- Email: [email protected]
S1150 Influence of Parental Height in Disease Characteristics of Pediatric-Onset Inflammatory Bowel Disease
S1148
AGA Abstracts
A Comparison Between Solid-State vs. Water Perfused Catheters for Colonic Manometry in Children Olivia Liem, Frances L. Connor, Narasimha S. Reddy, Hayat M. Mousa, Marc A. Benninga, Carlo Di Lorenzo BACKGROUND: Solid state manometry catheters with portable data loggers offer many potential advantages over traditional water-perfused systems, such as prolonged recordings in a more physiologic ambulatory setting and the lack of risk for water overload /intoxication. The use of solid state catheters has not been evaluated in comparison to perfused catheters in children. AIM: To compare data provided by solid state (SS) and water-perfused (WP) catheters in children undergoing colonic manometry studies. METHODS: The catheters were attached together such that ports were at the same location and simultaneous recordings were obtained using both SS and WP catheters (both 8 channels, 10cm apart) in 9 children with chronic constipation or fecal incontinence who were referred for colonic manometry. Signals were recorded for a minimum of 1 h during fasting, 1 h after ingestion of a meal and 1 h after administration of bisacodyl (0.2mg/kg, max 5mg). SS signals from the data logger were analyzed against the perfused signals. All high amplitude propagated contractions (HAPC), the most important feature of normal colonic motility, were evaluated for spatial and temporal features including their duration, amplitudes and propagation velocities. RESULTS: SS signals showed more stable baselines than WP signals. A total of 48 HAPCs were detected with SS and 32 with WP. All WP-HAPC were also observed with SS. The median amplitude of SS-HAPC was not different from WP-HAPC (110 vs. 112 mmHg). The duration of SS-HAPC, however, was greater from that of WP-HAPC (22.3 vs. 18.6s, p<0.001). The velocity of HAPC propagation did not differ between the two catheters (SS: 0.69 vs. WP: 0.65 cm/sec). CONCLUSIONS: SS catheters connected to a portable data logger are more sensitive in recording HAPC in children compared to the more traditional WP assembly. Solid-state catheters offer potential advantages over WP catheters in children, being portable, safer to use and may provide data over a more prolonged period of time.
S1146 No Effect of Gut-Directed Hypnotherapy On Rectal Sensitivity in Children with Functional Abdominal Pain and Irritable Bowel Syndrome Maartje M. van den Berg, Arine M. Vlieger, Carla Menko-Frankenhuis, Marloes E. Bongers, Marc A. Benninga Introduction: Gut-directed hypnotherapy has recently been shown to be highly effective in treating children with functional abdominal pain (FAP) and irritable bowel syndrome (IBS). This study was conducted to determine to what extend this treatment success is due to improvement of rectal sensitivity. Methods: 46 patients (8-18 years) with FAP (n=28) and IBS (n=18), according to the ROME II criteria, were randomized to either 12 weeks of standard medical therapy (SMT) or gut-directed hypnotherapy (HT). To assess rectal sensitivity a pressure-controlled intermittent distension protocol (barostat) was performed before and after therapy. Rectal hypersensitivity (RHS) was defined as pain thresholds ≤ 9 mmHg above minimal distension pressure. Results: RHS was found in 6/23 (27%) patients in both treatment groups at baseline. No difference in rectal pain thresholds was identified between IBS and FAP patients (p=0.87). After treatment, clinical remission (> 80% improvement in abdominal pain scores) was seen in 13% SMT patients compared to 57% HT patients (p= 0.005). Similar treatment outcomes were found in patients with and without RHS (p=0.35). Rectal pain thresholds had not changed significantly after treatment in both groups. In SMT patients pain thresholds changed from 16.6 (7.8) mmHg (mean (SD)) at baseline to 18.6 (8.5) mmHg at 12 weeks (p=0.10) and in HT patients from 20.7 (11.5) mmHg to 22.5 (10.1) mmHg (p=0.10). Subgroup analysis in patients with RHS showed a significant increase in pain thresholds from 7.0 (2.5) mmHg to 15.5 (9.9) mmHg in SMT patients (p=0.02) and from 7.5 (1.6) mmHg to 25.5 (12.8) mmHg in HT patients (p=0.03), but this increase in pain thresholds was not significant different between treatment groups (p=0.09) and not related to treatment success (p=0.60). Conclusion: In this small patient sample, clinical success achieved with HT can not be explained by improvement of rectal sensitivity. In contrast to earlier findings hypersensitivity of the rectum was found in a minority of children with IBS. Further studies are needed to confirm our finding that no association exists between rectal barostat findings and clinical symptoms in children with pain related functional gastrointestinal disorders.
