2014 IBD Abstracts
(Non-Hispanic Black 39 [58%]; Non-Hispanic White 63 [22%]) insurance payor status also differed among groups (P < 0.05). There were no statistically signi cant
differences between races or insurance payor status (Medicaid versus private) with respect to medications, radiologic imaging, or laboratory testing completed in the ED (Table 2). Nearly all patients in each race group were discharged from the ED. CONCLUSIONS: Race and insurance payor status does not appear to impact the evaluation and treatment of children and adolescents with Crohn's disease in the ED setting.
Patients With Ulcerative Colitis Do Not Have a Decreased Risk of Diverticulosis
Kinnucan Jami1, Tomal Justin1, Rubin David2
1In ammatory Bowel Disease Center, University of Chicago, Chicago, Illinois, 2The
University of Chicago Medicine, In ammatory Bowel Disease Center, Chicago, Illinois
BACKGROUND: Diverticulosis is common in patients after age 60, with a prevalence of up to 40%. Studies outside the United States (U.S.) have described a lower prevalence of diverticular disease in patients with in ammatory bowel disease (IBD) than in non-IBD. The aim of this study was to assess the prevalence and risk factors for diverticulosis in ulcerative colitis (UC) in a US tertiary center.
METHODS: We performed a retrospective review in patients over 50 years of age who underwent screening or surveillance colonoscopy from January 2006 to December 2013. Pathology was reviewed for all patients to eliminate cases of segmental colitis associated with diverticula (SCAD). We rst assessed the prevalence of diverticulosis in UC patients compared to patients without IBD. Then we performed a nested case-control study comparing UC patients with diverticulosis (cases) to UC patients without diverticulosis (controls) to identify clinical predictors (disease extent or disease duration) of diverticulosis. Cases were matched 1:1 fashion based on age and gender. Cases were excluded if there was lack of data or a matched control. Statistical analysis included a Student's t-test, Chi-squared test, relative risk (RR) and adjusted RR using logistic regression.
RESULTS: We identi ed 573 UC patients and 16,695 patients without IBD who underwent colonoscopy. 25.1% of UC patients had diverticulosis, and 46.7% of non- IBD patients had diverticulosis, P < 0.01. UC patients were younger (61.1 years versus
62.8 years, P < 0.01) and less likely to be female (46.3% versus 59.4%, P < 0.01). On univariate analysis the RR of diverticulosis 0.56 (95% CI, 0.49-0.64). On multivariate
analysis adjusting for age and gender, RR of diverticulosis 1.03 (95% CI, 1.02-1.03).
In the nested case-control study, 118 patients with UC with diverticulosis (cases)
were matched to 118 patients with UC without diverticulosis (controls). The mean age of cases was 65.1 years (SD 8.9 year), and 42.3% were female, this was similar in the control group. The majority of diverticula were limited to the sigmoid/descend- ing colon (68.6%). There were no differences between cases and controls in disease
duration (16.7 year versus 18.7 year, P 0.27) or extent of disease (pancolitis 50% versus 60.1%, P 0.24).
CONCLUSIONS: In this US-based study, UC patients had similar rates of diverticulosis
as non-IBD patients undergoing screening or surveillance colonoscopy. There were no identi ed predictors for the development of diverticulosis in patients with UC. These results are different than previously reported data in IBD patients in other countries and may represent unique dietary factors that contribute to the development of diverticulosis.
Intensive Granulocyte-monocyte Adsorptive Apheresis Twice a Week for the Treatment of Steroid-na ve Ulcerative Colitis
Miyakawa Maki, Tanaka Hiroki, Sakemi Ryosuke, Nasuno Masanao, Motoya Satoshi, Imamura Akimichi
IBD Center, Sapporo Kosei General Hospital, Sapporo, Japan
BACKGROUND: Intensive granulocyte-monocyte adsorptive apheresis (GMA) admin- istered twice a week for the treatment of ulcerative colitis (UC) achieves remission more effectively and faster than weekly administration. However, there are relatively few reports regarding intensive GMA for the treatment of steroid-na ve UC patients. The objectives of this study were to investigate the ef cacy of GMA for Japanese steroid-na ve UC patients, and to identify clinical predictors of remission in these patients after GMA treatment.
