERD increased the chances of complete resolution of throat-clearing (OR 4; 95% CI, 1.3, 14) and nocturnal cough (OR 4; 95% CI, 1.4, 12), regardless of treatment with PPI or placebo, suggesting that it may be rare to have silent refluxers. The investigators concluded that a small subset of individuals with throat-clearing can improve with PPI therapy and that few people experience complete resolution of dry cough, globus sensation, hoarseness, or sore throat with twice-daily PPI therapy, despite the potential benefit that those with throat clearing and nocturnal cough may experience, especially if accompanied by GERD symptoms.
NS Clinicians are aware of the long list of extraesophageal symptoms associated with reflux, such as chronic cough, sore throat, globus sensation, and sinusitis. The disconnect between what clinicians think should happen in theory and what happens in actuality occurs when many, if not most, of these patients with putative extraesophageal or supraesophageal reflux symptoms do not appear to respond to medical therapy. From a clinical standpoint, gastroenterologists are left in an odd position because they receive patient referrals from otolaryngologists or primary care physicians due to symptom complexes that appear to be consistent with extraesophageal reflux but are not responding to therapy. Gastroenterologists are often the first to suggest that the diagnosis is incorrect and that treatment must be changed.
The authors of this study randomized patients with potential extraesophageal reflux symptoms to either PPI therapy or placebo. What is really remarkable is the lack of efficacy of PPIs in these patients: for the vast majority of the symptoms assessed, patients did no better with medical therapy than with placebo. In fairness, the bar was set quite high, as the primary outcome was complete response to therapy. Certainly, if a patient has multifactorial symptomology and, for instance, postnasal drip is causing throat clearing as well as GERD, one might expect that partial response would be the rule and not the exception.
Nevertheless, this study shows that the likelihood of response to PPI therapy is small among patients with extraesophageal reflux and, without classic GERD symptoms (ie, heartburn), the likelihood of responding to PPI therapy becomes even smaller. Pragmatically, this suggests that if a patient's only potential manifestation of reflux is an extraesophageal symptom of reflux, the likelihood of response to therapy is exceedingly small and it may be worthwhile to look for other etiologies and think about other potential treatments for that symptomology.
Abstract 375 Methylation Biomarkers in Prediction of Response to Photodynamic Therapy of Barrett Esophagus
Wang and associates at the Mayo Clinic in Rochester, Minn., and Johns Hopkins University in Baltimore, Md., note that although photodynamic therapy has been approved in the treatment of Barrett esophagus with highgrade dysplasia (HGD), it is associated with cutaneous photosensitivity and strictures and 23% of patients do not achieve treatment response. The authors had previously found that loss of heterozygosity of p16 was important in determining the lack of ablation response. To confirm this finding, they assessed whether the gene promoter hypermethylation, another mechanism of gene inactivation, was also correlated. A retrospective cohort study was conducted among 30 patients with Barrett esophagus and HGD who underwent photodynamic therapy: 15 patients had responded to the therapy, defined as follow-up biopsy showing the complete elimination of dysplasia, and 15 patients had not experienced treatment response, defined as continued HGD posttreatment. Tissue block slides obtained prior to therapy were blinded and evaluated via extracted genomic DNA for hypermethylation of the promoter regions of RUNX, HPP1, and p16. DNA was treated with bisulfite to convert unmethylated cytosines into uracils before methylation-specific polymerase chain reaction (MSP), and the status of DNA methylation was found by real-time quantitative MSP with the ABI 7700 Sequence Detection System (Taqman). To show the percentage of DNA methylated for each gene, a normalized methylation value was created.
Univariate log transformed logistic analysis (adjusted for non-Gaussian distribution) was used to correlate the normalized mean values of the methylated genes to the photodynamic therapy response of each patient. The level of promoter methylation significantly correlated with methylation of the p16 promoter on univariate analysis (P=.045), even though the correlation was not significant upon performing multivariate analysis (P=.08). The authors concluded that these preliminary data provide independent corroboration that intact p16 is essential for photodynamic therapy in Barrett esophagus and that methylation of the promoter of p16 with subsequent inactivation appears to be detrimental to the treatment response, as the authors showed in an earlier study with p16 LOH.
NS There is currently much excitement regarding the utilization of ablation therapy for Barrett esophagus. However, many clinicians worry that if we are not successful in completely ablating the disease, we may cause a detrimental effect to the residual tissue and somehow activate or push it toward cancer. We also worry that tissue that has had enough genetic derangement may lose the ability to respond to ablative therapy, leaving us with essentially “unablatable” disease. One way to assess this possibility may be to examine biomarkers in the residual tissue and study both patients who respond to therapy and patients who do not. Clinicians have wondered for some time whether these biomarkers could help them predict which patients may respond to ablation therapy. The authors of this study examined hypermethylation, a marker for gene inactivation, on the promoter regions of multiple genes that are thought to be important in the regulation of cell cycle and avoidance of carcinogenesis. The results of the study revealed the interesting trend that the likelihood of the methylation of the p16 promoter region corresponded to the response to ablation therapy; in other words, if a patient's gene was inactivated by methylation, the patient was less likely to respond. These results suggest that clinicians may potentially be able to predict which patients may be good candidates for ablation therapy and which may require surgery or other forms of therapy.
In addition, the results suggest the possibility that patients who respond to ablation are those with relatively benign cellular mechanics to begin with. One might wonder whether such patients are those who are less likely to develop cancer anyway, whereas the patients who fail to respond are actually the ones with worse initial prognoses. In other words, ablative therapy may perform best in patients who least require it. These researchers and others doing similar work may shed light on this possibility in the future and aid clinicians in risk-stratifying patients and better selecting therapy for Barrett esophagus and HGD.
Abstract 376 No Occult Cancer at Esophagectomy in Patients With Barrett Esophagus With High-grade Dysplasia Who Have Undergone Surveillance With the Seattle Biopsy Protocol
Wang and coworkers at the Johns Hopkins Hospital in Baltimore, Md., note that although studies in the surgical literature have reported a prevalence as high as 30–43% for occult cancer in patients undergoing prophylactic esophagectomy for Barrett esophagus with HGD, the type of presurgical endoscopic surveillance biopsy protocol used was unclear. The authors sought to determine whether the prevalence of occult cancer at the time of esophagectomy was significantly reduced by the use of the Seattle endoscopic biopsy protocol in patients with HGD. The authors reviewed the medical records of patients who had undergone prophylactic esophagectomy for Barrett esophagus with HGD at Johns Hopkins Hospital from 1994 to 2006 to determine whether the Seattle endoscopic biopsy protocol was used during presurgical endoscopy. Pathology reports were also examined to determine whether occult adenocarcinoma was identified during surgery.
