How does cirrhosis impact mortality, morbidity, and resource utilization in non-variceal upper gastrointestinal bleeding? A nationwide analysis.

Introduction: Upper gastrointestinal bleeding is common in liver cirrhosis patients. Studies have described the prognostic impact of liver disease in non-variceal upper gastrointestinal bleeding (NVUGIB), but a direct subgroup comparison is lacking using a large database. Aim: To study the impact of NVUGIB on hospital-based outcomes in patients with cirrhosis. Material and methods: This is a retrospective study using Nationwide Inpatient Sample (NIS) employing International Classification of Diseases (ICD-10) codes for adult patients with a primary diagnosis of NVUGIB. Mortality, morbidity, and resource utilization were compared. Analyses were performed using STATA, proportions were compared using Fisher exact test, and continuous variables using Student's t-test. Confounding variables were adjusted using propensity matching, multivariate logistic, and linear regression analyses. Results: Of 107,001,355 discharges, 957,719 had a diagnosis of NVUGIB. Of those, 92,439 had cirrhosis upon admission. NVUGIB patients with cirrhosis had higher adjusted odds of mortality and intensive care unit (ICU) admission than patients without cirrhosis (adjusted odds ratio (AOR) for mortality 1.31, p < 0.001, ICU admission AOR = 1.29, p < 0.001). NVUGIB patients with cirrhosis had shorter length of stay (LOS) by 0.44 days (p < 0.001), greater hospital costs per day ($3114 vs. $2810, p < 0.001), and lower odds of acute kidney injury (AOR = 0.81, p < 0.001). In addition, the cirrhotic patients had higher odds of receiving endoscopic therapy (AOR = 1.08, p < 0.001). There was no difference between the 2 groups' requirements of packed red blood cell transfusion, parenteral nutrition, hypovolaemic shock, and endotracheal intubation. We also identified novel independent predictors of mortality from NVUGIB in cirrhosis patients. Conclusions: Cirrhosis presents greater mortality and morbidity burden and greater healthcare resource utilization from NVUGIB.

The Association of Lactose Intolerance With Colon and Gastric Cancers: Friend or Foe?

BACKGROUND: Lactose intolerance (LI) appears usually in later ages when the lactase enzyme becomes deficient or absent in the small intestine. Conflicting results have been reported in the literature about the association of lactose intolerance with various gastrointestinal malignancies. Hence, our aim was to study the association between LI, colon cancer (CCa), and gastric cancer (GC) using a large database. METHODS: A cross-sectional study was performed using the National Inpatient Sample (NIS) database between 2004 and 2014. We identified adult patients (18-90 years) who were diagnosed with LI (study group) using appropriate International Classification of Diseases, Ninth Revision (ICD-9) codes. The control group comprised patients who did not have a diagnosis of LI. We identified the diagnosis of CCa and GC in both study and control groups using the ICD-9 codes. Univariable and multivariable logistic regression analyses were performed to assess the association between LI, CCa, and GC. RESULTS: The total population comprised 71,360,501 patients, of which 57,909 (0.08%) were diagnosed with LI. LI patients were older (62 vs 51 years) with more females (61.5% vs 60.1%) and less African American patients (11.8% vs 14.3%) (p <0.0001 for all). In addition, LI patients had more smoking (12.4% vs 12%) and obesity (15% vs 8.9%). On the other hand, patients in the LI group had less alcohol use (3.8% vs 4.2%) (p <0.0001). After adjusting for the age, gender, race, smoking, alcohol, obesity, and inflammatory bowel disease, the LI group had a slightly lower rate of CCa (OR 0 .974, 95%CI 0.906-1.048, p = 0.486) and a lower rate of GC (OR: 0.993, 95%CI 0.924-1.068, p =0.853); however, the results were not statistically significant. CONCLUSION: Patients with lactose intolerance may have a lower risk of colon and gastric cancer. However, these findings were not statistically significant. Further studies are needed to understand this association.

Management of Refractory Gastrointestinal Bleeding in Hereditary Hemorrhagic Telangiectasia with Bevacizumab.

