- Email: [email protected]
Fecal lactoferrin distinguishes inflammatory from non-inflammatory causes of symptoms in patients with ileal-pouch anal anastomosis (IPAA)
S76
Abstracts
Purpose: Non-steroidal anti-inflammatory drugs (NSAIDs) have been implicated in small bowel ulceration and stricture formation, which are difficult to detect. We report on 12 patients referred either for obscure gastrointestinal bleeding [10] or abdominal pain [2] in which NSAID associated strictures and ulcers were detected by CE and/or intra-operative endoscopy. Methods: In the study population, the mean age was 60.2 yrs [range 18-77]; 3 were male and 9 were female. 9 had iron deficiency anemia and had received an average of 4 units of blood (range 2–20). 11 of 12 patients underwent CE after a prior workup that included a minimum of upper endoscopy [mean 1.41]; colonoscopy [mean 2.0] and all had enteroclysis or a SBFT. Reported NSAID consumption was from 1 to 30 years. 2 took NSAIDs for more than 10 years prior to CE. Results: Strictures with associated ulceration were found in all patients using CE while 2 had findings on radiography. The number of strictures per patient varied from 1 to 23. 10 of 11 patients had transient capsule retention of up to 2 weeks duration, 3 of 11 patients had capsule retention requiring surgical retrieval. 6 of 11 required surgical treatment consisting of intraoperative enteroscopy [6], small intestinal resection [6] and/or stricturoplasty [1] based on CE findings. Pathological exam of surgical specimens revealed strictures comprised of circumferential webs associated with a few discrete ulcers. Macroscopically, the most common findings were short concentric ring-like stenoses, with clearly demarcated ulceration, usually at the innermost margin of the web. In addition some strictures had no serosal representation, and were only located by CE and confirmed by intraoperative enteroscopy. The most frequent histological finding was fibromuscular hyperplasia that broadened the plica circularis. This was associated with disruption and disappearance of the muscularis propria, particularly towards the top of the plica. This was limited to the submucosa and was remarkable for the lack of inflammatory infiltrate. Conclusions: NSAID injury to the small intestine is more common than previously appreciated. CE provides a sensitive means of detecting NSAID lesions in the small intestine. Complete surgical treatment requires intraoperative enteroscopy to detect all stenoses. The pathology of these lesions may be characteristic for NSAID lesions of the small intestine and is distinct from Crohn’s disease. 221 RANDOMIZED CONTROLLED TRIALS CLAIMING EQUIVALENCY IN DIGESTIVE DISEASES: ARE THEY METHODOLOGICALLY RIGOROUS? Jill M. Tinmouth, M.D., Leah S. Steele, M.D., George A. Tomlinson, Ph.D., Richard Glazier, M.D. M.P.H.*. University of Toronto, Toronto, ON, Canada. Purpose: Traditionally, randomized controlled trials (RCTs) have attempted to demonstrate the superiority of one intervention over another. However, when efficacious treatment already exists, it may be more useful to prove that an intervention is equivalent, or at least not inferior, to the standard of care. The purpose of this paper is to determine whether claims of equivalency in digestive diseases trials are supported by the evidence. Methods: Medline (1989 –2002) was searched for RCTs using the MeSH headings “exp therapeutic equivalency” and “exp digestive diseases” and the text words “equivalence”, “equal”, “equals” or “equivalent” yielding 890 articles. Of these, 73 articles met the inclusion criteria. These articles were then reviewed using previously published criteria for equivalency (Greene et al., Ann Int Med 2000). Results: Amongst the included articles, 33% stated a priori that their aim was to test for equivalence. The magnitude of absolute differences between the interventions studied ranged from 0 to 58% for articles reporting their results as proportions (n⫽54). For articles reporting results as continuous values (n⫽16), the magnitude of relative differences between interventions ranged from 0 to 33%. Ten percent (7 of 73) of the articles described differences of 20% or more between “equal interventions”. Thirty four percent (34%) of articles set a boundary to test equivalence and confirmed it statistically while 52% used non-significant superiority tests (i.e., p-value ⬎ 0.05). Fourteen percent (14%) failed to use any statistical test to
