significantly < 0 m the 4 age groups at diagnosis (G1 -0.66, G2 -1.05, G3 -1.18, G4 -117) and oornlaliaed after one year in the 2 younger groups and at 2 years in the older goups. Height gains (delta z-scores) were present at one year ( group A: 0.81 vs group B: 030 ) and at 3 years (group A: 141 vs. group B: 0.51 ) Height gain was > 0 at 2 and 3 years even in children who wmv above average at the time of diagnosis (group B) In the population studied, 83% of patients with CD had a normal height for age (z > -2.0) at diagnosis. CONCLUSION: Even children who were of average height at diagnosis had catch- up growth and maintained it after three ),ears of a GFD, indicating prior suboptimal growth .Weight recovery was parallel to that of height. The overweight observed in group 2 during the third year roight be explained by the family attitude towards these children, aiming at a laster nutritional recovery"

T1354

The Efflca W of Ondansetron in Attenuating Symptoms of Cyclic Vomiting Bhanu K. Sunku, B. U. la Terming Maa, John R. Hayes

Background: Cyclic vomiting syndrome (CVS) is characterized by a clinical pattern of recurrent episodes of severe vomiting between which the child is normal. Because the pathophysinlo D, is unknovm, empiric therapy has been directed towards either preventing eDsodes , or aborting them in patients with infrequent episodes or attacks refractor}" to prophylaxis. In this study, we prospectively examined the efficacy of oral and IV ondansetmn used during acute episodes of CVS. Methods: A total of 25 patients meeting the consensus criteria for CVS were studied prospectively using a questionnaire: 17 patients were prescribed oral ondansetron and 14 patients received IV ondansetron dosed at 0.3 mg/kg, both adminis- tered at the onset of the episode Six patients were evaluated for both the IV and PO forms in separate episodes. Parents filled out a questinnnaim to e'<aluate the child's symptoms of pallor, tiredness, nausea, vomiting, abdominal pain and headache on a graded scale of 0 to 4 before arid after receiving medication at the point of improvement in symptoms or at two hours if no response was noted sootier' Side effects of dry' mouth, headache, abdominal pain and drowsiness were also recorded. Results: Of the 25 patients evaluated, 17 were female, the mean age was 94 yrs and the average duration of episodes was 69.8 hrs. All patients treated with ondansetmn had a significant reduction (p<.001) in all symptoms. The greatest difference in pre- and post-treatment scores was found in symptoms of nausea and vomiting for both the oral and IV groups. With the exception of fatigue and headache, more than half of the subjects achieved 50% reduction in sy'mptom scores. The median time of onset of the therapeutic effect was 60 minutes for nausea and vomiting. Nine out of 25 patients had tolerable side eftects including drowsiness (6), dry mouth (3), headache (2) and abdominal pain (2). Conclusions: In this study" of therapeutic response, we were able to document a significant improvemem from use of ondans-etron in acute episodes of CvS. The greatest improvement was seen in symptoms of nausea and vomiting. Approxi- mately one-third of patients had mild side effects. Ahhough the exact mechanism of action in CVS is unknown, it is likely that antagonism of 5HT~ receptors in the chemotrigger zone may reduce nausea and emesis

T1355

Comparison of Fecal L/pose Test and Fecal Elastase-1 Test in the Assessment of Exocrine Pancreatic Function in Cystic Fibrosis jaroslasv Walkowiak, Karl-Heinz Heraig, Kryst?/ma Strzykala, Julinsz Prayslawski, Marian Graymis/awski, Marian Krawczy'nski