During the study period, 39 patients underwent prophylactic esophagectomy for Barrett esophagus with HGD. Among patients who had undergone endoscopic surveillance with the Seattle biopsy protocol, no invasive adenocarcinomas (0/15) were found within resected surgical specimens, whereas among patients who had not undergone the Seattle biopsy protocol, 33% (8/24) did have invasive adenocarcinoma in resected surgical specimens. The investigators concluded that the prevalence of occult adenocarcinoma in patients with Barrett esophagus with HGD is very low with the use of the Seattle biopsy protocol during endoscopic surveillance. Therefore, they recommended the reexamination of the standard care of esophagectomy for Barrett esophagus with HGD.
NS The question of whether or not occult cancer can adequately be ruled out with endoscopic surveillance in the setting of HGD is an important and controversial one. Multiple papers in the surgical literature suggest that undetected cancer is present in many patients in whom HGD is found. This has understandably made some clinicians nervous about following these patients endoscopically as opposed to treating them invasively. This paper attempts to determine whether cancer can be identified if patients with HGD are surveyed aggressively enough, as well as how effective the Seattle biopsy protocol detects occult cancers in patients who subsequently go on to esophagectomy. In this study, 15 of the 39 patients underwent the Seattle biopsy protocol, which is a very intensive, rigorous course of 4-quadrant biopsies every 1 cm with a jumbo forceps. The remainder of patients underwent a variety of other biopsy protocols. Interestingly, among the patients who underwent the rigorous Seattle biopsy protocol, clinicians determined during esophagectomy that no case of cancer had been missed. On the other hand, among patients who received a cursory protocol instead of the Seattle protocol, 1 of 3 patients did have an invasive adenocarcinoma that was missed by the biopsy protocol. These results suggest that with an aggressive biopsy and highly standardized protocol, patients can be followed and clinicians can feel fairly confident that they are not likely to miss a cancer that is already present. An important caveat here is that the overall number of subjects in this study is small, so this conclusion must be viewed with caution.
This topic is important because it is hotly debated whether or not intensive endoscopic surveillance is an appropriate strategy for patients with HGD. The present study suggests that if clinicians are going to follow patients with HGD by intensive endoscopic surveillance, a rigorous biopsy protocol should be used. The aggressiveness of the biopsy protocol with respect to how extensively the patients are surveyed is an important determinant of its success.
Abstract 33 A Prospective Clinical Trial of Allergy Testing and Food Elimination Diet in Adults with Eosinophilic Esophagitis (EE)
Gonsalves and associates at Northwestern University in Chicago, Ill., and Children's Memorial Hospital in Chicago, Ill., observe that dietary elimination has not been formally tested in adults with eosinophilic esophagitis (EE), even though the six-food elimination diet (SFED) and elemental diet have shown efficacy in children with EE. In this ongoing trial, the authors assess the efficacy of food elimination diet on histologic improvement in adults with EE, as well as on symptomatic improvement based on dysphagia and SF36 scores, examining adults newly diagnosed with EE (defined as at least 20 eosinophils/high power field [eos/hpf]), or adults previously diagnosed who had relapsed. Patients undergo Bravo or 24-hour pH testing and those tested positive receive twice-daily PPI therapy prior to esophagogastroduodenoscopy (EGD). After conducting skin prick tests (SPT) for food and aeroallergens, patients undergo a 6-week trial of SFED (which eliminates milk, soy, egg, wheat, nuts, and seafood). Also eliminated from patients' diets are any other foods testing positive on SPT. Patients are not allowed concomitant treatment for aeroallergens. After 6 weeks, patients undergo repeat EGD with proximal and distal esophagus biopsies. EE resolution is defined as the peak count of no more than 5 eos/hpf, and biopsies are examined for peak eos/hpf and stained for ki-67 and p63, markers of epithelial hyperplasia, pretreatment and posttreatment. SF36 and dysphagia symptom scores will be noted pretrial and posttrial.
As of the paper presentation, 9 patients (4 men and 5 women, mean age 36, range 25–54) are enrolled. Dysphagia is the most common symptom, with a mean duration of 7.7 years. Food and aeroallergens have been found in all patients, with common food allergens to nuts, soy, wheat, eggs, and seafood and common aeroallergens to tree, grass, ragweed, and dust mite. Only 3 patients have completed the trial and undergone repeat EGD as of paper presentation. In proximal biopsies, the average eos/hpf pretreatment and posttreatment were 52 (±26) and 20 (±16), as opposed to 52 (±28) and 31 (±29) in distal biopsies. One patient achieved complete EE resolution, 1 patient experienced significant eos reduction but no resolution, and 1 patient reported no change. Staining for ki-67 and p63 revealed marked reduction in epithelial hyperplasia in responding patients. Posttreatment, symptom scores decreased 30%, whereas SF36 scores were lower than the general population but did not reveal significant change. The authors concluded that allergy testing and elimination diet are feasible and successful in reducing histologic eosinophilia, epithelial hyperplasia, and symptoms in some adults with EE, confirming a role for food allergens in this cohort. Because aeroallergens were not treated, they may be responsible for triggering patients who did not experience response, and predictors of treatment failure in this group need to be examined in future studies.
NS Though this study is ongoing and very few patients have completed it thus far, it should be highlighted because it addresses an important topic. The most appropriate and best way to treat EE has not yet been found. Additionally, very little is known regarding the pathogenesis of this condition. In children, it is thought that EE may be caused by an allergic reaction that may be at least partially mediated by foodstuffs. It is not unusual for children who have EE to undergo treatments that include elimination diet and allergy testing. The connection between allergy and EE in adults has been more elusive, and the data are somewhat divided as to whether or not EE truly represents an allergic reaction. It certainly has some components of allergy in its biology; however, it is not clear in adults.
These investigators are testing only patients who are newly diagnosed or have active relapsed EE, which is an important distinction because it excludes individuals who have quiescent disease and may not have the same biology.