AJG – Vol. 98, No. 9, Suppl., 2003
support their claims of equivalency. Sample size calculations were done in 30% of articles. Only 8 of 73 (11%) articles met all the criteria of stating an aim a priori, setting and testing an equivalency boundary, and calculating the necessary sample size to demonstrate equivalency. Conclusions: Trials in digestive diseases in which interventions are reported to be equivalent tend to be poorly supported by the evidence. In particular, more than half the articles mistakenly concluded that “failing to detect a significant difference” was synonymous with a finding of equivalence. Erroneous claims of equivalency are potentially dangerous and may lead to substandard patient care. 222 EFFECTS OF THE CCK1-RECEPTOR ANTAGONIST DEXLOXIGLUMIDE ON NIPPOSTRONGYLUS BRASILIENSISALTERED JEJUNAL SENSITIVITY IN RATS Lionel Bueno*, Gianfranco Caselli, Cecile Bonnafous, Patrick Griffin, Harvey Schneier, Lucio Rovati, Massimo D’Amato. INRA, Toulouse, France; Forest Research Institute, Jersey City, NJ and Rotta Research Laboratorium, Monza, Italy. Purpose: To evaluate the effects of dexloxiglumide on N. brasiliensisaltered jejunal sensitivity in rats. Methods: Male Wistar rats (8 groups of 8) were infected by subcutaneous administration of N. brasiliensis infective larvae. Rats received oral treatment with dexloxiglumide (30 or 100 mg/kg) or its vehicle (placebo) every day from Day 12 (ie, the peak of inflammatory response) to Day 27 postinfection. On Day 30, intestinal sensitivity was tested before and after administration of dexloxiglumide (5 or 20 mg/kg intraperitoneal[IP]) or its vehicle. A latex balloon inserted into the jejunum was inflated (12.5–100 mm Hg, 25 seconds every 5 minutes) to cause distention. Intraluminal pressure and volume were measured by a barostat connected to the balloon, and systemic blood pressure (BP) was recorded from a sidearm off of the left carotid cannula. A depression in BP correlated with an increase in visceral pain. Results: Dexloxiglumide treatment (Days 12–27) reduced the amplitude of the depression in BP for balloon pressures of 25, 50, and 75 mm Hg in a dose-dependent manner, reaching statistical significance (P⬍0.05) at the 100 mg/kg dose, both as a single data point at 50 mm Hg pressure, and as a global response evaluation (AUC of depressor response over applied intestinal pressure; 1350 ⫾ 128, 821 ⫾ 164, and 622 ⫾ 164, for controls, 30 mg/kg, and 100 mg/kg, respectively). Moreover, in the equation proposed by Ness and Gebhart [Brain Res. 1988;450:153 — (⌬ blood pressure)2 ⫽ (slope ⫻ distention) ⫹ intercept], the slopes of curves obtained in animals treated with 30 or 100 mg/kg dexloxiglumide (7.8 ⫾ 1.44 and 6.55 ⫾ 1.44, respectively) were significantly different from the slope of the control animals (18.58 ⫾ 2.08, P⬍0.05). IP administration of dexloxiglumide to chronically treated animals had no additional effect on the reduction of BP depression in response to jejunal distention beyond what was observed in chronically treated animals. Dexloxiglumide did not affect the change in the volume associated with increasing pressure, ie, jejunal tone. Conclusions: Chronic treatment with dexloxiglumide reduces or suppresses N. brasiliensis-induced intestinal hypersensitivity to distention. Chronic treatment with dexloxiglumide may therefore reduce visceral hypersensitivity observed in patients with IBS. 223 FECAL LACTOFERRIN DISTINGUISHES INFLAMMATORY FROM NON-INFLAMMATORY CAUSES OF SYMPTOMS IN PATIENTS WITH ILEAL-POUCH ANAL ANASTOMOSIS (IPAA) Mansour A. Parsi, M.D., Bret A. Lashner, M.D.*, Bo Shen, M.D., Jean-Paul Achkar, M.D., Feza Remzi, M.D., John Goldblum, M.D., Dahai Lin, B.S., Jason T. Connor, M.S., Dhanasekaran Ramasamy, M.D., Nish Shah, M.D., Marlene Bambrick, Victor W. Fazio, M.D. The Cleveland Clinic Foundation, Cleveland, OH.