Ot~ective: In pediatric patients, indirect tests am preferred because of their less invasive character. Among those, fecal elastase-1 test has so far been shown been shown to have the highest senmtivity and specificity. However, the role of the ti:cal lipase test in the diagnostic work up for pancreatic insufficiency in cystic hbmsis patients has not been defined. Therefbre, the aim of the present study was to compare the sensitivity and the specificity of fecal lipase (lP) test to the fecal e]astase-1 (El) test in the assessment of exocrine pancreatic function. Material gr methods: 63 CF patients and 95 HS on normal fat intake were evaluated. In all sultjects, E1 concentration (H.1SA) and LP activity (ELISA) ',',,ere measured. In 50 HS, ~mple-to-sample (n = 3) variation from the same stool and day-to-day variation from three consecutive stools were determined twice. The presence of pancreatic insufficiency patients was documented in 55 pancreatic it~nfficient CF patients by tire determination of fecal fat excretion and in 12 pancreatic sufficient patients by' the (hrect test The sensitivity" and specificity ot the E1 test and LP test were compared. Results: The sample-to-sample variation (mean + SEM 13 2 + 1.2 % vs. 23.4 + 2.2%) and day-to-day variation (mean + SEM: ]6.3 + 1.2% vs 325 + 26%) were significantly lower (p<0.0001) for E1 determinations tMn for LP measuremenl:s With the cut-off levds giving tire same specificity ['or both tests (95 8%), the sensitivity ot fecal elasrase-1 test (87.3%) was significantly higher (p<0.04) than Ihat of fecal lipase test (778%). In conclusion, tecal lipase test is less sensitive than fecal elastase-i test w'ith a comparable specificity in the assessment of CF pancreatic involvement.

T1356

Autuimmune Hepatitis in Children: Predictors for Orthotopic Liver Transplant Evaluation Mahmoud Sabri, John Peters, Maria Clavell, Seema Khan

Background: Autoimmune hepatitis (AIH) is a chronic liver disease that may, progress to cirrhosis and death if left untreated. Data at initial presentation on predictors of outcome in children with AIH are scarce. Aim: To assess dinicohistologic parameters as predictors oi medical treatment outcome in children with AIH. Methods: We re,hewed retrospectively climcal data related to 15 children (14 temales), median age 11 y (range, 2-16) with A1H followed at our center since 1989. Clinical presentation, laboratory tests, and liver histology were compared between two groups: gp A with patients either listed or post orthotopic liver transplant (OLT) (n = 5), and gp B with patients in clinical remission with medical manage- ment (n= 10). We defined clinical remission as hepatic transaminases less than 2-fold normal; liver histology, was staged according to prev/onaly published criteria ~. Results:

Initially, mean total bilirnbin, serum albumin, and ALT were 66 mg/dL, 2.9 mg/dl and 450 IU/L in gp A vs. 4 mg/dL, 3.8 mg/dL and 878 IU/L m gp B respectively (p = NS). The mean prothmmbin time (PT) of 17 seconds in gp A patients was prolonged cmnpared to 13.3 sec in gp B (p =0.05). All patients in gp A were diagnosed with type I AIH; in gp B 6 had type 1 AIH, 1 patient had type 2 AIH, while 3 were only ANA positive (p=NS) Liver histology at presentation revealed at least stage 2 cirrhosis in all (100%) in gp A cmnpared to 5 (50%) in gp B. Mean steroid dose used for induction of remission was 09 m gp A in contrast to 1.3 mg/kg/d in gp B. In the tbrmer gp none of the patients attained remission, 2 are post OLT, whereas 2 of the 3 0 L T listed patients are now on FK506 In gp B, all patients are on azathioprine/6-mercaptopurine in a state of remission, all but 2 are off steroids. There was no correlation between albumin, PT, and the type of AIH with cirrhotic and non-cirrhotic patients in this series. Conclusion: In children with AIH, prolonged Vr may predict poor response to medical management and may indicate the need tbe an early evaluation for OLT. We did not flnd a correlation of initial hepatic syntbetic fnnctinns with cirrhosis, but they may be a predictor of poor response in a larger cohort of patients ~Batts et al. Chronic hepatitis. An update on terminology and reporting. Am J Surg Pathol 1995;19:1409-17

Clinical ~ n and follow.up.