This study has a fairly rigorous primary outcome variable. The researchers are essentially looking for near complete resolution of eosinophils in the epithelium with a count of less than 5 eos/hpf, which represents a dramatic reduction, given that all the patients started with more than 20 eos/hpf. In addition, these are very preliminary data. Nevertheless, patients who underwent this elimination diet (removing nuts, soy, wheat, eggs, and seafood from their intake) revealed a marked decrease in the number of eosinophils per high power field from preand posttherapy. This was quantified by both proximal and distal biopsies, each of which showed an approximate 50% decrease in the number of eosinophils in the biopsy. The total numbers of subjects in the study are obviously very small. It is unclear what the results will be as more patients complete the trial, but it is important to ask treatment questions in rigorous ways, such as in this study. Studies such as this one will lead to improved treatment of this disease because currently we do not know what constitutes optimal therapy and for how long fairly caustic treatments such as chronic steroid therapy may be potentially involved. Certainly, this kind of study has the potential to shed light on the pathogenesis of the condition in adults, as well as the most appropriate therapy that could potentially be pursued long term.
Presentations in IBD
Abstract 387 Chromoendoscopy for Colonic Dysplasia Surveillance in Inflammatory Bowel Disease
Kandiel and colleagues from the Cleveland Clinic, Cleveland, Ohio, noted that studies from Europe and Japan have shown a 3–4-fold increase in detection of precancerous lesions with the utilization of chromoendoscopy compared to conventional colonoscopy in ulcerative colitis (UC). They sought to reproduce these findings in a population of US patients with UC, Crohn's colitis (CD), or indeterminate colitis (IC). Patients were eligible for enrollment if they had extensive UC or IC (extent proximal to the splenic flexure) or CD involving at least 30% of colon. Two colonoscopies were performed in each patient, the first using conventional methods and the second after spraying 0.1% indigo carmine dye. The number of dysplastic lesions identified with each technique was compared. Surveillance biopsies were also obtained every 10 cm after chromoendoscopy to detect dysplasia in normal-appearing mucosa.
The authors presented preliminary data in 33 patients from a study designed to evaluate a total of 100 patients. All identified dysplastic lesions were adenoma-like masses (ALMs) with low-grade dysplasia (LGD) with the exception of one irregular looking dysplasia-associated lesion or mass (DALM) that was identified with conventional colonoscopy in a 25-year-old patient with severe UC. Examination with chromoendoscopy yielded identification of 2 additional ALMs in UC patients that were not seen with conventional colonoscopy. No flat dysplastic lesions were seen with chromoendoscopy. Researchers performed 1,293 random surveillance biopsies of normal-appearing flat mucosa and identified only 1 focus of LGD in a patient with CD. One patient with CD, who had 2 ALMs, was the only patient with more than 1 dysplastic lesion. The authors concluded that chromoendoscopy may improve the identification of ALMs in UC patients. Further, in this population with a mean age of 49–50 years, a higher frequency of ALMs was seen in CD (33%) than in UC (10%). Finally, the authors concluded that the utility of random biopsies of normal-appearing flat mucosa may not be diagnostically useful and should be re-evaluated.
Abstract 388 Methylene Blue Dye-spray Targeted Biopsies are Superior to Standard Colonoscopic Surveillance Biopsies for Detecting Dysplasia in Patients With Ulcerative and Crohn's Colitis: A Prospective Endoscopic Trial
Marion and associates from the Divisions of Gastroenterology and Pathology at the Mt. Sinai School of Medicine in New York, NY, noted that patients with chronic UC or Crohn's colitis (disease length >8 years) are at an increased risk of developing colorectal cancer. Current screening methods rely on random biopsy sampling at colonoscopy but may fail to detect dysplasia in many patients, whereas dye spraying can aid in the detection of subtle mucosal abnormalities. The researchers sought to prospectively compare dye-spray technique using methylene blue to standard colonoscopic surveillance in detecting dysplasia.
Patients (N=102; 64 men; 79 with UC; 23 with Crohn's colitis) were enrolled. Following a standard bowel preparation each patient was examined in one of three methods: using standard office endoscopic equipment with a method of standard surveillance colonoscopy and 4 random biopsies every 10 cm (for a total of at least 32 samples); via targeted biopsy protocol; utilizing methylene blue (0.01%) dye spray, segmentally applied throughout the colon, with the provision that any pit-pattern abnormality or lesion rendered visible by the dye spray would be targeted and biopsied. Each patient had a single examination that included two passes of the colonoscope, and specimens were reviewed in a blinded fashion by a pathologist. The three methods were then compared with each patient acting as his or her own control.
The authors found that the targeted-biopsy-with-dye-spray method revealed significantly more dysplasia (16 patients with LGD and 1 patient with HGD) than standard colonoscopy with random biopsy (3 patients with LGD; P=.001) or the targeted protocol with no dye spray (8 patients with LGD and 1 patient with HGD; P=.057). Targeted biopsies with and without dye spray or “max” targeted biopsies detected dysplasia in 20 patients compared with 3 using random biopsy (P=.0002; two-tailed exact McNemar Test). The authors concluded that methylene blue dye-spray-targeted biopsies performed during surveillance colonoscopy provide superior detection of dysplasia when compared to current standard techniques, and recommend that this procedure be adopted as the standard of care in preventing colorectal cancer in patients with colitis.
FV The detection of precancerous dysplasia in IBD is an ongoing challenge for clinicians because lesions can present as flat, subtle, and difficult to detect on conventional colonoscopy. Further, few physicians actually take the 32 random colonic biopsies that are recommended to detect dysplasia. To improve visibility, and thus detection, of these subtle dysplastic lesions, chromoendoscopy has been advocated as an enhancing visualization and examination technique. Chromoendoscopy is a technique where different dyes, typically methylene blue or indigo carmine, are applied to the gastrointestinal mucosa during endoscopy in order to better characterize and highlight specific changes. This staining method allows the visualization of mucosal features that might not otherwise be evident, and could potentially improve the sensitivity of endoscopic examination.
Of the two chromoendoscopy studies presented at DDW, the smaller was performed by Kandiel and colleagues from the Cleveland Clinic in a low-risk population, utilizing indigo carmine stain. The larger study by Marion and colleagues from Mt. Sinai Hospital in New York was performed in a high-risk population, nearly 40% of whom had a previous dysplasia diagnosis, and utilized methylene blue stain. Kandiel found that chromoendoscopy resulted in a 1.7-fold increase in the detection of dysplasia over random biopsy, although this increase was not statistically significant. These results were comparable to the five previously published studies on this topic that found nonstatistically significant 1.6–3.5-fold increases in dysplasia detection with chromoendoscopy. The results from Mt. Sinai showed a statistically significantly greater detection of dysplasia with chromoendoscopy compared to random biopsy and a nearly statistically significant detection of dysplasia versus a targeted-biopsy protocol with no dye spray. These two prospective studies add to a small but growing body of literature suggesting that chromoendoscopy increases dysplasia detection when compared to conventional surveillance strategies.