AJG – September, Suppl., 2003
Abstracts
Purpose: Symptoms of increased stool frequency, urgency, and abdominal pain in patients with IPAA, can be due to inflammatory conditions such as pouchitis, cuffitis, or Crohn’s disease (CD) or to non-inflammatory conditions such as irritable pouch syndrome (IPS). Differentiating between these entities usually requires pouch endoscopy and biopsy. Non-invasive means of diagnosis are preferable. Methods: 54 consecutive subjects who had undergone proctocolectomy with IPAA for inflammatory bowel disease were prospectively evaluated. Exclusion criteria were 1) use of antibiotics, NSAIDs, 5-ASA products, corticosteroids, or immunosuppressives within 2 weeks of study entry, and 2) IPAA for familial adenomatous polyposis. All subjects had measurement of fecal lactoferrin concentration (FL) and underwent pouch endoscopy with biopsy, with calculation of the pouchitis disease activity index (PDAI). Pouchitis was defined as a PDAI score of ⱖ7. Cuffitis was defined as endoscopic and histological inflammation of the rectal cuff. IPS was defined as presence of symptoms with normal pouch endoscopy and histology. Results: 15 subjects were asymptomatic and 39 had symptoms (20 had pouchitis, 3 had cuffitis, 3 had CD, and 13 had IPS). FL levels were significantly higher in patients with inflammatory conditions (median 166, Interquartile range (IQR) 72.5– 467) compared to those with IPS (median 3.1, IQR 0.9 – 6.4) or asymptomatic patients (median 6.0, IQR 1.3–11.9); P⬍ 0.001. FL distinguished patients with IPS from those with pouchitis, cuffitis or Crohn’s disease with a sensitivity of 96% and specificity of 91% at cut-off level of 13 mcg/ml. A desired sensitivity of 100% was achieved at a cut-off level of 7 mcg/ml. At this cut-off level the specificity of the test was 83% (Table). Sensitivity and specificity of fecal lactoferrin at different cut-off levels FL Cut-off level 7 mcg/ml 13 mcg/ml 21 mcg/ml 23 mcg/ml
Sensitivity
Specificity
100% 96% 92% 92%
83% 91% 92% 100%
FL, Fecal lactoferrin
Conclusions: Fecal lactoferrin can serve as a sensitive, non-invasive initial screening test in an algorithm for the evaluation of symptomatic patients with IPAA. If FL levels are low (⬍ 7 mcg/ml), IPS can be diagnosed. If FL levels are high, pouch endoscopy is warranted to distinguish between different causes of inflammation. 224 MILK INGESTION SHOULD NOT BE DISCOURAGED IN HIVRELATED DIARRHEA Jill M. Tinmouth, M.D., Richard Glazier, M.D., George A. Tomlinson, Ph.D., Sharon Walmsley, M.D., Allan H. Steinhart, M.D., Gabor P. Kandel, M.D.*. University of Toronto, Toronto, ON, Canada. Purpose: Despite the efficacy of highly active antiretroviral therapy (HAART), diarrhea remains a significant problem for HIV-infected patients. Many HIV practitioners advise empiric lactose avoidance for their patients with diarrhea but this approach may be of questionable benefit. The purpose of this trial was to test whether the effect of ingesting lactose was equivalent to placebo for HIV patients with diarrhea, regardless of underlying lactase deficiency. Methods: In this double-blind, crossover trial of HIV-infected patients with chronic diarrhea, patients underwent lactose and placebo study periods in random order. During each study period, each patient ingested 240mL (1 cup) of either low fat milk, containing 12g lactose, or lactose-free milk. An 8 hr stool collection and concurrent hydrogen breath test (HBT) were performed. The primary outcome was the mean difference in stool weights between the lactose and placebo periods. Equivalency was declared if the difference in stool weight between the two periods excluded ⫹167g/8hr with 95% certainty. Based on reported stool weights for HIV-infected