S1149 Magnetic Resonance Imaging of the Lumbosacral Spine in Children with Chronic Constipation and Functional Non-Retentive Fecal Incontinence: A Prospective Study Noor Bekkali, Marloes E. Bongers, Eveline E. Hagebeuk, Michiel P. van Wijk, Rick R. Van Rijn, Marc A. Benninga Background: In a retrospective study magnetic resonance imaging (MRI) of the spine demonstrated an incidence of 9% of spinal abnormalities in children with intractable constipation. In asymptomatic adults, incidental spinal abnormalities such as lipomas are found in 1.55%. Aim: To prospectively determine the prevalence of lumbosacral spine abnormalities in children with constipation or functional non-retentive fecal incontinence (FNRFI). Methods: Pediatric patients aged 6-18 years referred for functional constipation or FNRFI (ROME III criteria) to our tertiary center were included. A pediatric neurologist performed complete neurological examination, with special attention for the lumbo-sacral spine region. MRI of the lumbo-sacral spine was performed. The neurologist was informed about the clinical status, but not about the MRI outcome. Results: Seventy four patients with constipation were included (46% male; age (±SD) 11.2 (±2.9) years). Mean symptom duration was 41.7 (±32.1) months. In none of patients abnormalities were found on neurological examination. One occult spina bifida and one terminal filum lipoma were seen on MRI. Both constipated patients had no urinary incontinence at intake. Defecation frequency normalized and fecal incontinence disappeared after medical and behavioral treatment in both children. Thirteen children with FNRFI were included (53.8% male; age 9.9 (±1.21) years) with mean symptom duration of 34.0 (±35.3) months. Again, no abnormalities were found on neurological examination. MRI showed a terminal filum lipoma in one child. This child had urinary incontinence at intake. Fecal and urinary incontinence disappeared after behavioral treatment. No neurosurgical treatment was performed in those children due to clinical improvement with conventional therapy. Conclusion: Spinal cord abnormalities were found in 2.7% of constipated children and 7.7% of FNRFI patients. Overall, terminal filum lipomas were found in 2.3%. It remains unclear whether the abnormalities found, play a role in the pathophysiology of constipation and/or FNRFI. More children are needed to reveal the true prevalence of spinal cord abnormalities in these children.
S1147 Gastrointestinal (GI) Permeability Is Associated with Trait Anxiety in Children with Functional Abdominal Pain (FAP) and Irritable Bowel Syndrome (IBS) Robert J. Shulman, Danita I. Czyzewski, Mariella Lane, Monica Jarrett, Robert L. Burr Background: FAP and IBS affect 10-15% of school age children and bear many physiological similarities to irritable bowel syndrome (IBS) in adults (e.g., functional pain, visceral hyperalgesia). Animal models of IBS have suggested a relationship between neonatal stress and increased GI permeability later in life (Pediatr Res 2007;62: 240). We hypothesized that psychological and physiological measures related to the experience of stress would predict GI permeability in children with FAP/IBS. Design/Methods: Children (age 7-10 yr.) were identified by chart review in pediatrician's or pediatric gastroenterologist's offices. Children (n=99) met Rome II criteria for FAP or IBS. Phone screening confirmed current symptoms. After instruction in the home by research assistants the children completed the State-Trait Anxiety Inventory for Children (STAIC) which measures both stable tendencies to experience anxiety (Trait) and current level of anxiety (State). They then underwent measurement of GI permeability to measure gastric (sucrose), small intestinal (lactulose/mannitol ratio), and colonic (sucralose/lactulose ratio) permeability. A first morning urine collection was obtained to measure norepinephrine, cortisol, and creatinine excretion. Results: The mean age of the children was 8.5 ± 0.1 yr. (mean ± SEM). Mean percent recovery for sucrose was 0.023 ± 0.002. The lactulose/mannitol and sucralose/lactulose ratios (per m2) were 0.064 ± 0.004 and 1.134 ± 0.104, respectively. Mean urinary norepinephrine/creatinine ratio was 22.5 ± 0.0 and cortisol/creatinine ratio was 34.3 ± 0.0. A series of regression equations were performed. In the first, the lactulose/mannitol ratio (i.e., small intestinal permeability) was predicted by the Trait anxiety score: F(1, 91) = 5.23; P = 0.025, r = 0.21). When urinary norepinephrine/creatinine ratio and cortisol/creatinine ratio were included in the model the prediction of permeability was stronger: F(3, 73) = 4.73; P = 0.005, r = 0.37. There was no relationship between Trait anxiety score and gastric or colonic permeability or between State anxiety score and GI permeability. Conclusions: There is a positive relationship between the stable tendency to experience anxiety states and small intestinal permeability in children with FAP/IBS. Catecholamine and cortisol excretion magnify the relationship. These data lend support to the hypothesis that stress (anxiety, norepinephrine and cortisol levels) may affect GI permeability. Whether this is related to pre- and/or perinatal stress in children needs to be determined.
AGA Abstracts
