Faculty Bibliography

Introduction: Previous clinical trials and societal guidelines have established preoperative risk factors for postoperative recurrence (POR) of Crohn's disease (CD). However, in patients who are in endoscopic remission at time of their index postoperative ileocolonoscopy, risk factors for later development of POR is unknown. This study aimed to assess the relative risk of POR after initial endoscopic remission.
Method(s): Retrospective cohort study of adult CD patients who underwent ileocolonic resection (ICR) between the years of 2009-2020. Patients with bowel continuity restoration and index postoperative ileocolonoscopy with modified Rutgeerts' score of <= i2a disease were included. Postoperative ileocolonoscopies were assessed by retrospective application of Rutgeerts' score. The primary outcome was a composite outcome of objective recurrence including radiologic (active disease on cross-sectional enterography), endoscopic (Rutgeerts score >= i2b), and surgical (repeat ICR) POR after initial endoscopic remission.
Result(s): 232 CD patients (median age 35 years, 45.2% stricturing, 15.5% >=2 prior ICR) with initial endoscopic remission (98 i0, 49 i1, 85 i2a disease) were included in the study. The median time to index postoperative colonoscopy was 334 days. 131 (56.5%; 46.9% i0, 53.1% i1,69.4% i2a) patients went on to develop composite POR at median interval of 2.84 years. Risk of composite POR was associated with index Rutgeerts' score (p=0.008; 41.4% endoscopic POR, 38.3% radiologic POR, 14.2% surgical POR). On multivariable Cox proportional hazard regression, patients with i1 disease (aHR: 1.76 95% CI [1.04-2.99]; p=0.036) or i2a disease (aHR; 2.24 [1.49-3.36]; p<0.001) at initial postoperative colonoscopy were at increased risk for composite POR relative to patients with i0 disease (Figure 1; Table 1). Additionally, patients that were exposed to biologics prior to ICR were at increased risk for late POR relative to biologic-naive patients (aHR: 1.84 [1.15-2.95]; p=0.012). Patients with i2a disease was independently associated with decreased time to radiologic (aHR: 1.66 [1.04-2.65]; p=0.033), endoscopic (aHR:2.85 [1.76-4.59]; p<0.001 ), and surgical (aHR: 2.42 [1.11-5.29]; p=0.027) POR (Table 1).
Conclusion(s): Postoperative CD recurrence following early endoscopic remission is common. Early mild ileal or anastomotic inflammation confers increased risk for late recurrence. Future studies to determine surveillance strategies for patients in initial endoscopic remission are needed

Introduction: Traveler's diarrhea is the most common travel related illness and is often encountered in the outpatient setting. Most pathogens that cause traveler's diarrhea are poorly characterized by conventional stool testing such as stool culture and ova and parasite exam. Multiplex polymerase chainreaction based gastrointestinal pathogen panels (GI PCR) offer advantages in sensitivity and turnaround time, however their impact on clinical management is unclear. Our aim was to evaluate the clinical impact of GI PCR testing in outpatients presenting with traveler's diarrhea.
Method(s): We performed a retrospective study comparing outpatients presenting with acute gastrointestinal symptoms at an academic medical center who received stool testing with GI PCR from September 2015 to April 2019, to patients who presented with similar symptoms from February 2009 to April 2012, when GI PCR was unavailable, and received stool culture and ova and parasite exam ("conventional testing"). Patients who received GI PCR were matched by age and sex to patients who received conventional testing. We recorded pathogens isolated, demographic data, history of travel within 30 days of presentation, presenting symptoms and rates of antibiotic therapy.
Result(s): We identified 1,018 outpatients who received GI PCR or conventional stool testing. (n5509 each). There were 70 patients in the conventional cohort and 78 in the GI PCR cohort, respectively, with traveler's diarrhea. In patients who received GI PCR, we observed higher diagnostic yield in patients with traveler's diarrhea who received GI PCR, compared to patients without a history of recent travel (54 vs. 39%, p50.01, Table 1). There was no difference in diagnostic yield in patients who received conventional stool testing (9 vs 10%, p50.74). Patients with travelers' diarrhea were more likely to have pathogenic species of Escherichia coli (E. coli) identified on testing (45 vs. 22%, p, 0.01, Table 2), particularly Enteroaggregative E. coli (23 vs. 6%, p, 0.01). Patients with traveler's diarrhea were more likely to receive antibiotics whether they received GI PCR (38 vs. 27%, p50.04) or conventional testing (35 vs. 25%, p50.06).
Conclusion(s): GI PCR was associated with a higher diagnostic yield in patients presenting with travelers' diarrhea. Patients with travelers' diarrhea were significantly more likely to have pathogenic E. coli species identified on PCR testing, with up to 45% of patients testing positive