S77
patients on protease inhibitors, this value reflects a change of approximately 50% in stool output. Results: Results are reported for 49 of 50 patients studied (1 patient violated the protocol). Forty eight (98%) were male with a median age of 42 (20 – 62). Median CD4 and viral load were 390 (20 –1100) cells/mm3 and 112 (⬍50 –⬎500,000) copies/mL respectively. Forty patients (82%) were on HAART and the median number of BMs/day was 4 (1–18). Twenty patients (41%) reported avoiding dairy products. Ten patients (20%) were lactase deficient (LD) by HBT. The mean difference in stool weights between the lactose and placebo periods was ⫺41.5g/8hr (95% C.I.: ⫺78.3 to ⫺4.8). For the subset of patients who were lactase deficient, the mean difference in stool weights between study periods was 11.3g/8hr (95% C.I.: ⫺33.3 to 55.9). Conclusions: We have shown that severity of diarrhea after lactose ingestion, as measured by stool weight, is equivalent to placebo in the entire sample as well as in the lactase deficient subset. Therefore, avoidance of reasonable quantities of milk (240 mL, equivalent to 12g lactose) is not justified in HIV-infected patients with diarrhea. Given the risks of malnutrition and osteopenia in this population, these patients can be encouraged to include reasonable quantities of dairy products in their diets.
225 HIGH YIELD OF CAPSULE ENDOSCOPY IN PATIENTS WITH CELIAC DISEASE AND RECURRENT ABDOMINAL PAIN Andrea N. Culliford, M.D., Moshe Rubin, M.D., Peter H.R. Green, M.D.*. New York Presbyterian Medical Center–Columbia University, New York, NY. Purpose: Patients with Celiac Disease may continue to have symptoms of diarrhea and abdominal pain while adhering to a gluten-free diet. Our aim was to evaluate the small bowel with wireless capsule endoscopy in this group to help define the etiology of their symptoms. Methods: 20 patients with biopsy proven celiac disease with recurrent abdominal pain (N⫽14), diarrhea (N⫽13), or both (N⫽8) were evaluated with wireless capsule endoscopy. All studies were reviewed by two independent endoscopists. Results: The mean age of patients was 57 years and the mean number of years on a gluten-free diet was 5.8 years. Abnormalities seen at wireless capsule endoscopy included villous atrophy (N⫽13), mucosal fissuring (N⫽13), edema (N⫽10) and layering or stacking of folds (N⫽11), a sign that has not previoulsy been reported. The findings were predominantly in the duodenum, but did extend into the jejunum and ileum. Unexpected findings in 12 patients were as follows: a stricture not seen on prior small bowel series in one patient, ulceration of the jejunum, ileum or both in 10 patients, and 1 patient with a polypoid mass lesion. The stricture was found in a patient with abdominal pain. The ulcers were seen in patients with diarrhea (N⫽5), abdominal pain (N⫽8) or both (N⫽5). Of the patients who had abdominal pain as their presentation, 64% had ulceration or stricture. No patient had Crohn’s disease or used NSAIDS. One patient with pain and ulceration had a clonal T cell expansion identified in their duodenal biopsy raising the possibility that the ulceration was due to a lymphoma. Conclusions: Wireless capsule endoscopy is a valuable diagnostic test in the evaluation and characterization of small bowel mucosal abnormalities in patients with celiac disease who develop recurrent symptoms, especially abdominal pain, while on a gluten-free diet.