S1150 Influence of Parental Height in Disease Characteristics of Pediatric-Onset Inflammatory Bowel Disease Jessica J. Lee, Johanna C. Escher, Melissa J. Shuman, Peter Forbes, Richard J. Grand Background: Impairment of linear growth remains a major challenge to patients with pediatric-onset inflammatory bowel disease (IBD). Despite development of advanced nutritional and medical therapy, the prevalence of growth failure (GF) in IBD has not decreased over the past decade. Although GF in IBD has been attributed mostly to nutritional and disease-related factors, genetics may be the critical predetermining factor. The role of parental height in the growth pattern of pediatric-onset IBD has not been addressed previously. Aim: To identify the impact of parental height and disease-related factors on growth in children with IBD. Methods: 246 probands with IBD diagnosed between the ages of 1-18 (177 Crohn's disease (CD), 62 ulcerative colitis (UC), and 7 indeterminate colitis (IC)) were included. Biological family members (234 mothers, 190 fathers, and 38 siblings) were included to study the family growth phenotypes. Results: GF was defined as height for age Z-score < -1.64 in more than one measurement since diagnosis. Fifty-six patients (22.9%) had GF, and most had CD (87.5% CD vs. 8.9% UC and 3.6% IC). There was a significant left-shift in the mean height Z-score for GF group (Z-score -1.68, p<0.001, whereas, the mean Z-score for non-growth failure (NGF) group was -0.05. GF and NGF groups did not have statistically significant differences in gender, age of onset, disease behavior, disease location, extraintestinal manifestations, history of operations, family history of IBD, or use of corticosteroids. Both mothers and fathers of the GF group showed significantly lower
A-188
last follow-up (OR5.08 [2.73-9.45]) and conversely less proctitis (E1) (1% v 17%, OR0.07 [0.01-0.49]). Unless stated otherwise, all p values <0.00001. Conclusion: Children with IBD have a more extensive disease phenotype at follow-up. The reasons for the stark contrast between children and adults with IBD may reflect differences in a combination of genetic and environmental factors. S1153 An Induction Dose of 12 Mg/D Budesonide Improves Clinical Efficacy in Comparison to a Standard Continuous Dose of 9 mg/d Active Pediatric Crohn's Disease Arie Levine, Michal Kori, Gabriel Dinari, Efrat Broide, Ron Shaoul, Baruch Yerushalmi, Markus Proels, Roland Greinwald AIM: To compare the efficacy and tolerability of two doses of oral budesonide in pediatric patients with active Crohn's disease. METHODS: Pediatric patients with active Crohn's disease (Pediatric Crohn's disease activity index [PCDAI] at baseline: > 12.5 and < 40) were eligible for this double-blind, randomized, multicenter study. Patients received 9 mg/d (n= 35) oral budesonide for 7 weeks or 12 mg/d for 4 weeks followed by 9 mg/d (n=35) for further 3 weeks. Primary endpoint was the mean change of PCDAI from baseline to week 7 between the treatment groups. All patients received 6 mg/d for an additional 3 weeks. RESULTS: 71 patients were randomized. 70 patients were evaluable for the intention-totreat (= ITT)-analysis and 61 patients for the per-protocol (= PP) analysis.Comparative data for the primary and secondary end points in the two groups are presented in Table 1 and Table 2.Serious adverse events occured in 28.6% of the 9 mg/day group and in 17.1% of the 12 mg/day group, however there were no drug related serious adverse events in either group. CONCLUSION: Oral budesonide was effective for inducing remission with both doses. An induction dose of 12 mg/d oral budesonide followed by 9 mg/day appears to improve clinical efficacy in mild to moderately active pediatric Crohn's disease. Table 1.Primary efficacy endpoint: Mean change of PCDAI from baseline to week 7
S1151 A Detailed Investigation Into Epidemiological Risk Factors for Childhood Onset Inflammatory Bowel Disease in Scotland Johan Van Limbergen, Elizabeth Hobbs, Richard K. Russell, Elaine Nimmo, Hazel E. Drummond, Linda Smith, Peter M. Gillett, Paraic McGrogan, Lawrence Weaver, Michael W. Bisset, Gamal Mahdi, Jack Satsangi, David C. Wilson Introduction: The incidence of inflammatory bowel disease (IBD) is increasing among children in Scotland and is now among the highest worldwide. Both environmental and genetic risk factors are implicated in the aetiology of IBD. We aimed to investigate the association of asthma, vaccinations and breastfeeding with paediatric IBD in Scotland. Methods: 439 children diagnosed with IBD <17 y (274 CD, 101 UC, 54 IBDU; median age 11.2 years (Q1-Q3: 8.7-13.1) were compared with more than 25,000 population-matched controls. Children with IBD and their parents were interviewed face-to-face to obtain data on breastfeeding, immunisation history (MMR, diphtheria, tetanus, pertussis, HIB3) and medical history. Control breastfeeding data for children born in 1996 was collected at the health visitor's first review visit (around 10 days old) and were available for Lothian and Greater Glasgow (Information Services Division of NHS Scotland). Immunisation history was recorded at 24 months for children born in 1993 in Lothian, Greater Glasgow and Grampian (idem). Doctor-diagnosed asthma prevalence data in children aged 2-15 from across the whole of Scotland were provided by the Scottish Health Survey 2003. Unifactorial analyses were performed using the χ2-test. Breastfeeding data were stratified for geographical region and affluence (using the Carstairs Deprivation Score - DepCat). Results: History of asthma was associated with IBD and CD (115/439 (26%) and 74/274 (27%) v 521/2965 (18%) in controls: p <0.0001, OR 1.67 (1.32-2.10) and p=0.0002, OR 1.74 (1.31-2.30)). Analysis of controls showed marked differences in breastfeeding rates (>1 week) between Lothian and Greater Glasgow (4368/8196 (53%) v 2270/6775 (33%) p<10-4 OR 2.26 (2.12-2.42)). After stratification for geographical region, there was no significant difference of breastfeeding rates between IBD/CD/UC cases v controls (all p>0.05). Breastfeeding rate in IBD cases (stratified by DepCat (1-2, 3-5, 6-7)) was not different from controls (all p>0.05). Breastfeeding rates in IBD cases were significantly different between DepCat 1-2 and DepCat 6-7 (53/79 (67%) v 17/67 (25%), p<10-4 OR 6.00 (2.91-12.36)). This difference was also present in controls (4226/6648 (64%) v 1768/7750 (23%) p<10-4 OR 5.90 (5.49-6.35)). There was no significant association between immunisation history and IBD/CD/UC (p>0.05). Conclusion: In the high incidence Scottish population, we have shown the novel association of childhood onset IBD and CD with asthma. We did not observe any association between paediatric IBD/CD/UC and either immunisation history or breastfeeding after stratification for geographical region and affluence.