Hereditary hemorrhagic telangiectasia (HHT) is an autosomal dominant disorder resulting in vascular malformations of several organs including the pulmonary, cerebral, and gastrointestinal systems. One sequela is recurrent gastrointestinal (GI) bleeding. Bevacizumab (Bev) is emerging as an effective treatment of recurrent gastrointestinal bleeding in HHT. Bev is a recombinant monoclonal antibody that inhibits vascular endothelial growth factor (VEGF), an integral part of angiogenesis.

P040 Latent Tuberculosis Infection In A Patient With A Negative Interferon Gamma Release Assay Prior to Biologic Initiation for Crohn's Disease.

CASE: Biologic therapy has revolutionized the treatment of inflammatory bowel disease (IBD). Nevertheless, it is well known that biologics are associated with an increased risk of opportunistic infections such as tuberculosis (TB). Screening tests indicated before biologic initiation, namely interferon gamma release assay (IGRA) or tuberculin skin tests (TST), may be falsely negative. We present the case of a male with uncontrolled Crohn's disease (CD) with a negative screen for TB and how a high index of suspicion led to the discovery of latent tuberculosis infection (LTBI) and the appropriate deferment of biologic therapy. A 63-year-old South Asian male with coronary artery disease status post bypass graft, prostate cancer, diabetes and iron deficiency anemia presented with fatigue and dizziness for one week. He reported dark stools since immigration from Bangladesh two years ago. Vital signs and examination were unremarkable. Laboratory studies revealed a gradual decline in hemoglobin from 9 g/dL to 7 g/dL over several months. Esophagogastroduodenoscopy was unremarkable. Colonoscopy revealed friability, inflammation and erythema of the ileocecal valve and terminal ileum (TI) with minimal strictures. Biopsy was consistent with CD. He was started on azathioprine. Subsequent computed tomography (CT) enterography revealed thickened TI extending over a length of 9-10 centimeters. Due to disease severity and continued symptoms, he was referred to our tertiary care center where it was determined that he may benefit from biologic therapy. Patient underwent IGRA testing which was negative. Given his immigration from an endemic area and use of immunosuppressants, a CT of the chest was obtained which was significant for tree-in-bud opacities within the right upper lobe and multiple nodules. Transbronchial biopsy of the largest nodule showed necrotizing granulomatous inflammation. Bronchoalveolar lavage culture was positive for Mycobacterium tuberculosis. The patient was diagnosed with LTBI. Anti-tuberculin therapy was initiated, and biologic therapy was postponed. Biologic agents are increasingly utilized in the treatment of IBD. These agents confer an increased risk of infections i.e. TB and poor outcomes from severe disease. Patients with IBD are also susceptible to infections due to the immunosuppression imposed by the disease itself. The risk of TB varies with the biologic used with the highest relative risks associated with adalimumab and infliximab. Screening for LTBI is indicated in all patients for which biologics are being considered. The goal is to identify and treat individuals at high risk for TB reactivation. A thorough history and examination should be performed to assess pertinent signs and symptoms. While IGRA has demonstrated higher sensitivity and specificity compared to TST, immunosuppressed patients may have a false negative result as in our case. Combination testing may be the most sensitive approach though evidence is limited. In cases of high suspicion, imaging with plain radiograph or CT scan is a useful tool in supporting the diagnosis of LTBI and may show predominantly upper lobe cavitary disease. Biologic therapy for IBD should be deferred until at least one month of treatment for LTBI has been completed.

Association between Inflammatory Bowel Disease and Lactose Intolerance: Fact or Fiction.