226 CELIAC DISEASE IN A COMMUNITY GASTROENTEROLOGY PRACTICE Lucı´a C. Fry, M.D., Swarnjit Singh, M.D., Klaus E. Mo¨ nkemu¨ ller, M.D.*. East Valley Gastroenterology and Hepatology, Chandler, AZ. Purpose: Although most studies in Europe have shown that the overall prevalence of celiac disease (CD) is 1:150 to 1:300, in the U.S.A. CD is
Abstracts
Purpose: Non-steroidal anti-inflammatory drugs (NSAIDs) have been implicated in small bowel ulceration and stricture formation, which are difficult to detect. We report on 12 patients referred either for obscure gastrointestinal bleeding [10] or abdominal pain [2] in which NSAID associated strictures and ulcers were detected by CE and/or intra-operative endoscopy. Methods: In the study population, the mean age was 60.2 yrs [range 18-77]; 3 were male and 9 were female. 9 had iron deficiency anemia and had received an average of 4 units of blood (range 2–20). 11 of 12 patients underwent CE after a prior workup that included a minimum of upper endoscopy [mean 1.41]; colonoscopy [mean 2.0] and all had enteroclysis or a SBFT. Reported NSAID consumption was from 1 to 30 years. 2 took NSAIDs for more than 10 years prior to CE. Results: Strictures with associated ulceration were found in all patients using CE while 2 had findings on radiography. The number of strictures per patient varied from 1 to 23. 10 of 11 patients had transient capsule retention of up to 2 weeks duration, 3 of 11 patients had capsule retention requiring surgical retrieval. 6 of 11 required surgical treatment consisting of intraoperative enteroscopy [6], small intestinal resection [6] and/or stricturoplasty [1] based on CE findings. Pathological exam of surgical specimens revealed strictures comprised of circumferential webs associated with a few discrete ulcers. Macroscopically, the most common findings were short concentric ring-like stenoses, with clearly demarcated ulceration, usually at the innermost margin of the web. In addition some strictures had no serosal representation, and were only located by CE and confirmed by intraoperative enteroscopy. The most frequent histological finding was fibromuscular hyperplasia that broadened the plica circularis. This was associated with disruption and disappearance of the muscularis propria, particularly towards the top of the plica. This was limited to the submucosa and was remarkable for the lack of inflammatory infiltrate. Conclusions: NSAID injury to the small intestine is more common than previously appreciated. CE provides a sensitive means of detecting NSAID lesions in the small intestine. Complete surgical treatment requires intraoperative enteroscopy to detect all stenoses. The pathology of these lesions may be characteristic for NSAID lesions of the small intestine and is distinct from Crohn’s disease. 221 RANDOMIZED CONTROLLED TRIALS CLAIMING EQUIVALENCY IN DIGESTIVE DISEASES: ARE THEY METHODOLOGICALLY RIGOROUS? Jill M. Tinmouth, M.D., Leah S. Steele, M.D., George A. Tomlinson, Ph.D., Richard Glazier, M.D. M.P.H.*. University of Toronto, Toronto, ON, Canada. Purpose: Traditionally, randomized controlled trials (RCTs) have attempted to demonstrate the superiority of one intervention over another. However, when efficacious treatment already exists, it may be more useful to prove that an intervention is equivalent, or at least not inferior, to the standard of care. The purpose of this paper is to determine whether claims of equivalency in digestive diseases trials are supported by the evidence. Methods: Medline (1989 –2002) was searched for RCTs using the MeSH headings “exp therapeutic equivalency” and “exp digestive diseases” and the text words “equivalence”, “equal”, “equals” or “equivalent” yielding 890 articles. Of these, 73 articles met the inclusion criteria. These articles were then reviewed using previously published criteria for equivalency (Greene et al., Ann Int Med 2000). Results: Amongst the included articles, 33% stated a priori that their aim was to test for equivalence. The magnitude of absolute differences between the interventions studied ranged from 0 to 58% for articles reporting their results as proportions (n⫽54). For articles reporting results as continuous values (n⫽16), the magnitude of relative differences between interventions ranged from 0 to 33%. Ten percent (7 of 73) of the articles described differences of 20% or more between “equal interventions”. Thirty four percent (34%) of articles set a boundary to test equivalence and confirmed it statistically while 52% used non-significant superiority tests (i.e., p-value ⬎ 0.05). Fourteen percent (14%) failed to use any statistical test to