METHODS: We retrospectively analyzed the data of active UC patients who received intensive GMA treatment at the IBD Center, Sapporo Kosei General Hospital from April 2010 to April 2014. We excluded patients with a Lichtiger clinical activity index (CAI) score of #4 and those who received additional medical treatments within 1 week before or after GMA treatment. GMA was performed using the Adacolumn (JIMRO, Takasaki, Japan). Each patient received GMA twice a week for 5 weeks with a maximum of 10 treatments. The ef cacy of GMA treatment was evaluated based on the decrease in the Lichtiger CAI. Clinical remission was de ned as a CAI score of
#4 within 10 GMA treatments. We investigated the remission rate and time required for remission in UC patients who received GMA treatment classi ed by steroid response (steroid-na ve, steroid-dependent, and steroid-resistant).
RESULTS: For the 78 patients (25 females, 53 males; mean age, 42.5 years) analyzed, the mean duration of disease was 5.7 years before the intensive GMA treatment and the mean CAI score was 9.6 at baseline. We classi ed 44, 29, and 5 patients as pan- colitis type, left-sided colitis type, and proctitis type UC, respectively. In terms of steroid response, 32, 30, and 16 patients were classi ed as steroid-na ve, steroid- dependent, and steroid-resistant, respectively. Of all patients, clinical remission was achieved in 42 (54%) with a mean time to remission of 15.0 days. Clinical remission of steroid-na ve, steroid-dependent, and steroid-resistant patients was achieved in 18 (56%) with a mean time of 16.2 days, 16 (53%) with a mean time of 14.8 days, and 8 (50%) with a mean time of 12.8 days. The remission rates and times to remission were not statistically signi cant between the steroid-naive, steroid-dependent, and steroid-resistant groups. Of the steroid-na ve patients, age was a statistically signif- icant predictor of remission (the mean ages of the remission and non-remission
groups were 38.4 years and 51.8 years respectively, P 0.030). In particular, of the 20 steroid-na ve patients aged #50 years, clinical remission was effectively achieved
in 15 (75%).
CONCLUSIONS: In UC patients who underwent intensive GMA treatment, the remission rates and times to remission were not statistically signi cant among the different steroid responses. Therefore, intensive GMA treatment was found to be as effective for steroid-naive patients as it is for steroid-dependent and steroid-resistant patients. However, in the case of steroid-naive patients, a younger age might be suggested as a possible predictor of remission.
Characteristics and Behavior of Elderly-Onset In ammatory Bowel Disease: A Multi-Center U.S. Study
Hou Jason1, Feagins Linda2, Waljee Akbar3
1Baylor College of Medicine, Houston, Texas, 2Dallas VA Medical Center, Dallas, Texas, 3University of Michigan, Ann Arbor, Michigan
BACKGROUND: Several European studies have highlighted potential differences in in ammatory bowel disease (IBD) distribution and disease course between adult- onset and elderly-onset IBD. However, there is a paucity of studies comparing adult- and elderly- onset Crohn's disease (CD) and ulcerative colitis (UC) populations in the
U.S. The aim of this study was to compare disease characteristics and behavior of adult- and elderly- onset IBD in a multi-center U.S. study.
METHODS: We reviewed the medical records of patients with IBD seen at 3 VA medical centers between 1999 and 2009 (Houston VA and Ann Arbor VA) and 1999 to 2014 (Dallas VA). Chart review was performed to identify and con rm patients with IBD and extract data regarding IBD type, behavior, location, age at, sex, and race. IBD characteristics and behavior were classi ed according to the Montreal Classi cation. Patients were classi ed based on age at IBD diagnosis as adult onset (18-64 years of age) or elderly onset ( 65 years of age). Statistical analyses were performed using Chi square analyses and Fisher's exact test for discrete variables and t-tests for continuous variables. Univariate and multivariate logistic regression adjusting for race, gender, and IBD type were performed.