Mean change of PCDAI from baseline to week 7 compared between the treatment groups did not reach significance (ITT analysis: p = 0.2571; PP analysis: p = 0.0895; two-sample t-test). Table 2.Secondary efficacy endpoints (ITT) at week 7:
S1154 A Comprehensive Review of the English Language Evidence Base for the Management of Paediatric Inflammatory Bowel Disease (IBD) Illustrates the Difficulties for Evidence-Based Clinical Guideline Formation David C. Wilson, Working Group Inflammatory Bowel Disease
S1152 Childhood Onset Inflammatory Bowel Disease Has a More Extensive Disease Phenotype in Comparison to Adult Onset IBD Using the Montreal Classification Johan Van Limbergen, Richard K. Russell, Hazel E. Drummond, Nicola Round, Linda Smith, Elaine Nimmo, Ian D. Arnott, Peter M. Gillett, Paraic McGrogan, Lawrence Weaver, Michael W. Bisset, Gamal Mahdi, David C. Wilson, Jack Satsangi
Background and aim: Clinical guidelines, coupled with clinical experience and expertise, play an important role in the delivery of appropriate, effective and efficient health care. We aimed to develop evidence-based clinical guidelines for the medical and nutritional management of paediatric IBD; the first step was comprehensive review of the evidence base. Methods: A search strategy was developed to identify existing guidelines and evidence of efficacy of medical and nutritional treatment of children and adolescents (aged <18 years) with IBD. A computer assisted search of the English language literature for drug and nutritional interventions in this group was performed using the following on-line databases: MEDLINE (1966-2006), EMBASE (1980-2006), CINAHL (1983-2006), Cochrane Library, Web of Science - Science Citation Index (1981-2006), and BIOSIS Previews. Abstracts of major paediatric gastroenterology meetings were searched and recognised experts in the field contacted. The titles and abstracts were examined by 2 reviewers to exclude irrelevant publications. Full manuscripts of the remaining papers were critically appraised by at least 2 reviewers using the SIGN criteria (Br Med J 2001; 323: 334-6). Results: There were 3 evidence-based guidelines for management of IBD identified, none of the highest methodological quality. 5227 publications were identified to end-December 2006, but 4849 were excluded on initial examination of title and abstract. Many manuscripts were excluded as not relevant (mainly because they did not include children or childhood and adult data were inseparable), so 224 publications were critically appraised. 1 systematic review (of nutritional management versus steroids for Crohn's disease) and 13 RCTs were identified, but all had major methodological flaws (Evidence Level 1-); methodological problems abound, such as lack of details on randomisation, lack of intention to treat analysis, and small sample sizes. All other studies were poor quality cohort studies (EL 2-), case series (EL 3) or expert opinion (EL 4). No evidence-based recommendations could therefore be made. Summary and Conclusions: Despite the widespread prevalence of paediatric IBD, comprehensive review reveals a paucity of evidence on the effectiveness of interventions, and a great need for adequately powered, high quality studies. We can provide no evidencebased recommendations for practice, and instead will proceed to formal and methodologically robust consensus-based guidelines, which will be informed by available paediatric and high quality adult evidence, and multidisciplinary experience and expertise.
Introduction: The Montreal classification of IBD was designed to address the problems associated with the Vienna classification in describing childhood onset CD, involvement of the upper GI tract and perianal disease. We compared the phenotypic characteristics of a paediatric IBD cohort with an adult cohort from the same homogenous population at last follow-up. Methods: 1713 IBD patients (416 IBD patients <17 y at diagnosis (276 CD, 99 UC, 41 IBDU) and 1297 adults (>17 y) (596 CD, 701 UC)) were phenotyped. In paediatric IBD, the median duration of follow-up (Q1-Q3) was 3.7 years (1.7 - 6.0) (CD) and 3.5 years (1.1-4.8) (UC). In adult IBD, median duration of follow-up was 10.3 years (3.8-20.6) (CD) and 8.9 years (4.2-16.5) (UC). The Montreal classification was applied to score CD location (L1-ileal, L2-colonic, L3-ileocolonic & L4-upper GI disease); CD behaviour (B1inflammatory, B2-stricturing, B3-penetrating); L4 was added as a modifier to L1-3; the suffix p was added to B1-3 for perianal CD involvement; UC extent was defined as (E3-extensive, E2-left sided, E1-proctitis). 98% of patients were Caucasian. Results: Compared to adult onset IBD, childhood onset IBD was characterised by more familial disease (34% v 20% OR2.1 [1.6-2.8]) and less smokers (1% v 18% OR0.05 [0.02-0.14]). Erythema nodosum was more common in children than in adults (8% v 4% p= 0.007 OR2.3 [1.2-4.1]). Childhood onset CD featured a strong male predominance (59% v 36%, OR2.6 [1.9-3.4]). At last follow-up, disease location was more extensive in children with more (a) L2+L4 and L3+L4 (16% v 2% OR11.6 [5.4-25.2] and 43% v 3% OR23.3 [13.4-40.6]) and (b) less L1 and L2 (3% v 32% OR0.06 [0.03 -0.1] and 15% v 36% OR0.3 [0.2 -0.5] respectively). CD behaviour at follow-up showed no differences when analysed at the same time interval post diagnosis. 53/196 (27%) of childhood CD patients already had L3+L4 disease at diagnosis. Within the first two years after diagnosis, 56/196 (29%) progressed to more extensive disease. These changes were not due to the inclusion of findings on upper GI endoscopy in 41/56 (73%) but due to progression from local oral/perianal, L1, L2 or L4 disease to L3 disease. Childhood onset UC was characterised by extensive colitis (E3) in 82% of children v 48% of adults at
A-189
AGA Abstracts
AGA Abstracts
mean height Z-scores than their NGF counterparts (-0.63 vs. 0.08 for mothers; -0.36 vs. 0.28 for fathers; p<0.001). GF patients had significantly lower mean target heights based on measured parental heights (172.5 cm for males and 160 cm for females) compared to NGF patients (177.4 cm for males and 165.2 cm for females). Of 107 patients who had reached ≥ 18 years of age after study enrollment, mean final height Z-score was -1.29 for the GF group as compared to 0.06 in the NGF group. For both growth groups, the ratio of measured final height to target height was close to 1 (0.97 for GF and 0.99 for NGF). Of 31 GF patients who reached adult heights, only 9 (29%) remained in the GF group whereas 22 (71%) did not. Conclusion: Parental heights were significantly lower in GF patients with pediatric-onset IBD, and there was no significant difference between target height and patients' observed final height. Most GF patients reached their target height and were no longer growth impaired when final height was reached. Hence, genetics may be the key determinant of growth in IBD.