~pA GpB n (f=md.) 5 (5) lo (9) Age (yr), median (range) 12 (8-15) 10 (2-16) Symptom duration (wk), mean 5 13.5 Follow-up (yr), median (range) 2 (1.5-7) 2.5 (0.5-13) Jaundice (%) 90 70 Abdominal pain (%) 40 70 FalJgue (%) 0 30 oe .~ (~ ) 4o 30

T1357

CDP571, a Humanized Monoclonal Antibody to TNFmt, a Well Tolerated Alternative in Crohn's Disease Patients who have Experienced Hypersensitivity Reactions to Infliximab Stephen Hanauer, Daniel Present, Stephan R, Targan, kori Ram, Jatin Patel, wilham J. Sandbom

Background: CDP571 is a novel humanized monoclonal antibody to tumor necrosis factor (TNF)-cc The aim of the study was to assess the tolerabifity of a single infusion of CDP571 in Crohn's disease patients with a histor}, of hypersensitivity' to infliximab. Hypersensitivity to inlliximab has been reported to correlate with the presence of human anti-chimeric antibodies (HACA), which may also decrease the plasma concentrations of the drag. Methods: 22 adult patients with Crohn's disease who had experienced either acute hypersensi- tivity reactions (AHR) ( n = l l ) or delayed hypersensitivity reactions (DHR) (n= 11) to infliximab were enrolled m this multicenter, open-label study Blood samples were taken at screening to confirm non-cross reactivity of plasma with CDP571 and for baseline analysis of HACA. Patients received a single intravenous dose of CDP57I (10 mg/kg) over 2 hours and were subsequently assessed for 4 weeks by standard clinical and laboratory analyses. Plasma samples were analyzed for CDP571 plasma concentrations and antibodies to CDP571 Crohn's disease was monitored using the CDAL Results: All patients had previously received ->2 doses of infliximab and 17/22 were found to have HACA. No cross-reactivity to CDP571 was found 21/22 patients completed the study One patient was withdrawn 13 days after the infusion due to disease progression. One patient experienced an acute reaction 1.7 hours into the infusion of CDP571, devdoping cough/dyspnea, which was treated *,vith diphenylhydramme. One patient experienced a possible DHR 16-20 days post-infusion, developing urticaria, headache and artbralgias similar to previous reaction experienced with inflLximab, but less severe. No antibodies to CDP571 were detected in an}, of the 22 patients and the plasma concentrations of CDP571 were consistent with those observed in previous studies. During the study, 11 patients experienced 44 adverse events (AEs) - - mostly mild or moderate. 3 serious AEs were reported, none of which were considered related to the stud?/" medication. Clinical response (decrease in CDAi ->70 points) was observed in 9/22 (41%) patients by" Week 4, with no differences between the AHR and DHR groups. 5 of the 22 patients (23%) showed a decrease in CDAI >100 points. Conclusion: CDP571 is well tolerated and may provide an ahernative option in patients who have previously experienced hypersensitivity to inlhximab.

T1358

A Prospective Cohort Study of the Influence of Nonsteroidal Anti-inflammatory Drugs on Disease Activity in Outpatients with Inflammatory Bowel Disease (IBD) Gregory F. Bonner, Alifiya Fakhri, Sridhar R. Vennamaneni

Aims: We examined whether use of nonsteroidal anti-inflammatory drags (NSAIDs) in oupatients with IBD was associated with increased severity of disease activity. Methods: 426 outpatients with Crohn's disease (CD) and 203 with ulcerative colitis (UC) were followed from November 1997 to April 2002. Patients were questioned at each ,,/sit regarding any use of prescription or OTC NSAIDs and a clinical disease activity index (modified Harvey Bradshaw (MHB) or Lichtiger score (LS)) was obtained A moderately active flare was defined as an MHB score of 7 or LS of 8. Patients with an ostomy or J-pouch surgery were excluded. NSAID use was dwided into low-dose (ASA 325 mg/d or less or equivalent doses of other NSAIDs) or high-dose, hrdividual patients could have multiple visits and each visit was counted separately. Statistical analysis was performed using the GEE regression model. Results: CD patients had 1315 visits with no NSAIDs, 215 visits with tow-dose NSAIDs and 139 visits with high-dose NSAIDs. UC patients had 495 visits w'ith no NSA1Ds, 112 low- dose NSAIDs, and 49 high-dose NSAIDs. For CD, the average MHB score was 4.07 for the no-NSAID group, 4.24 for tow-dose NSAIDs (p=0.46) and 4.78 for high-dose NSAIDs