RESULTS: A total of 1,665, patients were con rmed to have IBD based on chart review and included in the analyses, including 724 CD, 876 UC and 65 IBD unclassi ed patients. The cohort was 92% male, and 83% of those with identi able race were non- Hispanic Caucasian. Males were diagnosed at a later age then females, 46.89 (SD 0.45) versus 35.62 (SD 1.09) years of age, P < 0.001. 272 patients were 65 years of age at IBD diagnosis, and 1,393 patients were < 65 years of age at IBD diagnosis. After adjust- ing for IBD type and gender, Caucasians were more likely than non-Caucasians to have elderly- onset IBD (aOR 2.26, 95% CI, 1.36-3.76). Patients with UC were more likely than CD patients to have elderly- onset IBD (aOR 1.50, 95% CI, 1.11-2.03). Compared to
patients with adult- onset CD, patients with elderly onset CD were more likely to have isolated colonic disease (45% versus 32%, P 0.048) and non-stricturing, non-pene-
2014 IBD Abstracts
trating phenotype (71% versus 53%, P 0.031), but less likely to have perianal com- plications (7% versus 19%, P 0.018). No signi cant difference in the distribution of UC extent by age of diagnosis was observed.
CONCLUSIONS: There are several signi cant differences in the disease characteristics between adult- and elderly- onset IBD, including race, IBD type, and Crohn's disease distribution and behavior. Our observations of differences in CD distribution and behavior based on age of onset among a U.S. cohort are similar to those reported in European populations. We also observed that patients with UC were more likely to have elderly-onset disease compared to patients with CD, although UC extent among elderly-onset UC patients was not signi cantly different from adult-onset UC.
Clinical and Serological Factors Associated With Response to In iximab in Chronic Refractory Pouchitis and Pouch Complications
Kelly Orlaith1, Rosenberg Morgan2, Stempak Joanne3, Tyler Andrea4, Silverberg Mark2
1Zane Cohen Center for In ammatory Bowel Disease, Mount Sinai Hospital, Toronto, Canada, 2Mt Sinai Hospital, Toronto, Canada, 3ZCC for IBD, Mount Sinai Hospital, Toronto, Canada, 4ZCC for IBD, Mt Sinai Hospital, Toronto, Canada
BACKGROUND: Pouchitis refractory to rst-line therapies and pouch complications remains problematic following colectomy and ileal-pouch anal anastomosis (IPAA) for ulcerative colitis (UC). Evidence for in iximab (IFX) use in this setting and factors predicting its success remain limited. Here, we investigated IFX use in refractory pouchitis and clinical and serological variables associated with treatment response. METHODS: Patients were identi ed from the Mount Sinai Hospital IBD and Pelvic Pouch Databases, Toronto, Canada. Clinical, endoscopic and serological data were reviewed from 579 individuals who underwent colectomy and IPAA from 2000 to 2014.
Patients with chronic refractory pouchitis and Crohn's Disease- like outcome treated with IFX were included. Pretreatment parameters were measured within 4 weeks of induction and IFX response at a median of 9 (initial) and 48 weeks (sustained) respec- tively. Complete response was de ned as symptomatic and endoscopic resolution with modi ed Pouchitis Disease Activity Index (mPDAI) score < 5. Partial response was de ned as improvement in symptoms with reduction in mPDAI .2. Endoscopic muco- sal healing, pouch excision, pouch complications and CRP were recorded. Serum was analyzed for antibodies against Saccharomyces cerevisiae (ASCA), OmpC, CBir1 and perinuclear neutrophil cytoplasm (pANCA). Fisher's exact testing and Kruskal-Wallis detected differences in clinical and serological variables between subgroups. Logistic regression models were applied to estimate odds ratio (OR) and 95% con dence interval (CI) of clinical and serological factors with initial and sustained clinical response as dependent variables. Results were deemed signi cant if P-value was #0.05.