AGA Abstracts
A Comparison Between Solid-State vs. Water Perfused Catheters for Colonic Manometry in Children Olivia Liem, Frances L. Connor, Narasimha S. Reddy, Hayat M. Mousa, Marc A. Benninga, Carlo Di Lorenzo BACKGROUND: Solid state manometry catheters with portable data loggers offer many potential advantages over traditional water-perfused systems, such as prolonged recordings in a more physiologic ambulatory setting and the lack of risk for water overload /intoxication. The use of solid state catheters has not been evaluated in comparison to perfused catheters in children. AIM: To compare data provided by solid state (SS) and water-perfused (WP) catheters in children undergoing colonic manometry studies. METHODS: The catheters were attached together such that ports were at the same location and simultaneous recordings were obtained using both SS and WP catheters (both 8 channels, 10cm apart) in 9 children with chronic constipation or fecal incontinence who were referred for colonic manometry. Signals were recorded for a minimum of 1 h during fasting, 1 h after ingestion of a meal and 1 h after administration of bisacodyl (0.2mg/kg, max 5mg). SS signals from the data logger were analyzed against the perfused signals. All high amplitude propagated contractions (HAPC), the most important feature of normal colonic motility, were evaluated for spatial and temporal features including their duration, amplitudes and propagation velocities. RESULTS: SS signals showed more stable baselines than WP signals. A total of 48 HAPCs were detected with SS and 32 with WP. All WP-HAPC were also observed with SS. The median amplitude of SS-HAPC was not different from WP-HAPC (110 vs. 112 mmHg). The duration of SS-HAPC, however, was greater from that of WP-HAPC (22.3 vs. 18.6s, p<0.001). The velocity of HAPC propagation did not differ between the two catheters (SS: 0.69 vs. WP: 0.65 cm/sec). CONCLUSIONS: SS catheters connected to a portable data logger are more sensitive in recording HAPC in children compared to the more traditional WP assembly. Solid-state catheters offer potential advantages over WP catheters in children, being portable, safer to use and may provide data over a more prolonged period of time.
S1146 No Effect of Gut-Directed Hypnotherapy On Rectal Sensitivity in Children with Functional Abdominal Pain and Irritable Bowel Syndrome Maartje M. van den Berg, Arine M. Vlieger, Carla Menko-Frankenhuis, Marloes E. Bongers, Marc A. Benninga Introduction: Gut-directed hypnotherapy has recently been shown to be highly effective in treating children with functional abdominal pain (FAP) and irritable bowel syndrome (IBS). This study was conducted to determine to what extend this treatment success is due to improvement of rectal sensitivity. Methods: 46 patients (8-18 years) with FAP (n=28) and IBS (n=18), according to the ROME II criteria, were randomized to either 12 weeks of standard medical therapy (SMT) or gut-directed hypnotherapy (HT). To assess rectal sensitivity a pressure-controlled intermittent distension protocol (barostat) was performed before and after therapy. Rectal hypersensitivity (RHS) was defined as pain thresholds ≤ 9 mmHg above minimal distension pressure. Results: RHS was found in 6/23 (27%) patients in both treatment groups at baseline. No difference in rectal pain thresholds was identified between IBS and FAP patients (p=0.87). After treatment, clinical remission (> 80% improvement in abdominal pain scores) was seen in 13% SMT patients compared to 57% HT patients (p= 0.005). Similar treatment outcomes were found in patients with and without RHS (p=0.35). Rectal pain thresholds had not changed significantly after treatment in both groups. In SMT patients pain thresholds changed from 16.6 (7.8) mmHg (mean (SD)) at baseline to 18.6 (8.5) mmHg at 12 weeks (p=0.10) and in HT patients from 20.7 (11.5) mmHg to 22.5 (10.1) mmHg (p=0.10). Subgroup analysis in patients with RHS showed a significant increase in pain thresholds from 7.0 (2.5) mmHg to 15.5 (9.9) mmHg in SMT patients (p=0.02) and from 7.5 (1.6) mmHg to 25.5 (12.8) mmHg in HT patients (p=0.03), but this increase in pain thresholds was not significant different between treatment groups (p=0.09) and not related to treatment success (p=0.60). Conclusion: In this small patient sample, clinical success achieved with HT can not be explained by improvement of rectal sensitivity. In contrast to earlier findings hypersensitivity of the rectum was found in a minority of children with IBS. Further studies are needed to confirm our finding that no association exists between rectal barostat findings and clinical symptoms in children with pain related functional gastrointestinal disorders.
S1149 Magnetic Resonance Imaging of the Lumbosacral Spine in Children with Chronic Constipation and Functional Non-Retentive Fecal Incontinence: A Prospective Study Noor Bekkali, Marloes E. Bongers, Eveline E. Hagebeuk, Michiel P. van Wijk, Rick R. Van Rijn, Marc A. Benninga Background: In a retrospective study magnetic resonance imaging (MRI) of the spine demonstrated an incidence of 9% of spinal abnormalities in children with intractable constipation. In asymptomatic adults, incidental spinal abnormalities such as lipomas are found in 1.55%. Aim: To prospectively determine the prevalence of lumbosacral spine abnormalities in children with constipation or functional non-retentive fecal incontinence (FNRFI). Methods: Pediatric patients aged 6-18 years referred for functional constipation or FNRFI (ROME III criteria) to our tertiary center were included. A pediatric neurologist performed complete neurological examination, with special attention for the lumbo-sacral spine region. MRI of the lumbo-sacral spine was performed. The neurologist was informed about the clinical status, but not about the MRI outcome. Results: Seventy four patients with constipation were included (46% male; age (±SD) 11.2 (±2.9) years). Mean symptom duration was 41.7 (±32.1) months. In none of patients abnormalities were found on neurological examination. One occult spina bifida and one terminal filum lipoma were seen on MRI. Both constipated patients had no urinary incontinence at intake. Defecation frequency normalized and fecal incontinence disappeared after medical and behavioral treatment in both children. Thirteen children with FNRFI were included (53.8% male; age 9.9 (±1.21) years) with mean symptom duration of 34.0 (±35.3) months. Again, no abnormalities were found on neurological examination. MRI showed a terminal filum lipoma in one child. This child had urinary incontinence at intake. Fecal and urinary incontinence disappeared after behavioral treatment. No neurosurgical treatment was performed in those children due to clinical improvement with conventional therapy. Conclusion: Spinal cord abnormalities were found in 2.7% of constipated children and 7.7% of FNRFI patients. Overall, terminal filum lipomas were found in 2.3%. It remains unclear whether the abnormalities found, play a role in the pathophysiology of constipation and/or FNRFI. More children are needed to reveal the true prevalence of spinal cord abnormalities in these children.