Based on these and past results, some investigators advocate cessation of random, nontargeted biopsy and recommend implementation of dye-targeted biopsy as standard-of-care practice. Controversy centers around the fact that although chromoendoscopy does yield more detection of dysplasia, random biopsy still has the potential to detect additional lesions that are not endoscopically visualized, even with dye enhancement.
Many questions remain regarding the optimal surveillance protocol using chromoendoscopy. At this time, surveillance colonoscopy using chromoendoscopy is not standard of care in patients with inflammatory bowel disease (IBD).
Abstract 982 Infliximab Administered as 3-Dose Induction Followed by Scheduled Maintenance Therapy in IBD: Comparable Clinical Outcomes With or Without Concomitant Immunomodulators
Lichtenstein and associates from the University of Pennsylvania, Philadelphia, Penn., the Mayo Clinic, Rochester, Minn., and Centocor Pharmaceuticals analyzed data from four pivotal trials in CD and UC (ACCENT I & II, ACT I & II) to assess whether concomitant immunomodulator therapy with infliximab (Remicade, Centocor) is associated with better clinical outcomes when 3-dose induction followed by scheduled maintenance therapy is used in IBD. They found that response and remission rates did not differ significantly when infliximab was used with and without concomitant immunomodulators. Although there was a trend in favor of concomitant immunomodulator therapy in some trials, other trials showed a trend against their use. Three of the four trials showed a trend toward fewer hospitalizations when concomitant immunomodulators were used, and this trend was most pronounced in ACCENT I (luminal CD). The authors concluded that the use of concomitant immunomodulators in patients receiving scheduled maintenance infliximab therapy for IBD does not improve clinical outcomes over 54 weeks; the effect of the use of concomitant immunomodulators beyond 1 year of therapy would require further study. Hospitalization rates may be lower with the use of concomitant immunomodulators, especially in CD but, overall, the use of concomitant immunomodulators in patients receiving scheduled maintenance infliximab therapy for IBD should be considered optional, not mandatory (Table 1).
Table 1.
Meta-analysis Across Four Randomized Trials of Infliximab Efficacy With and Without Concomitant Immunosuppression
| Trial | Endpoint | With concomitant immunomodulators (%) | Without concomitant immunomodulators (%) | Odds ratio of outcome with concomitant immunomodulators (95% CI) | Statistical significance |
|---|---|---|---|---|---|
| ACT I (N=243) | Response Wk 30 | 52 | 51 | 1.05 (0.63–1.73) | NS |
| Response Wk 54 | 45 | 45 | 1.00 (0.60–1.65) | NS | |
| Remission Wk 30 | 34 | 36 | 0.91 (0.54–1.55) | NS | |
| Remission Wk 54 | 34 | 36 | 0.92 (0.54–1.55) | NS | |
| Hospitalization Wk 54 | 7.2 | 5.1 | NS | ||
| ACT II (N=241) | Response Wk 30 | 55 | 53 | 1.10 (0.66–1.84) | NS |
| Remission Wk 30 | 36 | 27 | 1.57 (0.90–2.72) | NS | |
| Hospitalizations Wk 30 | 3.9 | 5.8 | NS | ||
| ACCENT I (N=223) | Response Wk 30 | 63 | 53 | 1.53 (0.81–2.87) | NS |
| Response Wk 54 | 50 | 41 | 1.46 (0.79–2.71) | NS | |
| Remission Wk 30 | 52 | 39 | 1.71 (0.93–3.17) | NS | |
| Remission Wk 54 | 37 | 32 | 1.24 (0.66–2.35) | NS | |
| Hospitalizations Wk 54 | 1.9 | 5.3 | NS | ||
| ACCENT II (N=96) | Maintenance of fistula closure Wk 30 | 60 | 65 | 0.80 (0.33–1.95) | NS |
| Maintenance of fistula closure Wk 54 | 43 | 48 | 0.83 (0.34–2.02) | NS | |
| Hospitalization Wk 54 | 3.3 | 4.6 | NS |
FV For many years it has been standard practice to prescribe immunomodulators (azathioprine [AZA], 6-mercaptopurine [6-MP], or methotrexate) to IBD patients receiving infliximab therapy. The rationale was that concomitant immunomodulator use reduced immunogenicity, formation of antibodies against infliximab, infusion reactions, and loss of response. However, concerns of increased infection and malignancy risk, particularly in light of rare but fatal cases of hepatosplenic T-cell lymphoma seen in 11 patients receiving combination infliximab and AZA or 6-MP, have caused physicians to question the risk/benefit profile of concomitant immunomodulator use in patients receiving maintenance infliximab therapy.
To address this question, Lichtenstein and colleagues performed a post-hoc analysis of four pivotal randomized, controlled trials of infliximab in CD and UC, the ACT I, ACT II, ACCENT I, and ACCENT II trials, assessing outcomes in patients treated with infliximab monotherapy versus those treated with concomitant immunomodulators. Despite conventional wisdom regarding the efficacy of immunomodulator cotherapy, in all four trials, similar response and remission rates were observed in patients receiving monotherapy and those receiving immunomodulators. Three of the four trials did show a trend toward fewer hospitalizations when concomitant immunomodulators were used, but overall outcomes were not significantly affected.
The Infliximab Maintenance Immunosuppressive Discontinuation trial, also presented at DDW (Van Assche et al., Abstract 734), showed similar results when patients on immunomodulator therapy were either continued for 6 months or more after the start of infliximab or discontinued immediately. No clinical or endoscopic benefit over infliximab monotherapy was found at 2-year follow-up, although those patients who continued to receive immunomodulator therapy had higher infliximab trough levels. Similar data were presented for other biologic agents, including certolizumab, natalizumab, and adalimumab, showing no increased benefit from concomitant immunomodulator therapy. Though the data presented by Lichtenstein and colleagues represent a post-hoc analysis, they further confirm the concept that physicians should not feel obligated to reflexively prescribe an immunomodulator in all patients who are receiving infliximab. This is a particularly important distinction in pediatric patients, who make up the majority of those who have developed associated hepatosplenic T-cell lymphoma.