RESULTS: Thirty-one patients were included (33% male; age 32.6 6 2.6 [mean 6 SE]). 26% were on combined therapy with immune-modulator. Seventy-four present achieved post-induction response at median 9 weeks (IQR [6.5-13]) (48% complete, with concomitant mucosal healing). At median 48 weeks (IQR [25-71.5]), 62.6% retained response (29.6% complete). There were signi cant reductions in mPDAI in the entire cohort from a pre-induction score of 8.5 6 0.3 to 2 6 3.4 at nal
follow-up (mean, SE; P < 0.002) and in CRP from 29.48 6 6.2 mg/L to 5.76 6
1.6 mg/L (P < 0.001). Age, gender, smoking, pre-pouch disease extent and CRP did not affect response to IFX. Pretreatment mPDAI score did not signi cantly affect sustained IFX response but higher pretreatment mPDAI scores were negatively associated with early mucosal healing (P 0.056; OR 0.5, 95% CI, [0.2-1.010]).
Patients with pretreatment mPDAI .10 were less likely to have initial mucosal
healing (P 0.03). Initial mucosal healing was associated with sustained complete response (P < 0.05; [OR] 13.2; 95% con dence interval [CI], 1.0-165). Presence of ASCA was associated with higher initial mPDAI (P 0.016), but no difference in treatment response. IFX responders had a lower number of positive antibody titers
(2 positive titers versus 3, mean; P < 0.05). Having .2 positive titers was associated with longer term non-response to IFX though not signi cantly.
CONCLUSIONS: IFX was effective in the short- and longer term in patients with chronic refractory pouchitis and pouch complications. Mucosal healing after induction is the strongest clinical association with a complete sustained response to IFX in this group and pretreatment mPDAI may help predict this.
Cardiovascular Morbidity in Hospitalized Patients With In ammatory Bowel Disease Compared to Patients Without In ammatory Bowel Disease and Other Autoimmune Diseases
Badamas Jemilat1, Al Kazzi Elie2, Limketkai Berkeley3, Post Wendy4, Hut ess Susan3
1Johns Hopkins School of Medicine, Baltimore, Maryland, 2Johns Hopkins University - School of Medicine, Baltimore, Maryland, 3Johns Hopkins University School of Medicine, Baltimore, Maryland, 4Johns Hopkins School of Medicine, Baltimore, Maryland
BACKGROUND: Coronary artery disease (CAD) is common among the general population and individuals with autoimmune disorders such as systemic lupus erythematosus (SLE), and rheumatoid arthritis (RA) due to in ammation that
increases the risk of atherosclerosis. The prevalence of cardiovascular morbidity in in ammatory bowel disease in the United States is not well established. We aimed to examine the prevalence of cardiovascular morbidity among hospitalized in ammatory bowel disease (IBD) patients compared with the general hospitalized population and hospitalized RA and SLE patients in the 2011 Nationwide Inpatient Sample (NIS).
METHODS: NIS 2011 contains a 20% strati ed sample of all discharges from non- federal acute care hospitals in the United States. Each record in the NIS represents a single hospital discharge and includes patient demographics, hospital character- istics and up to 25 International Classi cation of Diseases 9th revision, clinical modi cation (ICD-9-CM) codes. All individuals with information on age and sex aged 18 to 100 with complete coding history were eligible. Hospitalizations with ICD-9-CM codes 555.x (Crohns disease) or 556.x (ulcerative colitis) were considered IBD. Comparisons were made with hospitalizations without diagnosis of IBD (non-IBD), SLE (710.0), and RA (714.0). Cardiovascular morbidity was de ned by acute myocardial infarction (410.x) or CAD (411.x, 412.x, 413.x and 414.x) as the primary diagnosis. Odds ratios (OR) comparing the prevalence of acute MI and CAD in IBD compared with non-IBD, SLE and RA were computed adjusting for age, gender, diabetes, hypertension, hyperlipidemia and smoking.