S1147 Gastrointestinal (GI) Permeability Is Associated with Trait Anxiety in Children with Functional Abdominal Pain (FAP) and Irritable Bowel Syndrome (IBS) Robert J. Shulman, Danita I. Czyzewski, Mariella Lane, Monica Jarrett, Robert L. Burr Background: FAP and IBS affect 10-15% of school age children and bear many physiological similarities to irritable bowel syndrome (IBS) in adults (e.g., functional pain, visceral hyperalgesia). Animal models of IBS have suggested a relationship between neonatal stress and increased GI permeability later in life (Pediatr Res 2007;62: 240). We hypothesized that psychological and physiological measures related to the experience of stress would predict GI permeability in children with FAP/IBS. Design/Methods: Children (age 7-10 yr.) were identified by chart review in pediatrician's or pediatric gastroenterologist's offices. Children (n=99) met Rome II criteria for FAP or IBS. Phone screening confirmed current symptoms. After instruction in the home by research assistants the children completed the State-Trait Anxiety Inventory for Children (STAIC) which measures both stable tendencies to experience anxiety (Trait) and current level of anxiety (State). They then underwent measurement of GI permeability to measure gastric (sucrose), small intestinal (lactulose/mannitol ratio), and colonic (sucralose/lactulose ratio) permeability. A first morning urine collection was obtained to measure norepinephrine, cortisol, and creatinine excretion. Results: The mean age of the children was 8.5 ± 0.1 yr. (mean ± SEM). Mean percent recovery for sucrose was 0.023 ± 0.002. The lactulose/mannitol and sucralose/lactulose ratios (per m2) were 0.064 ± 0.004 and 1.134 ± 0.104, respectively. Mean urinary norepinephrine/creatinine ratio was 22.5 ± 0.0 and cortisol/creatinine ratio was 34.3 ± 0.0. A series of regression equations were performed. In the first, the lactulose/mannitol ratio (i.e., small intestinal permeability) was predicted by the Trait anxiety score: F(1, 91) = 5.23; P = 0.025, r = 0.21). When urinary norepinephrine/creatinine ratio and cortisol/creatinine ratio were included in the model the prediction of permeability was stronger: F(3, 73) = 4.73; P = 0.005, r = 0.37. There was no relationship between Trait anxiety score and gastric or colonic permeability or between State anxiety score and GI permeability. Conclusions: There is a positive relationship between the stable tendency to experience anxiety states and small intestinal permeability in children with FAP/IBS. Catecholamine and cortisol excretion magnify the relationship. These data lend support to the hypothesis that stress (anxiety, norepinephrine and cortisol levels) may affect GI permeability. Whether this is related to pre- and/or perinatal stress in children needs to be determined.
AGA Abstracts
S1150 Influence of Parental Height in Disease Characteristics of Pediatric-Onset Inflammatory Bowel Disease Jessica J. Lee, Johanna C. Escher, Melissa J. Shuman, Peter Forbes, Richard J. Grand Background: Impairment of linear growth remains a major challenge to patients with pediatric-onset inflammatory bowel disease (IBD). Despite development of advanced nutritional and medical therapy, the prevalence of growth failure (GF) in IBD has not decreased over the past decade. Although GF in IBD has been attributed mostly to nutritional and disease-related factors, genetics may be the critical predetermining factor. The role of parental height in the growth pattern of pediatric-onset IBD has not been addressed previously. Aim: To identify the impact of parental height and disease-related factors on growth in children with IBD. Methods: 246 probands with IBD diagnosed between the ages of 1-18 (177 Crohn's disease (CD), 62 ulcerative colitis (UC), and 7 indeterminate colitis (IC)) were included. Biological family members (234 mothers, 190 fathers, and 38 siblings) were included to study the family growth phenotypes. Results: GF was defined as height for age Z-score < -1.64 in more than one measurement since diagnosis. Fifty-six patients (22.9%) had GF, and most had CD (87.5% CD vs. 8.9% UC and 3.6% IC). There was a significant left-shift in the mean height Z-score for GF group (Z-score -1.68, p<0.001, whereas, the mean Z-score for non-growth failure (NGF) group was -0.05. GF and NGF groups did not have statistically significant differences in gender, age of onset, disease behavior, disease location, extraintestinal manifestations, history of operations, family history of IBD, or use of corticosteroids. Both mothers and fathers of the GF group showed significantly lower
A-188