Background/Aims: Inflammatory bowel disease (IBD) is a complex condition precipitated by genetic susceptibility and possibly a disturbed microbiome. The role of dairy foods in IBD is controversial. This study examined the association between lactose intolerance (LI) and IBD. Methods: Data on hospital admissions of all IBD adult patients were extracted from the National Inpatient Sample database between 2004 and 2014. The comorbidities and outcomes of interest were defined by querying all the diagnostic and procedural fields for the corresponding International Classification of Diseases 9th version (ICD-9) codes. Patients with IBD were defined as the "study group," and the patients who did not have IBD were defined as the "control group". LI was identified in both groups using the ICD-9 codes. Multivariate logistic regression was performed to examine the association between IBD and LI. Results: The total population was 71,342,237 patients, of which 598,129 (0.83%) had IBD. The IBD patients were younger (52 years vs. 57 years) and with fewer females (57.5% vs. 60.1%) (p<0.001 for all). After adjusting for the potential confounding factors, the IBD group had a significantly higher rate of LI (OR 2.71, 95% CI 2.55-2.88, p<0.001) compared to the non-IBD group. The findings were similar on the further stratification of IBD into Crohn's disease compared to the control group (OR 2.70, 95% CI 2.50-2.92, p<0.001) and ulcerative colitis compared to the control group (OR 2.71, 95% CI 2.46-2.98, p<0.001). Conclusions: IBD patients have a 2.7 times higher risk of LI. Screening for LI in this population is warranted to avoid confusing or overlapping symptomatology.

The Wnt-ß-Catenin-IL-10 Signaling Axis in Intestinal APCs Protects Mice from Colitis-Associated Colon Cancer in Response to Gut Microbiota.

Loss of immune tolerance to gut microflora is inextricably linked to chronic intestinal inflammation and colitis-associated colorectal cancer (CAC). The LRP5/6 signaling cascade in APCs contributes to immune homeostasis in the gut, but whether this pathway in APCs protects against CAC is not known. In the current study, using a mouse model of CAC, we show that the LRP5/6-ß-catenin-IL-10 signaling axis in intestinal CD11c+ APCs protects mice from CAC by regulating the expression of tumor-promoting inflammatory factors in response to commensal flora. Genetic deletion of LRP5/6 in CD11c+ APCs in mice (LRP5/6ΔCD11c) resulted in enhanced susceptibility to CAC. This is due to a microbiota-dependent increased expression of proinflammatory factors and decreased expression of the immunosuppressive cytokine IL-10. This condition could be improved in LRP5/6ΔCD11c mice by depleting the gut flora, indicating the importance of LRP5/6 in mediating immune tolerance to the gut flora. Moreover, mechanistic studies show that LRP5/6 suppresses the expression of tumor-promoting inflammatory factors in CD11c+ APCs via the ß-catenin-IL-10 axis. Accordingly, conditional activation of ß-catenin specifically in CD11c+ APCs or in vivo administration of IL-10 protected LRP5/6ΔCD11c mice from CAC by suppressing the expression of inflammatory factors. In summary, in this study, we identify a key role for the LRP5/6-ß-catenin-IL-10 signaling pathway in intestinal APCs in resolving chronic intestinal inflammation and protecting against CAC in response to the commensal flora.

Liver abscess and bacteremia caused by lactobacillus: role of probiotics? Case report and review of the literature.

BACKGROUND: Lactobacilli are non-spore forming, lactic acid producing, gram-positive rods. They are a part of the normal gastrointestinal and genitourinary microbiota and have rarely been reported to be the cause of infections. Lactobacilli species are considered non-pathogenic organisms and have been used as probiotics to prevent antibiotic associated diarrhea. There are sporadic reported cases of infections related to lactobacilli containing probiotics. CASE PRESENTATION: In this paper we discuss a case of an 82 year old female with liver abscess and bacteremia from lactobacillus after using probiotics containing lactobacilli in the course of her treatment of Clostridium difficile colitis. The Lactobacillus strain identification was not performed and therefore, both commensal microbiota and the probiotic product should be considered as possible sources of the strain. CONCLUSION: Lactobacilli can lead to bacteremia and liver abscesses in some susceptible persons and greater awareness of this potential side effect is warranted with the increasing use of probiotics containing lactobacilli.

Risk of Gastrointestinal Bleeding with Rivaroxaban: A Comparative Study with Warfarin.