AJG – Vol. 98, No. 9, Suppl., 2003
support their claims of equivalency. Sample size calculations were done in 30% of articles. Only 8 of 73 (11%) articles met all the criteria of stating an aim a priori, setting and testing an equivalency boundary, and calculating the necessary sample size to demonstrate equivalency. Conclusions: Trials in digestive diseases in which interventions are reported to be equivalent tend to be poorly supported by the evidence. In particular, more than half the articles mistakenly concluded that “failing to detect a significant difference” was synonymous with a finding of equivalence. Erroneous claims of equivalency are potentially dangerous and may lead to substandard patient care. 222 EFFECTS OF THE CCK1-RECEPTOR ANTAGONIST DEXLOXIGLUMIDE ON NIPPOSTRONGYLUS BRASILIENSISALTERED JEJUNAL SENSITIVITY IN RATS Lionel Bueno*, Gianfranco Caselli, Cecile Bonnafous, Patrick Griffin, Harvey Schneier, Lucio Rovati, Massimo D’Amato. INRA, Toulouse, France; Forest Research Institute, Jersey City, NJ and Rotta Research Laboratorium, Monza, Italy. Purpose: To evaluate the effects of dexloxiglumide on N. brasiliensisaltered jejunal sensitivity in rats. Methods: Male Wistar rats (8 groups of 8) were infected by subcutaneous administration of N. brasiliensis infective larvae. Rats received oral treatment with dexloxiglumide (30 or 100 mg/kg) or its vehicle (placebo) every day from Day 12 (ie, the peak of inflammatory response) to Day 27 postinfection. On Day 30, intestinal sensitivity was tested before and after administration of dexloxiglumide (5 or 20 mg/kg intraperitoneal[IP]) or its vehicle. A latex balloon inserted into the jejunum was inflated (12.5–100 mm Hg, 25 seconds every 5 minutes) to cause distention. Intraluminal pressure and volume were measured by a barostat connected to the balloon, and systemic blood pressure (BP) was recorded from a sidearm off of the left carotid cannula. A depression in BP correlated with an increase in visceral pain. Results: Dexloxiglumide treatment (Days 12–27) reduced the amplitude of the depression in BP for balloon pressures of 25, 50, and 75 mm Hg in a dose-dependent manner, reaching statistical significance (P⬍0.05) at the 100 mg/kg dose, both as a single data point at 50 mm Hg pressure, and as a global response evaluation (AUC of depressor response over applied intestinal pressure; 1350 ⫾ 128, 821 ⫾ 164, and 622 ⫾ 164, for controls, 30 mg/kg, and 100 mg/kg, respectively). Moreover, in the equation proposed by Ness and Gebhart [Brain Res. 1988;450:153 — (⌬ blood pressure)2 ⫽ (slope ⫻ distention) ⫹ intercept], the slopes of curves obtained in animals treated with 30 or 100 mg/kg dexloxiglumide (7.8 ⫾ 1.44 and 6.55 ⫾ 1.44, respectively) were significantly different from the slope of the control animals (18.58 ⫾ 2.08, P⬍0.05). IP administration of dexloxiglumide to chronically treated animals had no additional effect on the reduction of BP depression in response to jejunal distention beyond what was observed in chronically treated animals. Dexloxiglumide did not affect the change in the volume associated with increasing pressure, ie, jejunal tone. Conclusions: Chronic treatment with dexloxiglumide reduces or suppresses N. brasiliensis-induced intestinal hypersensitivity to distention. Chronic treatment with dexloxiglumide may therefore reduce visceral hypersensitivity observed in patients with IBS. 223 FECAL LACTOFERRIN DISTINGUISHES INFLAMMATORY FROM NON-INFLAMMATORY CAUSES OF SYMPTOMS IN PATIENTS WITH ILEAL-POUCH ANAL ANASTOMOSIS (IPAA) Mansour A. Parsi, M.D., Bret A. Lashner, M.D.*, Bo Shen, M.D., Jean-Paul Achkar, M.D., Feza Remzi, M.D., John Goldblum, M.D., Dahai Lin, B.S., Jason T. Connor, M.S., Dhanasekaran Ramasamy, M.D., Nish Shah, M.D., Marlene Bambrick, Victor W. Fazio, M.D. The Cleveland Clinic Foundation, Cleveland, OH.