Introduction: Health disparities in health care utilization and clinical outcomes are strongly linked to race, ethnicity, socioeconomic, and insurance status in many diseases. We aimed to measure racial and socioeconomic disparities in inflammatory bowel disease (IBD) care including disease characteristics, treatment, health care utilization, and clinical outcomes.
Method(s): We performed a retrospective chart review of patients 18 years and older with established IBD care at a private or public hospital in New York City. Established care was defined as having at least 1 colonoscopy and 1 clinic visit within the same hospital. Patient demographics, disease characteristics, and outcomes were collected. Multivariate logistic regression was performed and reported as odds ratios (OR) with 95% confidence intervals (CI).
Result(s): A total of 322 patients, 50% from each hospital, were included in the analysis. Public hospital IBD patients were more likely to have Medicaid and visit the ED (17% vs 6%, OR 5.3, CI 2.1-14, and 41% vs. 11%, OR 4.0, 1.5-11,) compared to private hospital patients. Focusing on race, Black IBD patients were more likely to be hospitalized (36%, OR 7.6 CI 1.0-161) compared to other races. In terms of insurance, Medicare patients were more likely to be hospitalized and require surgery (28%, OR 4.9 CI 1.1-22 and 16%, OR 14, CI 1.7-120) compared to uninsured patients or those with Medicaid or private insurance. Unemployed IBD patients had higher rates of surgery (18% vs. 9%, OR 3.6, CI 1.2-10). CD patients who were unemployed had higher rates of prolonged steroid use, ED visits, and hospitalizations (OR 8.5, CI 1.2-79, OR 14, CI 1.7-173, OR 15, 1.9-191) compared to employed CD patients. UC patients with Medicaid were more likely to be hospitalized (OR 3.6, 1.0-14) compared to uninsured patients or those with Medicaid or private insurance. UC patients with private insurance were more likely to achieve mucosal healing (OR 5.3, 1.1-31) compared to uninsured patients or those with Medicaid or Medicare.
Conclusion(s): Race, employment, and insurance status were linked with disparities in IBD care. Public hospital use, Black race, unemployment, and non-private insurance were associated with various adverse IBD outcomes including increased ED visits, hospitalizations, requirement for surgery, prolonged steroid use, and decreased rates of mucosal healing. Future work should focus on identifying interventions to resolve disparities for these patient populations

Introduction: The etiology of acute diarrhea in solid organ transplant (SOT) recipients requires rapid diagnosis to avoid severe infection or graft loss, especially for those requiring hospitalization. We aimed to characterize factors associated with hospitalization for enteric infection as detected by multiplex gastrointestinal PCR panel (GI panel) stool testing in SOT recipients.
Method(s): We performed a cross-sectional study at two academic institutions of inpatient and outpatient SOT recipients who underwent stool testing over 3 years with a FilmArray GI panel during an acute episode of diarrhea. A multivariate analysis was performed to identify factors predictive of hospitalization.
Result(s): Of 300 SOT recipients with an acute episode of diarrhea, 68% (n=204) were hospitalized at the time of sample collection (Table 1). On univariate analysis, hospitalized patients were less likely to have a pathogen identified on the GI panel compared to non-hospitalized patients (33% vs 53%, p, 0.001) - specifically, fewer viral enteric infections (14% vs 27%, P=0.007). Hospitalized patients were more likely to have fever (29% vs 6.3%, p< 0.001) and had a higher 30-day mortality than those not hospitalized (16.9% vs 4.2%; P=0.002). The number of immunosuppressive agents was not associated with hospitalization (P=0.913). In the multivariate analysis adjusting for age, sex, race, Charlson's comorbidity index, antimicrobial prophylaxis, recent hospitalization, transplant type, degree of immunosuppression, and transplant age, fever (odds ratio [OR] 6.8; 95% confidence interval [CI] 2.9- 19), nausea/vomiting (OR 2.5; 95% CI 1.2-5.4) and negative GI panel (OR 2.0; 95% CI 1.1-3.7) were associated with hospitalization. Compared to other organs, renal transplant (OR 0.318, 95% CI 0.133- 0.723) was associated with lower likelihood of hospitalization.
Conclusion(s): In SOT recipients presenting with acute diarrhea, symptoms of severe illness, such as fever and emesis, and a negative GI Panel, were associated with hospitalization. Renal transplant was inversely associated with hospitalization. Diarrheal illness in SOT patients may be caused by pathogens not identified on the GI panel or from non-infectious causes such as immunosuppressive, antimicrobial, or bowel regimen medications. Further elucidation of these non-infectious diarrheal etiologies in SOT recipients hospitalized for acute diarrhea may potentially avoid major complications, such as graft loss, severe infection, and death.