A-517 AGA Abstracts

(p=0.0072 vs no-NSAIDs). For UC the corresponding scores were 5.64, 5.46, and 6.20 respectively (p=NS). The probabdity of moderately acnve disease did not demonstrate significant dilfi:mnce with NSA1D use, though CD patients on high-dose NSA1Ds had a 1,27dbld increase likelihood of moderately active disease as compared with patients on no- NSAIDs. Subgroup analysis showed die increase in disease activity among CD patients taking high-dose NSAIDs was limited to patients with colonic niw~lvement. For CD patients with limited ileal involvement the MHB scorn was 4.01 with no-NSAIDs and 3.85 on high-dose NSAIDs. CD patients with ileocolonic or isolated colonic disease had an MHB score of 4.11 with no- NSAIDs but increased to 5 47 with high-dose NSAIDs (p = 0.008). Among the UC patients, there was a trend towards a higher disease activity score among patients with pancofitiS(as opposed to distal disease). The LS increa~d from 5.79 to 7.43 comparing the pancofitis no-NSAIDs vs. high-dose NSAIDs (p = 0.21). Condusions: Use of low-dose NS?dDs was not a ~ i a t e d with an increase in disease activity for these outpatients witfi either CD or UC, Use of high-dose NSA1Ds was associated with a higher level of disease activity among CD patients with colonic involvement.

T1359

Initial Response to Treatment (Rx) and Need for Intensified (Step-Up) Therapy in Children with Newly Diagnosed Inflammatory Bowel Disease (IBD): A Preliminary Investigation by the Pediatric IBD Collaborative Research Group James Markowitz, Jeffrey Hyams, Anne Griffiths, Athos Bousvaros, J. Fernando Del Rosario, Jonathan Ewans, Richard Grand, Aubrey Kate, Subra Kugathasan, David black, Adam Mezott, Maria Oliva-Hemker, Anthony Otley, Marian Pfeftkrkorn, Joel Rosh, Robert Rothbaum, Vasnndhara Tofia, Don W Cben

The response of children with IBD to current Rx in dinical practice settings has received little systelratic attention. We therefore enrolled newly diagnosed children with IBD from 16 US & Canadian pediatric GI centers, identified between Jan-Nov 2002, in a prospective, observational registry" designed to assess the charactenstics of the North American pediatric IBD population, and their outcomes to current treatments. Methods: All children were managed according to the dictates of their treating physicians, not by standard ILx protocols. Baseline & 30 day disease characteristics and Rx were recorded prospectively. Responses to ILx aiter 30 days were assessed in regard to need tbr step-up Rx after initial Rx was established. Subjects were considered to have required step-up if the Physician Global Assessment (PGA) had worsened between baseline and 30 days, if steroids or infliximah were prescribed bdbre 30 days if these agents were not started as initial P,x, or if the initial dosage of steroid was increased prior to 30 days after the start of ILx. Results: 113 newly diagnosed children with IBD (79 Crohn d ~ a s e (CD), 25 ulcerative colitis (UC), 9 indetermi- rate colitis (ICY), mean age 11.3 yrs (range 1.4-15.9 yrs), were evaluated. Initial disease activity was characterized by PGA as mild (16 CD, 8 UC, 4 ICY or moderate-severe (63 CD, 17 UC, 5 ICY. 28 children were initially treated with steroids alone, 38 with steroids + additional agents, and 47 with no initial sysmmic steroid. Aiter 30 days of 1Lx, only 3/25 (12%) UC and 1~ (11%) IC subjects required stepmp, irrespective of niitial ~x. However, 2/16 (125%) subjects with mild CD and 13/63 (20.6%) of those with moderate-severe CD required step-up. Among those with moderate-severe CD, 30% of children not innially receiving steruids required step-up, compared to only 16.3% of those receiving steroids initially" (p = 036). Among the 3 children with moderate-severn CD who initially were treated with budesonide, 2 required step-up by" 30 day, s. Conclusions: Children with newly diagnosed [BD respond well to initial Rx and only rarely require step-up within the first 30 days of Rx. Those with moderate-severe CD at diagnosis are the most likely to require steDup, nspecially if their initial Rx does not include prednisone. Among all children with CD initially" receiving pradniso~m (+- other agents), 87% can be expected to improve within the' first 30 days of Rx