last follow-up (OR5.08 [2.73-9.45]) and conversely less proctitis (E1) (1% v 17%, OR0.07 [0.01-0.49]). Unless stated otherwise, all p values <0.00001. Conclusion: Children with IBD have a more extensive disease phenotype at follow-up. The reasons for the stark contrast between children and adults with IBD may reflect differences in a combination of genetic and environmental factors. S1153 An Induction Dose of 12 Mg/D Budesonide Improves Clinical Efficacy in Comparison to a Standard Continuous Dose of 9 mg/d Active Pediatric Crohn's Disease Arie Levine, Michal Kori, Gabriel Dinari, Efrat Broide, Ron Shaoul, Baruch Yerushalmi, Markus Proels, Roland Greinwald AIM: To compare the efficacy and tolerability of two doses of oral budesonide in pediatric patients with active Crohn's disease. METHODS: Pediatric patients with active Crohn's disease (Pediatric Crohn's disease activity index [PCDAI] at baseline: > 12.5 and < 40) were eligible for this double-blind, randomized, multicenter study. Patients received 9 mg/d (n= 35) oral budesonide for 7 weeks or 12 mg/d for 4 weeks followed by 9 mg/d (n=35) for further 3 weeks. Primary endpoint was the mean change of PCDAI from baseline to week 7 between the treatment groups. All patients received 6 mg/d for an additional 3 weeks. RESULTS: 71 patients were randomized. 70 patients were evaluable for the intention-totreat (= ITT)-analysis and 61 patients for the per-protocol (= PP) analysis.Comparative data for the primary and secondary end points in the two groups are presented in Table 1 and Table 2.Serious adverse events occured in 28.6% of the 9 mg/day group and in 17.1% of the 12 mg/day group, however there were no drug related serious adverse events in either group. CONCLUSION: Oral budesonide was effective for inducing remission with both doses. An induction dose of 12 mg/d oral budesonide followed by 9 mg/day appears to improve clinical efficacy in mild to moderately active pediatric Crohn's disease. Table 1.Primary efficacy endpoint: Mean change of PCDAI from baseline to week 7
S1151 A Detailed Investigation Into Epidemiological Risk Factors for Childhood Onset Inflammatory Bowel Disease in Scotland Johan Van Limbergen, Elizabeth Hobbs, Richard K. Russell, Elaine Nimmo, Hazel E. Drummond, Linda Smith, Peter M. Gillett, Paraic McGrogan, Lawrence Weaver, Michael W. Bisset, Gamal Mahdi, Jack Satsangi, David C. Wilson Introduction: The incidence of inflammatory bowel disease (IBD) is increasing among children in Scotland and is now among the highest worldwide. Both environmental and genetic risk factors are implicated in the aetiology of IBD. We aimed to investigate the association of asthma, vaccinations and breastfeeding with paediatric IBD in Scotland. Methods: 439 children diagnosed with IBD <17 y (274 CD, 101 UC, 54 IBDU; median age 11.2 years (Q1-Q3: 8.7-13.1) were compared with more than 25,000 population-matched controls. Children with IBD and their parents were interviewed face-to-face to obtain data on breastfeeding, immunisation history (MMR, diphtheria, tetanus, pertussis, HIB3) and medical history. Control breastfeeding data for children born in 1996 was collected at the health visitor's first review visit (around 10 days old) and were available for Lothian and Greater Glasgow (Information Services Division of NHS Scotland). Immunisation history was recorded at 24 months for children born in 1993 in Lothian, Greater Glasgow and Grampian (idem). Doctor-diagnosed asthma prevalence data in children aged 2-15 from across the whole of Scotland were provided by the Scottish Health Survey 2003. Unifactorial analyses were performed using the χ2-test. Breastfeeding data were stratified for geographical region and affluence (using the Carstairs Deprivation Score - DepCat). Results: History of asthma was associated with IBD and CD (115/439 (26%) and 74/274 (27%) v 521/2965 (18%) in controls: p <0.0001, OR 1.67 (1.32-2.10) and p=0.0002, OR 1.74 (1.31-2.30)). Analysis of controls showed marked differences in breastfeeding rates (>1 week) between Lothian and Greater Glasgow (4368/8196 (53%) v 2270/6775 (33%) p<10-4 OR 2.26 (2.12-2.42)). After stratification for geographical region, there was no significant difference of breastfeeding rates between IBD/CD/UC cases v controls (all p>0.05). Breastfeeding rate in IBD cases (stratified by DepCat (1-2, 3-5, 6-7)) was not different from controls (all p>0.05). Breastfeeding rates in IBD cases were significantly different between DepCat 1-2 and DepCat 6-7 (53/79 (67%) v 17/67 (25%), p<10-4 OR 6.00 (2.91-12.36)). This difference was also present in controls (4226/6648 (64%) v 1768/7750 (23%) p<10-4 OR 5.90 (5.49-6.35)). There was no significant association between immunisation history and IBD/CD/UC (p>0.05). Conclusion: In the high incidence Scottish population, we have shown the novel association of childhood onset IBD and CD with asthma. We did not observe any association between paediatric IBD/CD/UC and either immunisation history or breastfeeding after stratification for geographical region and affluence.