Whether or not there are subgroups of patients who may benefit from concomitant immunomodulator therapy remains to be determined. The ongoing SONIC trial is looking at patients who are naive to both immunomodulators and infliximab and comparing monotherapy versus cotherapy. Results will compare outcomes at 6-month follow-up and should provide definitive information with regard to efficacy of different regimens.
Abstract 272 A Multicenter, Randomized, Phase 2a Study of Human Monoclonal Antibody to IL-12/23p40 (CNTO 1275) in Patients With Moderately to Severely Active Crohn's Disease
Sandborn and colleagues from the Mayo Clinic, in Rochester, Minn., as well as several additional centers in the United States, Canada, and Belgium note that the type 1 cytokines interleukin (IL)-12 and IL-23 are implicated in the pathophysiology of CD. They conducted a 54-week study to assess the safety and efficacy of either a single intravenous (IV) infusion or 4 subcutaneous (SC) injections of CNTO 1275 in patients with moderate-to-severe CD, including infliximab nonresponders.
A population of patients (N=104) with moderate to severe CD (Crohn's Disease Activity Index [CDAI] 220–450) refractory to treatment with 5-aminosalicylic acid (5-ASA), antibiotics, corticosteroids, infliximab, and/or immunomodulators were randomized to 1 of 4 groups (approximately 25 patients/group) to receive the following regimens: SC placebo at Weeks 0, 1, 2, 3 and SC CNTO 1275 (90 mg) at Weeks 8, 9, 10, 11 (Group 1); SC CNTO 1275 (90 mg) at Weeks 0, 1, 2, 3, and placebo at Weeks 8, 9, 10, 11 (Group 2); IV placebo at Week 0 and IV CNTO 1275 (4.5 mg/kg) at Week 8 (Group 3); or IV CNTO 1275 (4.5 mg/kg) at Week 0 and IV placebo at Week 8 (Group 4). An additional population of patients with previous failure of response or loss of response to infliximab were randomized to open-label therapy in either SC CNTO 1275 (90 mg) at Weeks 0, 1, 2, 3 (Group 5, n=14) or IV CNTO 1275 (4.5 mg/kg) at Week 0 (Group 6, n=13). Clinical response was defined as a reduction from baseline in the CDAI of 25% or more and 70 points or more. The primary endpoint was clinical response at Week 8 for the initial population (IV and SC combined), and data through Week 16 were presented.
At Week 8, 49.0% of patients in the randomized population receiving CNTO 1275 achieved clinical response versus 39.6% receiving placebo (P=.34). At Week 4 and at Week 6, 52.9% of patients receiving CNTO 1275 were in clinical response versus 30.2% receiving placebo (P=.02). At Week 8, 49.0% of patients receiving CNTO 1275 were in response, using the criterion of a 100 point or greater CDAI reduction versus 30.2% receiving placebo (P=.05). Among patients with prior infliximab experience in the randomized population (n=49), 59.1% receiving CNTO 1275 were in clinical response at Week 8 versus 25.9% receiving placebo (P=.02). In the open-label population, clinical response rates at Week 8 were 42.9% for SC-administered CNTO 1275 and 53.8% for IV-administered CNTO 1275. Similar proportions of adverse events were reported for CNTO 1275 and placebo patients. No serious infections occurred in the randomized population but 2 occurred among the open-label patients: disseminated histoplasmosis (in a patient with previous steroid, AZA, and approximate 3-year history of prior infliximab use) and food poisoning. The authors concluded that use of CNTO 1275 was well tolerated and provided benefit in patients with moderate-to-severe CD.
FV Throughout the near decade-old era of biologic therapy for IBD, the inflammatory cytokine TNF-alfa has been the sole target of approved agents. However, novel compounds targeting other mediators are likely to make up the next generation of biologic agents used to treat CD and UC. At DDW, several phase II clinical trials of these novel compounds were presented. Notable among them was Sandborn and colleagues' trial of CNTO 1275, a fully human monoclonal antibody targeted against the common P40 subunit of IL-12 and IL-23. The study concluded that there was short-term benefit seen with this novel monoclonal antibody in patients with moderate-to-severe CD, that it was well tolerated, and that this effect was most prominent in patients with prior infliximab experience, suggesting that patients who have taken infliximab can benefit from this alternative method of treatment. This study highlights the fact that promising biologics with alternative mechanisms of action are under active investigation in both the preclinical and clinical phases.
Abstract 639a A Randomized, Placebo-controlled Trial of the PPARy Ligand Rosiglitazone for Active Ulcerative Colitis
Lewis and associates from the University of Pennsylvania, Philadelphia, Penn., and other US centers conducted a multicenter, randomized, double-blind, placebo-controlled trial of the thiazolidinedione (TZD) rosiglitazone (Avandia, GlaxoSmithKline) to determine the efficacy of its anti-inflammatory properties in patients with mild to moderately active UC. Patients (N=105) refractory or intolerant to 5-ASA therapy were randomized to receive either oral rosiglitazone 4 mg bid or placebo bid for 12 weeks. Disease activity was measured with the Disease Activity Index (DAI), with mild-to-moderate activity defined as a score of 4–10, inclusively. Primary endpoint was clinical response (reduction in the DAI by 2 points or more) at Week 12. Secondary outcomes included clinical response defined as reduction in the DAI by 3 points or more, clinical remission (DAI ≤2), endoscopic remission (DAI <2 and mucosal appearance =0), and quality of life (increase in inflammatory bowel disease questionnaire by 16 points or more). Patients who did not complete 12 weeks of follow-up were defined as treatment failures for all outcomes. Analyses were performed on an intention-to-treat basis.
At the end of treatment, 23 patients (44%) treated with rosiglitazone and 12 patients (23%) treated with placebo had achieved clinical response (P=.03). Patients treated with rosiglitazone had higher rates of the secondary outcomes of clinical response (reduction in the DAI by 3 points or more) and clinical remission, but not endoscopic remission. Improvement in endoscopic appearance (P=.01), stool frequency (P=.04), bleeding (P=.21), and Physician's Global Assessment (P=0.03) were more common in the rosiglitazone arm. Quality of life was significantly improved at Week 8 (P=.01) but not at Weeks 4 (P=.48) or 12 (P=.14). The authors concluded that rosiglitazone is efficacious in the treatment of mild-to-moderately active UC refractory to or intolerant of 5-ASA and should be considered as a novel secondline therapy.