Introduction: Postoperative recurrence (POR) of Crohn's disease (CD) following ileocecal resection (ICR) is common. Endoscopic CD activity guides postoperative management. Histologic activity precedes endoscopic activity. In patients in endoscopic remission, the influence of histologic activity on risk of subsequent POR is unknown. We aimed to assess the impact of histologic activity on subsequent endoscopic POR in CD individuals in endoscopic remission following ICR.
Method(s): Adult CD patients who underwent ICR between the years 2009 to 2020 with no active disease (Rutgeerts' i0) on index colonoscopy with biopsies of the neoterminal ileum or anastomosis were included. Those without a follow up after the first endoscopy were excluded. The rate of composite POR, including endoscopic (>=i2b disease), radiologic, surgical, and pharmacologic recurrence (i.e. change in treatment), was compared between patients with histologic activity versus quiescence using chi-square and Kaplan-Meier survival analysis curves. Cox proportional hazard modeling was used to compare histologic activity to high-risk features of surgical recurrence and use of postoperative biologic prophylaxis.
Result(s): There were 79 CD patients (Mean age 34 years, 49.4% ileocolonic disease, 71.2% stricturing, 34.5% >=1 prior ICR, 27.8% active smoking, and 25.3% treated with postoperative biologic prophylaxis) with i0 disease and random endoscopic biopsies after ICR that were included. Overall composite recurrence rate was 58.2% over median follow up of 1242 days. Histologic activity was not associated with composite POR (P=0.49) including endoscopic (P=0.50), radiologic (0.99), surgical (0.78), or pharmacologic recurrences (Biologic: P=0.59; Immunomodulator: P=0.99; Steroid POR: P=0.54). Composite POR was not significantly different by histologic activity group (P=0.29) (Figure 1). Adjusting for high risk predictive factors of POR, histologic activity was not associated with composite POR (HR 1.71 (0.92-3.21); P=0.09). While active smoking was significantly associated (HR 2.29 (1.11-4.72); P=0.025) with composite POR, postoperative biologic prophylaxis was protective (HR 0.41 (0.20-0.81); P=0.01) (Table 1).
Conclusion(s): Histologic activity in endoscopically normal exam is not associated with endoscopic, radiologic, surgical, and pharmacologic postoperative recurrences. Future studies are required to assess the utility of endoscopic biopsies in CD patients with i0 disease to better guide the management of postoperative CD

Introduction: Multiplex gastrointestinal pathogen panels (GI PCR) for the detection of enteric infection offer advantages in turnaround time and sensitivity compared to conventional modalities of stool testing. Previous research on GI PCR has focused almost exclusively on hospitalized patients, however, most cases of gastroenteritis are managed in the outpatient setting. Our aim was to compare GI PCR to conventional stool testing in outpatients with gastroenteritis and evaluate the impact on clinical decision-making and management.
Method(s): We performed a retrospective study of outpatients presenting with acute gastrointestinal symptoms at an academicmedical center from September 2015 to April 2019 who received stool testing with GI PCR, and from February 2009 to April 2012 who received stool culture and ova and parasite exam ("conventional testing"). Patients who received GI PCR were matched by age and sex to patients who received conventional testing. We recorded pathogens isolated, demographic data, presenting symptoms, risk factors, antibiotic therapy, and downstream associated healthcare utilization including abdominal imaging, endoscopy, emergency room visits, hospitalizations, and abdominal surgeries. Univariate and multivariate linear regression was used to estimate the effect of testing on empiric antibiotic therapy.
Result(s): We identified 1,018 outpatients who received GI PCR or conventional stool testing. (n = 509 each). A pathogen was isolated in 208 (41%) patients with GI PCR and 49 (10%, P< 0.01, Table 1) with conventional culture. There were no significant differences in overall rates of antibiotic therapy or average duration of therapy, but GI PCR was associated with less empiric therapy (47% vs 70%, P < 0.01), and higher rates of narrowing (13 vs 4%) and discontinuation (3 vs 0%, P < 0.01) of empiric antibiotics after the initial visit, with a 1.6-fold lower likelihood (95% CI 1.04-2.57, P= 0.04) of initiating empiric antibiotics on multivariable analysis. There were no significant differences in other outcomes.
Conclusion(s): GI PCR testing detected more pathogens compared to conventional stool testing, without a significant increase in antibiotics prescribed. PCR testing was associated with changes in prescribing patterns for antibiotics, with a shift towards less empiric therapy