AJG – September, Suppl., 2003
Abstracts
Purpose: Symptoms of increased stool frequency, urgency, and abdominal pain in patients with IPAA, can be due to inflammatory conditions such as pouchitis, cuffitis, or Crohn’s disease (CD) or to non-inflammatory conditions such as irritable pouch syndrome (IPS). Differentiating between these entities usually requires pouch endoscopy and biopsy. Non-invasive means of diagnosis are preferable. Methods: 54 consecutive subjects who had undergone proctocolectomy with IPAA for inflammatory bowel disease were prospectively evaluated. Exclusion criteria were 1) use of antibiotics, NSAIDs, 5-ASA products, corticosteroids, or immunosuppressives within 2 weeks of study entry, and 2) IPAA for familial adenomatous polyposis. All subjects had measurement of fecal lactoferrin concentration (FL) and underwent pouch endoscopy with biopsy, with calculation of the pouchitis disease activity index (PDAI). Pouchitis was defined as a PDAI score of ⱖ7. Cuffitis was defined as endoscopic and histological inflammation of the rectal cuff. IPS was defined as presence of symptoms with normal pouch endoscopy and histology. Results: 15 subjects were asymptomatic and 39 had symptoms (20 had pouchitis, 3 had cuffitis, 3 had CD, and 13 had IPS). FL levels were significantly higher in patients with inflammatory conditions (median 166, Interquartile range (IQR) 72.5– 467) compared to those with IPS (median 3.1, IQR 0.9 – 6.4) or asymptomatic patients (median 6.0, IQR 1.3–11.9); P⬍ 0.001. FL distinguished patients with IPS from those with pouchitis, cuffitis or Crohn’s disease with a sensitivity of 96% and specificity of 91% at cut-off level of 13 mcg/ml. A desired sensitivity of 100% was achieved at a cut-off level of 7 mcg/ml. At this cut-off level the specificity of the test was 83% (Table). Sensitivity and specificity of fecal lactoferrin at different cut-off levels FL Cut-off level 7 mcg/ml 13 mcg/ml 21 mcg/ml 23 mcg/ml
Sensitivity
Specificity
100% 96% 92% 92%
83% 91% 92% 100%
FL, Fecal lactoferrin
Conclusions: Fecal lactoferrin can serve as a sensitive, non-invasive initial screening test in an algorithm for the evaluation of symptomatic patients with IPAA. If FL levels are low (⬍ 7 mcg/ml), IPS can be diagnosed. If FL levels are high, pouch endoscopy is warranted to distinguish between different causes of inflammation. 224 MILK INGESTION SHOULD NOT BE DISCOURAGED IN HIVRELATED DIARRHEA Jill M. Tinmouth, M.D., Richard Glazier, M.D., George A. Tomlinson, Ph.D., Sharon Walmsley, M.D., Allan H. Steinhart, M.D., Gabor P. Kandel, M.D.*. University of Toronto, Toronto, ON, Canada. Purpose: Despite the efficacy of highly active antiretroviral therapy (HAART), diarrhea remains a significant problem for HIV-infected patients. Many HIV practitioners advise empiric lactose avoidance for their patients with diarrhea but this approach may be of questionable benefit. The purpose of this trial was to test whether the effect of ingesting lactose was equivalent to placebo for HIV patients with diarrhea, regardless of underlying lactase deficiency. Methods: In this double-blind, crossover trial of HIV-infected patients with chronic diarrhea, patients underwent lactose and placebo study periods in random order. During each study period, each patient ingested 240mL (1 cup) of either low fat milk, containing 12g lactose, or lactose-free milk. An 8 hr stool collection and concurrent hydrogen breath test (HBT) were performed. The primary outcome was the mean difference in stool weights between the lactose and placebo periods. Equivalency was declared if the difference in stool weight between the two periods excluded ⫹167g/8hr with 95% certainty. Based on reported stool weights for HIV-infected
S77