FV 5-ASA is the cornerstone of therapy for mild-to-moderate UC. Patients who fail 5-ASA are often prescribed steroids or immunomodulators but resist this next step up in therapy. The thiazolidinedione ligands, including rosiglitazone, are PPAR-γ ligands proposed to have anti-inflammatory effects in patients with UC. Unlike most novel therapies for IBD, which are bioengineered monoclonal antibodies created almost exclusively for the treatment of IBD, rosiglitazone is a commonly used medication for Type 2 diabetes. This randomized controlled trial is notable not only in showing that a medication for diabetes can be effective in treating UC, but that patients with mild or moderate UC refractory or intolerant of 5-ASA may have an alternative to immunomodulators and steroids.
However, concurrent with the presentation of this data, a meta-analysis was published in the New England Journal of Medicine that linked rosiglitazone with a 43% increased risk of myocardial infarction in diabetes patients, which has prompted a review of this drug by the US Food and Drug Administration. In response to this review, interim results for the RECORD (Rosiglitazone Evaluated for Cardiac Outcomes and Regulation of Glycemia in Diabetes) study were recently published. This study is prospectively designed to look at cardiovascular outcomes in over 4,400 patients with type 2 diabetes. Fortunately, this interim analysis did not find that the rosiglitazone group was significantly different from the control group with respect to the endpoint of cardiovascular hospitalization or death.
Similarly, in the study by Lewis and colleagues, where patients were generally younger and did not have the underlying cardiovascular risk factor of diabetes, there were no significant cardiovascular complications or myocardial infarctions. The main adverse effect was lower extremity swelling, which is a known side effect of this class of medications. Despite the potential safety concerns in diabetic patients, the study can be interpreted as showing that rosiglitazone may be an effective, and hopefully safe, agent for selected patients with UC. Should short-and long-term safety concerns be adequately addressed, rosiglitazone may in the future prove useful in patients refractory or intolerant of 5-ASA and may reduce the need for steroids in this population.
Presentations in IBS
Abstract 535 Fecal Serine-Protease Activity: a Possible Pathophysiological Factor and Biomarker for Diarrhea-Predominant Irritable Bowel Syndrome
Gecse and associates from the University of Szeged in Hungary and the French National Institute for Agricultural Research note the recently observed high colonic luminal serine-protease activity that characterizes diarrhea-predominant irritable bowel syndrome (IBS-D) and posit that it may be responsible for commonly observed features in IBS, including increased intestinal permeability and visceral sensitivity. They investigated the origin of elevated serine-protease activity to evaluate whether it is a relevant pathophysiologic marker in IBS-D.
Fecal samples were collected from healthy subjects, patients with IBS (all subgroups, based on Rome II criteria), UC patients, and patients with acute infectious diarrhea. Fecal serine-protease, trypsin, mast-cell tryptase, and myeloperoxidase activity were measured photometrically. Fecal pancreatic elastase-1, human secretory leukocyte protease inhibitor, neutrophil elastase, and calprotectin concentrations were assayed by enzyme-linked immunosorbent assay. The authors found that serine-protease activity was significantly elevated in IBS-D and UC patients compared to healthy controls (63.9±8.8 and 62.2±14.0 vs 24.5±6.3 U/mg protein; P<.01 and P<.05, respectively). These increases were not present in subgroups with infectious diarrhea or other IBS subgroups. No significant difference was observed in fecal trypsin and mast-cell tryptase activity in UC, infectious diarrhea, or in any subgroups of IBS patients when compared to healthy controls. Likewise, fecal pancreatic elastase-1 concentration showed no significant difference among subgroups of IBS and UC patients when compared with healthy controls. However, levels were decreased in the feces of infectious diarrhea patients (P<.05). Secretory leukocyte protease inhibitor activity showed no significant difference among any of the studied groups. Markers of intestinal inflammation, including neutrophil-elastase, calprotectin, and myeloperoxidase activity were elevated only in stool samples of UC and infectious diarrhea patients.
The authors concluded that elevated fecal serine-protease activity in IBS-D patients is not due to increased intestinal transit, linked to increased serine-protease activity of epithelial origin, coupled with decreased luminal antiprotease activity, nor associated with mucosal inflammation. They also posited that elevated serine-protease activity is a relevant pathophysiologic marker in IBS-D, in the absence of inflammatory markers.
WC The search for a definitive IBS-associated biomarker has been ongoing since the syndrome was first recognized. Recent observations have noted that proteinase-activated receptors (PARs) may play an important role in the development of visceral hypersensitivity in patients with functional disorders like IBS or functional dyspepsia and that these PARs are cleaved by proteases. Thus, these authors sought evidence of consistently abnormal levels of proteases in patients with IBS. They found that fecal serine-protease activity is markedly elevated in patients with IBS-D but not other forms of IBS and that it is also elevated in UC. Despite further testing, they were not, however, able to pinpoint the mechanism that causes these raised levels.
There is some speculation that IBS-D may represent a form of IBD and that these serine proteases are emanating from overactive inflammatory cells like neutrophils. Whether or not this holds true, differentiation from similar measures seen in UC patients will need to be sorted out before this marker can be utilized clinically because IBS-D and UC are established and distinct disease states. Proven UC therapies, including 5-ASA and steroids, have been tried and have had no effect in IBS-D. From a treatment standpoint, these findings also suggest that protease inhibitors could be of potential benefit as a therapeutic option in IBS-D patients.
Abstract 986 Screening for Celiac Sprue in Patients With Suspected Irritable Bowel Syndrome: Results From a Prospective US Multi-Center Trial
The utility of screening for celiac disease in patients with suspected IBS remains unclear. Investigators from the University of Michigan Medical Center in Ann Arbor, the National Naval Medical Center in Bethesda, Md., and the Eisenhower Army Medical Center in Augusta, Ga., conducted a study to determine the performance characteristics of serologic tests for celiac disease in patients with suspected IBS and control subjects. Patients meeting Rome II criteria for nonconstipated IBS and healthy individuals scheduled to undergo colonoscopy for colorectal cancer screening were invited to undergo serologic screening for celiac disease. Subjects who agreed to participate underwent testing for immunoglobulin (Ig)G and IgA antigliadin (AGA), anti-endomysial (EMA), and anti-tissue transglutaminase (TTG) antibodies. Total serum IgA levels were also measured.
Thus far, 366 IBS patients and 276 controls have enrolled in this ongoing trial. At least one abnormal antibody test result was seen in 32 (8.7%) IBS patients and 8 (2.9%) controls (P=.005). IgG AGA was most commonly positive (23 IBS patients and 3 controls). Biopsy-proven celiac disease was subsequently confirmed in 4 (1.1%) IBS patients and 2 controls (0.7%, P =NS). Antibody test results for one IBS patient with celiac disease were not available. Table 2 shows performance characteristics for the different antibody tests given.