T1360

Infliximab Is Not a Bridge to Immunomodulators In Crohns Disease Jeltrey A. Tuvlin, Kevin G. Schaeter, Nitin P ~me, Russell D Cohen, Stephen B. Hanauer

Background: infliximab (IFX) induces and maintains remission m Crohn's disease (CD). The likelihood of long-term improvement without re-infusion after 1 to 3-induction inthaions, with or without immunomodulators, has not been determined. Methods: CD pts treated at the University ot Chicago IBD center with on-demand IFX induction regimens ( 1-3 infusions within 6 months) during the 3 years alter commercial release were prospectively analyzed tot initial efficacy duration oI response, and relapse within 12 months of the last infusion. lmmunomodulator therapy at baseline or subsequently added was determined and clinical improvement was compared to baseline according to a 3-point scale C0 = <50%, 1 = 50- 99%, 2 = 100%). Cfinieal relapse was defined as a decrease in clinical improvement score, need tigr repeat IFX infusion, increase in corticosteroid dose, or CD surgery'. Results: 268 (89%) ot 301 CD pts treated with IFX had adequate follow up. 57/268 (55% female) received only 1 (n= 23), 2 (n = 21), or 3 (n= 13) in|haions within 12 months. 39/57 pts initially responded to IFX. 34 (87%) of these 39 pts relapsed: by improvement scale (32), surgery (1), or steroids (1). Median time to relapse was 12 wks (range 3-57 wks). Only 5/268 (1.9%) pts received 1, 2, or 3 infusions and remained at their post-infusion level of improvement without relapse at 12 months. Baseline therapy, with immunomodulators at the time of initial infusmn did rmt prevent relapse (Table). Conclttsion~s: There is no evidence that infliximab- induced improvement of CD is a bridge to long-term immunomodulator therapy- without repeated on demand or maintenance infusions. Future studies should consider randomizing responders off immunomodulators

Irnmunomodulators In Pts Receiving 1-3 induction Infusions of Inmximab

Pts Re- # Well at 12 lapsed Bridged Pts

Total Pts Mo.s w/o (n=34) After Pts Bridged" to Immu- W ~ at 12 Me~s wto n=57 Relapse or Initial nomodulatots, n=5 Reinfusion Response Relapse or

Re~nfus~n (.=3~) Baseline Immune- 22 12/14 modulators (38.8%) 2/22 (9.1%)1' (85~7%):~ 5 115 No Baseline Immu- 35 nomodulatars (61,4%) 3/35 (8.6%) 22/25 (88%) NIA N/A ~dged" is defined by beginning immunomodalators within three rnonth~ of in~a~ i~ infusion or aftetw~ 1'p=ns (0.71), FisheCs Exact Test ~..p= ns (0,77), Chi ~ua~e wit~ Yates correction

T1361

Evaluations of Clinical Efficacy and Steroid Sparing Effect of Granulocyte and Monocyte Adsorptive Apheresis in Patients with Corticosteroid Dependent Ulcerative Colitis Hiroyuki H.anai, Ken Takeuehi, Takayuki Iida, Kotaro Tozawa, Isao Matsushita, Isao Matsushi-ta, Tatsum Tanaka, Tatsuro Tanaka, Yoshihiko Sato, Kazuto Kikuti, Toshio Nakamura, Yasuhiko Maruyama, Abby Saniabadi