patients on protease inhibitors, this value reflects a change of approximately 50% in stool output. Results: Results are reported for 49 of 50 patients studied (1 patient violated the protocol). Forty eight (98%) were male with a median age of 42 (20 – 62). Median CD4 and viral load were 390 (20 –1100) cells/mm3 and 112 (⬍50 –⬎500,000) copies/mL respectively. Forty patients (82%) were on HAART and the median number of BMs/day was 4 (1–18). Twenty patients (41%) reported avoiding dairy products. Ten patients (20%) were lactase deficient (LD) by HBT. The mean difference in stool weights between the lactose and placebo periods was ⫺41.5g/8hr (95% C.I.: ⫺78.3 to ⫺4.8). For the subset of patients who were lactase deficient, the mean difference in stool weights between study periods was 11.3g/8hr (95% C.I.: ⫺33.3 to 55.9). Conclusions: We have shown that severity of diarrhea after lactose ingestion, as measured by stool weight, is equivalent to placebo in the entire sample as well as in the lactase deficient subset. Therefore, avoidance of reasonable quantities of milk (240 mL, equivalent to 12g lactose) is not justified in HIV-infected patients with diarrhea. Given the risks of malnutrition and osteopenia in this population, these patients can be encouraged to include reasonable quantities of dairy products in their diets.
225 HIGH YIELD OF CAPSULE ENDOSCOPY IN PATIENTS WITH CELIAC DISEASE AND RECURRENT ABDOMINAL PAIN Andrea N. Culliford, M.D., Moshe Rubin, M.D., Peter H.R. Green, M.D.*. New York Presbyterian Medical Center–Columbia University, New York, NY. Purpose: Patients with Celiac Disease may continue to have symptoms of diarrhea and abdominal pain while adhering to a gluten-free diet. Our aim was to evaluate the small bowel with wireless capsule endoscopy in this group to help define the etiology of their symptoms. Methods: 20 patients with biopsy proven celiac disease with recurrent abdominal pain (N⫽14), diarrhea (N⫽13), or both (N⫽8) were evaluated with wireless capsule endoscopy. All studies were reviewed by two independent endoscopists. Results: The mean age of patients was 57 years and the mean number of years on a gluten-free diet was 5.8 years. Abnormalities seen at wireless capsule endoscopy included villous atrophy (N⫽13), mucosal fissuring (N⫽13), edema (N⫽10) and layering or stacking of folds (N⫽11), a sign that has not previoulsy been reported. The findings were predominantly in the duodenum, but did extend into the jejunum and ileum. Unexpected findings in 12 patients were as follows: a stricture not seen on prior small bowel series in one patient, ulceration of the jejunum, ileum or both in 10 patients, and 1 patient with a polypoid mass lesion. The stricture was found in a patient with abdominal pain. The ulcers were seen in patients with diarrhea (N⫽5), abdominal pain (N⫽8) or both (N⫽5). Of the patients who had abdominal pain as their presentation, 64% had ulceration or stricture. No patient had Crohn’s disease or used NSAIDS. One patient with pain and ulceration had a clonal T cell expansion identified in their duodenal biopsy raising the possibility that the ulceration was due to a lymphoma. Conclusions: Wireless capsule endoscopy is a valuable diagnostic test in the evaluation and characterization of small bowel mucosal abnormalities in patients with celiac disease who develop recurrent symptoms, especially abdominal pain, while on a gluten-free diet.
226 CELIAC DISEASE IN A COMMUNITY GASTROENTEROLOGY PRACTICE Lucı´a C. Fry, M.D., Swarnjit Singh, M.D., Klaus E. Mo¨ nkemu¨ ller, M.D.*. East Valley Gastroenterology and Hepatology, Chandler, AZ. Purpose: Although most studies in Europe have shown that the overall prevalence of celiac disease (CD) is 1:150 to 1:300, in the U.S.A. CD is