Table 2.
Performance Characteristics of Serologic Celiac Disease Markers
| IgG AGA | IgA AGA | EMA | IgA TTG | |
|---|---|---|---|---|
| Irritable Bowel Syndrome Patients Only | ||||
| Sensitivity | 100% (3/3) | 0% | 33.3% (1/3) | 33.3% (1/3) |
| Specificity | 94.5% (342/362) | 98.3% (359/365) | 99.7% (363/364) | 99.2% (361/364) |
| PPV | 13% (3/23) | 0% | 50% (1/2) | 25% (1/4) |
| Irritable Bowel Syndrome Patients and Controls | ||||
| Sensitivity | 60% (3/5) | 20% (1/5) | 60% (3/5) | 60% (3/5) |
| Specificity | 95.7% (536/560) | 98.3% (550/559) | 99.8% (559/560) | 99.5% (557/560) |
| PPV | 7.7% (2/26) | 10.0% (1/10) | 75% (3/4) | 50% (3/6) |
- AGA
antigliadin antibodies
- EMA
anti-endomysial antibodies
- Ig
immunoglobulin
- PPV
positive predictive value
- TTG
antitissue transglutaminase.
The authors observe that although celiac antibodies were identified significantly more commonly in IBS patients than controls, the prevalence of biopsyproven celiac disease was similar between groups. Contrary to published literature, EMA and TTG yielded marginal sensitivity and positive predictive value for celiac disease. The addition of IgG AGA identified patients with celiac disease missed by EMA and TTG but commonly yielded false-positive results. These preliminary results suggest that current recommendations to screen for celiac disease with EMA or TTG alone may not be adequate in clinical practice.
This study is supported by a grant from Prometheus Laboratories.
WC Our prospective multicenter study illustrates that although the prevalence of antibodies to celiac disease appears to be elevated in nonconstipated IBS patients, actual, biopsy-proven celiac disease is no more common in patients fulfilling Rome II criteria for IBS versus healthy individuals undergoing colon cancer screening, which is surprising and contrary to earlier research conducted in the United Kingdom.
For IBS patients, the prevalence of biopsy-proven celiac disease was 1.1% versus 0.7% for controls, which was not significantly different. The higher prevalence of celiac antibodies in the IBS versus control group may therefore seem unimportant. However, this mismatch may provide some insight into the pathogenesis of symptoms in some IBS patients. Most of the antibodies that were identified in IBS patients were antigliadin antibodies, confirming European studies that have reported similar findings. This, coupled with the fact that gliadin is the major protein in wheat, leads us to theorize that although these patients do not have true celiac disease, they may have a form of gluten sensitivity.
The other important point from this study concerns current recommendations for serologic screening in celiac disease. Most diagnostic guidelines for celiac disease recommend screening with TTG and/or EMA, and it has been said that the sensitivity and specificity of these tests for celiac disease exceed 90%. However, these numbers have been reported in studies of patients with known disease (ie, biopsy-proven sprue) versus patients with no disease or clearly negative biopsies. This is very different from examining a real-world population of patients with symptoms and a low prevalence of disease and then using these tests to screen for the disorder. This study is one of the first to give a detailed description of antibody profiles in patients with IBS symptoms diagnosed with celiac disease as they present in clinical practice, and it shows that EMA and TTG are not as sensitive in this low prevalence population as has been previously reported. This finding may be related to the fact that nearly all patients diagnosed with celiac disease in our study had partial villous atrophy or only intraepithelial lymphocytes. Importantly, none of the patients in our study diagnosed with biopsy-proven celiac disease had full villous atrophy.
Abstract 532 Effects of a Novel, First-in-Class Guanylate Cyclase-C Activator, Linaclotide Acetate (MD-1100), on Gastrointestinal and Colonic Transit and Bowel Habits in Patients With Constipation-Predominant Irritable Bowel Syndrome (C-IBS)
Andresen and colleagues from the Mayo Clinic in Rochester, Minn., and Microbia, Inc., in Cambridge, Mass., announced data from a randomized, double-blind, placebo-controlled study evaluating the effects of oral linaclotide, a novel, first-in-class agonist of the human guanylate cyclase-C (GC-C) for the treatment of constipation-predominant IBS (IBS-C). In previous human phase I studies, linaclotide was safe and well tolerated, decreased stool consistency, increased stool frequency, and increased ease of stool passage in healthy volunteers.
In the current study, 36 women with IBS-C were randomized to regimens of 100 μg or 1000 㚼g linaclotide daily or placebo. Participants underwent 5-day baseline and 5-day treatment periods, during which gastrointestinal transit (via validated scintigraphy) and bowel function (by daily diaries, including Bristol Stool Form Scale) were assessed. Treatment effects were evaluated using analysis of covariance with pairwise comparisons of each dose versus placebo. Baseline colonic transit was used as covariate.
Linaclotide did not detectably affect gastric emptying and orocecal transit times. However, there was a significant (P=.015) overall treatment effect on ascending colon emptying halftime, with a significant acceleration associated with linaclotide 1,000 μg/daily (P=.004). Treatment effects on overall colonic transit were also observed (P=.020 for the geometric center at 48 hours) with significant acceleration with linaclotide 1,000 daily versus placebo (P=.01). There were significant overall effects on stool frequency, stool consistency, ease of passage, and time to first bowel movement, with a strong dose response for stool consistency (overall P<.001). No safety issues were identified.
The investigators concluded that in women with IBS-C, linaclotide 1,000μg daily significantly accelerated ascending colon emptying and colonic transit at 48 hours and also improved stool consistency, stool frequency, ease of passage, and time to first bowel movement. Further studies of clinical endpoints in randomized controlled trials are warranted.
WC This study provides objective data on the pharmacodynamic effects of linaclotide on gastrointestinal and colonic transit. The investigators found that linaclotide had little effect on gastric emptying or small bowel transit but did accelerate colonic transit, as well as impact stool frequency and stool consistency. As a pharmacodynamic study, it was not powered to examine IBS symptoms like abdominal pain or global symptom relief. These findings do provide justification for proceeding to larger trials in patients with IBS-C and chronic constipation. Results from these large clinical trials will be available next year.