Corticosteroids are given to induce remission in most patients with active ulcerative colitis (UC), but frequent relapse is common in patients who initially, respond. Further, prolonged steroid therapy has been considered unwarranted and dangerous. Therefore, there is a long standing need for steroid free treatment of UC. However, increased granulocyte and mono- cyte/macrophage levels and their activation and prolonged survival time is a feature of active UC. Likewise, fecal calprotectin (a neutrophil protein) level increases during nitestinat inflammation and c~'m predict UC relapse. We thought that granulocyte and monocyte adsorptive apheresis (GMA) to reduce the circulating level of these leukocytes might reduce inflanunation and prevent relapse during steroid tapering in steroid dependent patients. Forty six patients, mean age 38+-14 yr, CAI (clinical activity index) 9.2, DAI (disease activity index) 8.6 were given 10 GMA sessious each, one session/week for 10 consecutive weeks, by using Adacolumn that contains cellulose acetate bgads of 2 mm in diameter as the column adsorptive carriers for granulocytes and monocytes/macrophages. Duration of one G~vk& sessions was 60 minutes, at 30mUminute. At week 12, CAI was 1.7, DAI 2.8 (n=46) and 39 of 46 patients were in remission. The mean dose of prednisolone, 18 mg/ patient/day at entry was reduced to l lmg at week 12 and to 4rag at week 20. Further, at week 20, 42 patients were in remission, 2 had improved and 2 had relapsed. The treatment was well tolerated and no serious side eft~cts were observed. In conclusion, GMA appeared to be an eftective adjunct to standard drug therapy of active UC by promoting remission and suppressing relapse during steroid tapering. It seems that reduction of activated granulo- cytes and monocytes alleviates inflammation and this might promote remission. However, the full efficacy of GMA is unlikely to be due to reduction of circulating levels of granulocytes and monocytes per se. Analysis of blood returning to the patients from the column outflow revealed elevated plasma level of IL-1 receptor antagonist which also should exert anti- inflammatory effect

T1362

Oral Immune Regulation Using Colitis Extracted Proteins - A new Mode of Treatment for Crohn's Disease: Results of a Phase l Clinical Trial Eran Israeli, Eran Goldin, Nilla Hemed, Barbara Thalenfeld, Dean Engelhardt, Elazar Rabbani, Yaron llan

Background: Oral immune-regulation is a new method for modulation of the pro- (Thl) / anti-inflammatory" (Th2) paradigm, and was recently' shown effective in amelioration of experimental colitis Currently available immunomodulatory treatments for Crohn's disease (CD) are not antigen specific. Aims: To evaluate the safety and possible efficacy of induction of oral immune regulation using colitis extracted proteins (CEP) in subiects with moderate to severe Crolm's disease. Methods: Ten CD subjects with CDAI score of 200 to 350 were treated orally three times a week for 16 weeks, with autologous colonic-mucosal protein- extract obtained by colonoseopy. Subjects were monitored for CDAI, Inflammatory-bowd- disease-questionnaire (1BDQ), and routine blood tests. The immune modulatory, effect of this treatment regimen was assessed by FACS analysis of T-lymphocyte subpopulations, CEP-speciflc IFNg ELISPOT assay, and serum cytokine levels. Results: Induction of oral immune regulation towards CEP appeared to significantly ameliorate disease activity. Dumg the course of treatment period, 10/10 subjects had clinical response (CDAIjavaseript:ap- pend char('(')70), and 7/10 subjects achieved clinical remission (CDAIjavascript:ap- pend char('(') l SO). Median nine to respot~se was five weeks. A signifieant increase in mean IBDQ score was noted on week 16 as compared to baseline (134javascr ipt :ap- pend char('_+')9vs.164javascript:append char('_+')2, p<0.05). No treatment related adverse events were noted in any of the subjects, High levels of CEP specific IFNg spot forming colonies were detected in 5 subjects prior to treatment and in all five, a marked decrease was observed The CD4+/CD8+ lymphocyte ratio, and peripheral NKT cell numbers, increased significantly, in 7/10 and in 5/10 subjects, respectively. A significant increase m the CD3 +/RA, CD19 +/RA, and CD25 +/45RO lylnphocytes was noted in the majority of treated subjects. No major changes occurred in CD3 +/45RO, CD4+/45RO, CD8 +/45RO, CD44 +/45RO, and CD 16 +/56 + RA lymphocytes subpopulations. A signifi- cant increase in ILl0, and IL4 serum cytokine levels was observed in 7/10 subjects during treatment period. Conclusions: lmmuna-regulation via oral administration of proteins extracted from colonic mucosa is a safe and possibly effective treatment for subjects with moderate to severe CD. This mode of treatment may provide a means of antigen specific immune modulation in the bowel ['or CD patients.

AGA Abstracts A-518