Introduction: Acute diarrhea is common in solid organ transplant (SOT) recipients. Although immunosuppression regimens overlap among transplant organ types, it is not known whether particular organ transplants are at higher risk for infectious diarrhea. We compared rates of enteric infection as detected by stool multiplex gastrointestinal PCR panel (GI panel) testing among multiple types of SOT recipients.
Method(s): We performed a cross-sectional study at two academic institutions of inpatient and outpatient recipients of single solid organ transplantation who underwent stool testing with a FilmArray GI panel during an acute episode of diarrhea over 3 years. Patients were stratified by transplant type and the primary outcome was infectious diarrhea by GI panel.
Result(s): Of 300 subjects, there were 58 (19%) heart, 65 (22%) liver, 68 (23%) lung, and 109 (36%) renal single transplant recipients. Use of mTOR inhibitors was less common in lung and renal transplant patients (P<0.001). Lung transplant patients were more likely to be on more immunosuppressive agents (P<0.001; Table) - especially steroids (P<0.001) - and more likely to be on antimicrobial prophylaxis for opportunistic infections (OI). A positive GI panel was identified in 118 (39%) patients, and was more common in renal transplant recipients (54% renal vs. 41% heart vs. 28% liver vs. 25% lung, P<0.001). Bacterial infection was less common in lung transplant recipients (8.8% lung vs. 34% renal vs. 22% heart vs. 22% liver, P=0.002). Hospitalization for acute diarrhea was more common in lung transplant recipients and less common in renal transplant recipients (78% lung vs. 58% renal vs. 68% liver vs. 76% heart, P=0.019). In the multivariate analysis controlling for age, sex, race, comorbidities, prophylaxis, recent hospitalization, and transplant age, use of mTOR inhibitors (Odds Ratio [OR] 3.7, 95% Confidence Interval [CI] 1.1-15), hospitalization (OR 2.3, 95% CI 1.3-4.3), and lung transplant (OR 2.5, 95% CI 1.0-6.4) predicted negative GI panel (Figure 1).
Conclusion(s): In SOT recipients with acute diarrhea, lung transplant patients - who receive more immunosuppression and OI prophylaxis - are more likely to be hospitalized and have a negative GI panel. Lung transplantation, hospitalization, and mTOR inhibitor use independently predict negative GI panel. Transplant and immunosuppression type should be accounted for during the management of acute diarrhea in SOT recipients.

Introduction: Previous studies have demonstrated that exposure to anti-TNFa medications or immunomodulators (IMM) does not increase the risk of new or recurrent cancer in patients with inflammatory bowel disease (IBD) and a previous history of cancer. There is little data regarding this risk after the use of newer biologics such as ustekinumab and vedolizumab. In this study, we assessed whether patients with IBD and a history of previous cancer who are exposed to these newer agents have an increased risk of developing subsequent cancer.
Method(s): We performed a retrospective cohort study of patients with IBD and cancer from a single academic medical center between January 2013 and December 2020. We recorded information on demographics, cancer type and treatment, and IBD characteristics and drug exposures. The primary exposure was type of IBD monotherapy after a cancer diagnosis. The primary outcome was development of new or recurrent cancer.
Result(s): Of 401 patients with IBD and a history of cancer, 37 subsequently received vedolizumab, 14 ustekinumab, 31 IMM, 41 anti-TNF, 11 combination anti-TNF with an IMM, and 267 were not exposed to any immunosuppression following a cancer diagnosis (Table 1). There were no differences in duration of IBD, median age at cancer diagnosis, or smoking history. During a total median follow-up of 52 months, 81 (20%) patients developed incident cancer including 6 (16%) exposed to vedolizumab, 2 (14%) to ustekinumab, 3 (10%) to IMM, 12 (29%) to anti-TNF, 2 (18%) to combination anti-TNF with an IMM, and 56 (21%) with no immunosuppression (P = 0.41). Sensitivity analyses assessing any history of exposure to vedolizumab or ustekinumab, inclusive of both single and multiple biologic exposures, also did not reveal an increased rate of incident cancer.
Conclusion(s): In this single-center study, vedolizumab or ustekinumab monotherapy in patients with IBD and a history of cancer was not associated with an increase in new or recurrent cancer compared with anti-TNF, IMM, or no immunosuppression. Prospective studies are needed to confirm this conclusion