RESULTS: A total of 63,387 of 6,683,491 (1.0%) hospitalizations in the database included a diagnosis of IBD with 40,420 CD (63.8%) and 22,967 UC (36.2%) hospitalizations. Median ages for CD, UC and non-IBD patients were 49, 56 and 58 respectively. 57.5% of IBD and 59.2% of non-IBD patients were females. Acute MI was present in 0.5% of CD, 0.9% of UC, 1.8% of non-IBD, 1.2% of SLE and 1.6% of RA hospitalizations. CAD was present in 1.2% of CD, 1.8% of UC 3.8% of non-IBD, 2.5% of SLE and 3.2% of RA hospitalizations. The odds of acute MI was lower in CD (OR 0.50; 95% CI, 0.45-0.56) and UC (OR 0.68; 95% CI, 0.61-0.76) compared with the non- IBD hospitalizations. The odds of CAD was lower in CD (OR 0.45; 95% CI, 0.41-0.49) and UC (OR 0.54; 95% CI, 0.43-0.60) compared with the non-IBD hospitalizations. Acute MI was also less common in IBD patients than in SLE (OR 0.55; 95% CI, 0.51-0.61) and RA (OR 0.63; 95% CI, 0.56-0.73) with the same trend for the CAD comparisons (P < 0.05).
CONCLUSIONS: The prevalence of MI and CAD during hospitalization was lower in IBD than in the general population and other autoimmune diseases. This suggests that IBD patients are unique among the autoimmune diseases. Compared with other autoimmune diseases, IBD patients may not need to be screened more aggressively for cardiovascular disease but should undergo routine screening per general population guidelines.
Patterns of Hospitalization and Biologic Use in an Incident Cohort of Crohn's Disease Patients
Al Kazzi Elie1, Limketkai Berkeley2, Brant Steven2, Hut ess Susan2
1Johns Hopkins University - School of Medicine, Baltimore, Maryland, 2Johns Hopkins University School of Medicine, Baltimore, Maryland
BACKGROUND: Monoclonal antibody-based treatments for Crohn's disease (CD) are thought to have the potential to alter disease course if used soon after diagnosis or before substantial mucosal damage occurs. However, identifying cohorts of newly diagnosed patients to examine the effects of early versus late use is challenging, as many patients do not see a health care provider who routinely prescribes biologic treatments as a rst line therapy early in the disease course. We aimed to examine patterns of hospitalization and biologic use among newly diagnosed CD patients. METHODS: A sample of 1,000 cases of CD was identi ed by at least two encounters
associated with International Classi cation of Diseases, Ninth Revision, Clinical Modi cation (ICD-9-CM) code 555 in the U.S. military electronic health record from an eligible population of all active duty personnel diagnosed between 2001 and 2012. Individuals with ICD-9-CM 555 encounters within 1 year of enlistment were excluded because they may have had prevalent disease (n 18). Inpatient encoun-
ters with an ICD-9-CM code 555 in the rst 4 diagnosis positions were considered
hospitalizations for CD. Outpatient Current Procedural Terminology (CPT) codes were used to identify the use of biologics (J0135-adalimumab, J0718- certolizumab, J1745-in iximab, J2323-natalizumab). Time to rst use of a biologic was examined using a Cox proportional hazards model by year of CD diagnosis (prior to 2007, 2007-2012), adjusted for age, gender, race and hospitalization at diagnosis. Time to hospitalization by early biologic use within 6 months of diagnosis was also examined.
RESULTS: The eligible 982 cases had a median age at diagnosis of 27 years (range 20-55 years) and median year of diagnosis of 2007. Median follow-up time after diagnosis was 2.0 years (range 0.02-11.6 years). Females comprised 17.6% of cases and 72.6% of cases were white. Hospitalizations occurred in 38.3% of patients with 48.2% of the hospitalizations occurring at the time of diagnosis. Multiple hospital-
izations occurred in 16.6% of patients. Biologics were used by 9.4% (N 92) of patients with 8.7% receiving in iximab and 0.9% (n 9) adalimumab; no case received certolizumab, natalizumab, or both adalimumab and in iximab. First hos-
pitalization preceded biologic use in 83.7% (n 77) of biologic users and 20.7%