Abstract 639 Lubiprostone Significantly Improves Symptom Relief Rates in Adults With Irritable Bowel Syndrome and Constipation (IBS-C): Data from Two, Twelve-Week, Randomized, Placebo-Controlled, Double-Blind Trials
Lubiprostone (Amitiza, Takeda/Sucampo) is a chloride channel activator currently indicated for the treatment of chronic idiopathic constipation in adults. In a recent dose-ranging, double-blind, placebo-controlled phase II study, lubiprostone improved relief of IBS symptoms and was well tolerated. To confirm these results, two 12-week, phase III, multicenter, double-blind, randomized, placebo-controlled studies of lubiprostone in adults with IBS-C were conducted by Drossman and associates at the University of North Carolina in Chapel Hill, the University of Michigan Health System in Ann Arbor, and Sucampo Pharmaceuticals in Bethesda, Md.
Adults (n=1,167) diagnosed with IBS-C (Rome II criteria) were enrolled and received lubiprostone (8 mg twice daily; n=780) or placebo (n=387) for 12 weeks. Most were female (91.6%) and 18–65 years of age (91.7%). Primary efficacy was determined utilizing a numeric scale response to the unique question: “How would you rate your relief of IBS symptoms (abdominal discomfort/pain, bowel habits, and other IBS symptoms) over the past week compared to how you felt before you entered the study?” The two highest points of a 7-point balanced scale (where the center point constituted no change in symptom status) were used for patients to qualify as a responder. Patients reporting at least moderate relief for 4 of 4 weeks or significant relief for 2 of 4 weeks were considered monthly responders. Overall responders were monthly responders for at least 2 out of 3 months. Nonresponders were defined as patients discontinuing for any reason or reporting an increase in rescue medication use, lack of efficacy, or moderately or significantly worse relief.
Patients receiving lubiprostone achieved greater overall response compared to those receiving placebo (lubiprostone 17.9% vs placebo 10.1%; P=.001). Individually, each study showed lubiprostone's superiority for overall response (P=.009 and P=.031). Secondary endpoints measured in both studies included abdominal discomfort/pain, stool consistency, straining, constipation severity, and quality of life and were statistically significantly improved in patients receiving lubiprostone versus placebo. Lubiprostone was well tolerated, with a similar incidence of serious adverse events (1% in each group) and related adverse events (lubiprostone 22%, placebo 21%). The most common treatment-related adverse events were nausea (8% vs 4%, respectively) and diarrhea (6% vs 4%, respectively).
WC An important distinction of this trial is the very rigorous endpoint utilized to minimize placebo response. In past IBS trials, symptom assessment scales have used graded verbal descriptors (eg, mild, moderate, marked) or binary (yes or no) to assess clinical improvements in global or individual IBS symptoms. A criticism of these types of scales is that they identify only if a patient improves with an intervention. The balanced scale used in this trial situated “no benefit” in the center and allowed patients to identify whether the intervention improved or worsened their IBS symptoms. This is the first time that a large, randomized, placebo-controlled IBS study has used this type of balanced scale.
Past IBS studies have been criticized because of placebo responses ranging from 35% to 50%. The extremely rigorous endpoint in this study successfully minimized the placebo response to 10%. The therapeutic gain for lubiprostone over placebo was approximately 8%, which is similar to other drugs that have been shown to benefit IBS patients, such as alosetron or tegaserod. It is clear from this study that, similar to these other drugs, lubiprostone benefits only a subset of the total IBS population. At this point in time, it is unclear which clinical factors identify patients who are more likely to improve with lubiprostone.
The other important finding from this study is the reduced rate of diarrhea associated with the 8-μg dose of lubiprostone, compared to earlier studies in patients with constipation that used a 24-μg dose. This finding tells us that diarrhea is a dose-dependent side effect of this agent. If lubiprostone is approved for the treatment of IBS-C, the 8-μg dose should be a valuable addition to the gastroenterologist's treatment armamentarium.
Abstract 537 Randomized Controlled Trial Shows Biofeedback to Be Superior to Alternative Treatments for Patients With Fecal Incontinence
Heyman and colleagues from the University of North Carolina at Chapel Hill and the Veterans Administration for Health Promotion and Disease Prevention in Durham, NC, compared electromyographic biofeedback to an alternative treatment for patients with fecal incontinence. Patients (N=148) participated in a 4-week run-in treatment consisting of medication, education, and behavioral strategies to prevent fecal incontinence. Of the total cohort, 35 (21%) reported adequate relief, precluding further participation, and 24 (14%) withdrew during run-in. The remaining 108 patients (83 females) were included in the intent-to-treat analysis. All subjects were randomly assigned to one of two treatment groups: electromyographic biofeedback and Kegel exercise or Kegel exercise alone. Behavioral strategies were discussed with all patients during 6 biweekly 1-hour sessions. Diary data (incontinence episodes and compliance with Kegel exercises and behavioral strategies) were reviewed in each session. The primary dependent measure was a report of adequate relief (yes or no) 3 months post-treatment.
Chi-Square analysis showed superiority of biofeedback plus Kegel exercise to Kegel exercise alone; 77% of biofeedback versus 40.6% of controls were successful (X2=14.2, P<.001). Secondary analyses at 3-month follow-up showed that biofeedback patients had significantly lower scores on the Fecal Incontinence Severity Index compared to controls (P=.001) and a trend favoring biofeedback over controls (P=.07) for fecal incontinence frequency (from diary data).
The investigators concluded that biofeedback for fecal incontinence is superior to education and conservative management and also superior to Kegel exercises alone.
This study was supported by National Institutes of Health grants R01 DK57048, R24 DK67674, and RR00046.
WC Biofeedback therapy can be performed by utilizing an electromyographic probe or a pressure manometer inserted into the anal sphincter to provide visual feedback for patients attempting to strengthen their pelvic floor and anal sphincter muscles. Patients with incontinence and a weak sphincter will often squeeze their gluteal muscles when exercising and require guidance to properly squeeze and strengthen the anal musculature instead.
This is one of the first well-done, relatively large randomized trials to show benefit from biofeedback therapy in patients with fecal incontinence. Data from smaller observational studies have shown that biofeedback may be helpful but the largest randomized study prior to this one, from St. Mark's Hospital in London [Norton et al. Gastroenterology. 2003;125:1320–1329], found no benefit to biofeedback versus Kegel exercises or advice and education. This study by Heyman and colleagues shows that biofeedback can be worthwhile, at least in a subset of patients. Currently, many insurance plans do not reimburse for biofeedback, and therapists often do not offer this service. Hopefully, this evidence will promote more widespread use of this potentially beneficial therapy.