Introduction: While ileocecal resection (ICR) often leads to remission of Crohn's disease (CD), most patients will relapse. Evidence-based guidelines suggest biologic prophylaxis to prevent post-operative recurrence (POR) in high-risk patients and endoscopy within 6-12 months of surgery to assess for POR. Post-operative monitoring intervals and biomarker/cross-sectional imaging use have not been protocolized, leading to diversity in management.We aimed to describe the surveillance approach for CD patients after ICR and determine the association with POR.
Method(s): This was a retrospective chart review of CD patients who underwent ICR with >=1 year of follow-up at an academic medical center. We grouped patients into high- (HR) and low-risk (LR) for POR per guidelines and compared baseline data, medication use, and surveillance approach. We then stratified patients by colonoscopy within 1 year, labs within 1 year, or imaging within 1 year to compare POR rates. Biomarker, radiographic, and endoscopic POR were defined as high CRP/fecal calprotectin (FC), inflammation on CT/MRE, and Rutgeerts >= i2, respectively.
Result(s): Of 154 patients, 142 were HR. Median follow-up time was 32 months (m). 9 (75%) of LR and 104 (73%) of HR patients had colonoscopies within 12m of surgery. 7 (58%) of LR and 60 (42%) of HR patients had another colonoscopy between 12-24m. CRP was assessed in>70% of LR and HR between both 6-12m and 12-24m. FC was done in 33% of LR and 16% of HR between 6-12m and in 17% of LR and 18% of HR between 12-24m. Of those who received biologics, prophylactic treatment was initiated in 3 (43%) of LR and 86 (67%) of HR, median time to start was 5 weeks. Endoscopic POR rates were similar between those with biomarkers within 1 year vs without (p=0.99), imaging within 1 year vs without (p=0.17), and colonoscopy within 1 year vs without (p=0.53). Of patients who had endoscopic POR, 10 (59%) of those with colonoscopy within 1 year of surgery went into remission on ensuing scopes, compared to 0 of those who did not have colonoscopy within 1 year of surgery (p=0.32).
Conclusion(s): Despite guidelines, over one-quarter of patients did not undergo endoscopic monitoring for POR within 1 year of ICR. Biomarker and imaging use varies. Multiple modes of surveillance within 1 year were not associated with reduced endoscopic POR, yet it may be associated with subsequent remission as it likely altered management. Future studies should establish a surveillance protocol to maximize relapse detection and maintain remission

Introduction: With an aging population, management of biologic therapy in IBD patients with active or recent cancer is challenging. We evaluated the comparative safety of non-tumor necrosis factor (TNF)-a directed therapy vs. TNFa antagonists vs. immunomodulator monotherapy (IMM) in IBD patients with active or recent cancer (<=5 years).
Method(s): Through the collaborative REACH-IBD (Rising Educators Academics and Clinicians Helping IBD) research initiative, we conducted a retrospective, multi-center cohort study. We included IBD patients from 12 centers with active cancer (Cohort A) or recent cancer within = years (Cohort B) who were treated with non-TNFa biologics vs. TNFa antagonists (reference) after cancer diagnosis.We excluded patients with nonmelanoma skin cancer. Primary outcomes were cancer progression (Cohort A) or new/recurrent cancer (Cohort B). We performed Cox proportional hazard analysis to compare the safety of different biologics.
Result(s): (Cohort A)We included 107 patients with active cancer (5416y, 62% male, 72% solid cancer, 400 person-year follow-up), of whom 35 were treated with non-TNFa biologics (29 vedolizumab, 6 ustekinumab), 45 with TNFa antagonists and 27 with IMM (Table 1). Overall, 19 patients had progression of cancer, 13 died and 20 were hospitalized for serious infection (Figure 1A). After adjusting for age and type of active cancer, there was no difference in the risk of cancer progression (non-TNFa biologics vs. TNFa antagonists: HR, 1.55; 95% CI, 0.48-5.03), mortality (HR, 2.74; 95% CI, 0.25-30.5) and serious infections (HR, 1.90; 95% CI, 0.15-24.0) between patients treated with non-TNFa biologics vs. TNFa antagonists. (Cohort B) We included 141 patients with recent prior cancer (5214y, 51% male, 86% solid cancer; duration of remission prior to starting biologics, 1719m) of whom 54 were treated with non-TNFa biologics (40 vedolizumab, 14 ustekinumab), 63 with TNFa antagonists and 24 with IMM (Table 1). Overall, 14 patients had recurrence of cancer (or developed new incident cancer) and 6 died (Figure 1B). After adjusting for age, type of prior cancer and duration of remission, there was no difference in the risk of cancer recurrence between non-TNFa biologics vs. TNFa antagonists (HR, 0.97; 95% CI, 0.16-5.75).
Conclusion(s): In IBD patients with active or recent cancer (within = years), non-TNFa-directed biologics and TNFa antagonists have comparable safety. Choice of biologic should be dictated by IBD disease severity, in collaboration with an oncologist