Mean change of PCDAI from baseline to week 7 compared between the treatment groups did not reach significance (ITT analysis: p = 0.2571; PP analysis: p = 0.0895; two-sample t-test). Table 2.Secondary efficacy endpoints (ITT) at week 7:
S1154 A Comprehensive Review of the English Language Evidence Base for the Management of Paediatric Inflammatory Bowel Disease (IBD) Illustrates the Difficulties for Evidence-Based Clinical Guideline Formation David C. Wilson, Working Group Inflammatory Bowel Disease
S1152 Childhood Onset Inflammatory Bowel Disease Has a More Extensive Disease Phenotype in Comparison to Adult Onset IBD Using the Montreal Classification Johan Van Limbergen, Richard K. Russell, Hazel E. Drummond, Nicola Round, Linda Smith, Elaine Nimmo, Ian D. Arnott, Peter M. Gillett, Paraic McGrogan, Lawrence Weaver, Michael W. Bisset, Gamal Mahdi, David C. Wilson, Jack Satsangi
Background and aim: Clinical guidelines, coupled with clinical experience and expertise, play an important role in the delivery of appropriate, effective and efficient health care. We aimed to develop evidence-based clinical guidelines for the medical and nutritional management of paediatric IBD; the first step was comprehensive review of the evidence base. Methods: A search strategy was developed to identify existing guidelines and evidence of efficacy of medical and nutritional treatment of children and adolescents (aged <18 years) with IBD. A computer assisted search of the English language literature for drug and nutritional interventions in this group was performed using the following on-line databases: MEDLINE (1966-2006), EMBASE (1980-2006), CINAHL (1983-2006), Cochrane Library, Web of Science - Science Citation Index (1981-2006), and BIOSIS Previews. Abstracts of major paediatric gastroenterology meetings were searched and recognised experts in the field contacted. The titles and abstracts were examined by 2 reviewers to exclude irrelevant publications. Full manuscripts of the remaining papers were critically appraised by at least 2 reviewers using the SIGN criteria (Br Med J 2001; 323: 334-6). Results: There were 3 evidence-based guidelines for management of IBD identified, none of the highest methodological quality. 5227 publications were identified to end-December 2006, but 4849 were excluded on initial examination of title and abstract. Many manuscripts were excluded as not relevant (mainly because they did not include children or childhood and adult data were inseparable), so 224 publications were critically appraised. 1 systematic review (of nutritional management versus steroids for Crohn's disease) and 13 RCTs were identified, but all had major methodological flaws (Evidence Level 1-); methodological problems abound, such as lack of details on randomisation, lack of intention to treat analysis, and small sample sizes. All other studies were poor quality cohort studies (EL 2-), case series (EL 3) or expert opinion (EL 4). No evidence-based recommendations could therefore be made. Summary and Conclusions: Despite the widespread prevalence of paediatric IBD, comprehensive review reveals a paucity of evidence on the effectiveness of interventions, and a great need for adequately powered, high quality studies. We can provide no evidencebased recommendations for practice, and instead will proceed to formal and methodologically robust consensus-based guidelines, which will be informed by available paediatric and high quality adult evidence, and multidisciplinary experience and expertise.
Introduction: The Montreal classification of IBD was designed to address the problems associated with the Vienna classification in describing childhood onset CD, involvement of the upper GI tract and perianal disease. We compared the phenotypic characteristics of a paediatric IBD cohort with an adult cohort from the same homogenous population at last follow-up. Methods: 1713 IBD patients (416 IBD patients <17 y at diagnosis (276 CD, 99 UC, 41 IBDU) and 1297 adults (>17 y) (596 CD, 701 UC)) were phenotyped. In paediatric IBD, the median duration of follow-up (Q1-Q3) was 3.7 years (1.7 - 6.0) (CD) and 3.5 years (1.1-4.8) (UC). In adult IBD, median duration of follow-up was 10.3 years (3.8-20.6) (CD) and 8.9 years (4.2-16.5) (UC). The Montreal classification was applied to score CD location (L1-ileal, L2-colonic, L3-ileocolonic & L4-upper GI disease); CD behaviour (B1inflammatory, B2-stricturing, B3-penetrating); L4 was added as a modifier to L1-3; the suffix p was added to B1-3 for perianal CD involvement; UC extent was defined as (E3-extensive, E2-left sided, E1-proctitis). 98% of patients were Caucasian. Results: Compared to adult onset IBD, childhood onset IBD was characterised by more familial disease (34% v 20% OR2.1 [1.6-2.8]) and less smokers (1% v 18% OR0.05 [0.02-0.14]). Erythema nodosum was more common in children than in adults (8% v 4% p= 0.007 OR2.3 [1.2-4.1]). Childhood onset CD featured a strong male predominance (59% v 36%, OR2.6 [1.9-3.4]). At last follow-up, disease location was more extensive in children with more (a) L2+L4 and L3+L4 (16% v 2% OR11.6 [5.4-25.2] and 43% v 3% OR23.3 [13.4-40.6]) and (b) less L1 and L2 (3% v 32% OR0.06 [0.03 -0.1] and 15% v 36% OR0.3 [0.2 -0.5] respectively). CD behaviour at follow-up showed no differences when analysed at the same time interval post diagnosis. 53/196 (27%) of childhood CD patients already had L3+L4 disease at diagnosis. Within the first two years after diagnosis, 56/196 (29%) progressed to more extensive disease. These changes were not due to the inclusion of findings on upper GI endoscopy in 41/56 (73%) but due to progression from local oral/perianal, L1, L2 or L4 disease to L3 disease. Childhood onset UC was characterised by extensive colitis (E3) in 82% of children v 48% of adults at
A-189
AGA Abstracts
AGA Abstracts
mean height Z-scores than their NGF counterparts (-0.63 vs. 0.08 for mothers; -0.36 vs. 0.28 for fathers; p<0.001). GF patients had significantly lower mean target heights based on measured parental heights (172.5 cm for males and 160 cm for females) compared to NGF patients (177.4 cm for males and 165.2 cm for females). Of 107 patients who had reached ≥ 18 years of age after study enrollment, mean final height Z-score was -1.29 for the GF group as compared to 0.06 in the NGF group. For both growth groups, the ratio of measured final height to target height was close to 1 (0.97 for GF and 0.99 for NGF). Of 31 GF patients who reached adult heights, only 9 (29%) remained in the GF group whereas 22 (71%) did not. Conclusion: Parental heights were significantly lower in GF patients with pediatric-onset IBD, and there was no significant difference between target height and patients' observed final height. Most GF patients reached their target height and were no longer growth impaired when final height was reached. Hence, genetics may be the key determinant of growth in IBD.