Introduction. The risk of gastrointestinal (GI) bleeding with rivaroxaban has not been studied extensively. The aim of our study was to assess this risk in comparison to warfarin. Methods. We examined the medical records for patients who were started on rivaroxaban or warfarin from April 2011 to April 2013. Results. We identified 300 patients (147 on rivaroxaban versus 153 on warfarin). GI bleeding occurred in 4.8% patients with rivaroxaban when compared to 9.8% patients in warfarin group (p = 0.094). GI bleeding occurred in 8% with therapeutic doses of rivaroxaban (>10 mg/d) compared to 9.8% with warfarin (p = 0.65). Multivariate analysis showed that patients who were on rivaroxaban for ≤40 days had a higher incidence of GI bleeding than those who were on it for >40 days (OR = 2.8, p = 0.023). Concomitant use of dual antiplatelet agents was associated with increased risk of GI bleeding in the rivaroxaban group (OR = 7.4, p = 0.0378). Prior GI bleeding was also a risk factor for GI bleeding in rivaroxaban group (OR = 15.5). Conclusion. The incidence of GI bleeding was similar between rivaroxaban and warfarin. The risk factors for GI bleeding with rivaroxaban were the first 40 days of taking the drug, concomitant dual antiplatelet agents, and prior GI bleeding.

Comparison of Ischemic Colitis in the Young and the Elderly.

BACKGROUND: Ischemic colitis is traditionally known as a disease of the elderly; however, its recognition among the young recently has increased. The aim of this study was to illustrate the features of ischemic colitis in a younger population. METHODS: Medical records of patients with ischemic colitis from January 2007 to January 2013 were reviewed. The study was conducted in 2 hospitals, and the patients were divided into 2 groups: < 50 and ≥ 50 years old. RESULTS: A total of 118 patients with ischemic colitis were identified. Fifteen patients (12.7%) were < 50 years of age; 103 patients (87.3%) were ≥ 50 years old. While drugs and vasculitis­as a group­was the most common precipitating factor for ischemic colitis in the younger age group, constipation was the most common precipitating factor in the older age group. All patients in the younger group had rectal bleeding vs 70.9% in the older group (P = 0.009). History of coronary artery disease, dyslipidemia, and hypertension were higher in the older group. Length of hospital stay was shorter in the younger group (3.4 days) than the older group (7.2 days). CONCLUSION: In this study, 12.7% of the patients were under age 50. All patients in this "young" age group experienced rectal bleeding and their hospital stay was shorter.

Gastrointestinal bleeding with dabigatran, a comparative study with warfarin: a multicenter experience.

BACKGROUND/AIMS: The risk of gastrointestinal (GI) bleeding with dabigatran when compared to warfarin has been controversial in the literature. The aim of our study was to assess this risk with the use of dabigatran. METHODS: We examined the medical records of patients who were started on dabigatran or warfarin from October 2010 to October 2012. The study was conducted in two hospitals. RESULTS: A total of 417 patients were included (208 dabigatran vs. 209 warfarin). GI bleeding occurred in 10 patients (4.8%) in the dabigatran group compared to 21 patients (10.1%) in the warfarin group (p=0.0375). Multivariate analysis showed that patients who were on dabigatran for ≤ 100 days had a higher incidence of GI bleeding than those who were on it for >100 days (p=0.0007). The odds of GI bleeding in patients who were on dabigatran for ≤ 100 days was 8.2 times higher compared to those who were on the drug for >100 days. The incidence of GI bleeding in patients >65 years old was higher than in those <65 years old (p=0.0453, OR=3). History of previous GI bleeding was another risk factor for GI bleeding in the dabigatran group (p=0.036, OR=6.3). The lower GI tract was the most common site for GI bleeding in the dabigatran group (80.0% vs. 38.1%, p=0.014). CONCLUSIONS: The risk of GI bleeding was lower with dabigatran. The risk factors for GI bleeding with dabigatran were the first 100 days, age >65 years, and a history of previous GI bleeding.

Risk of gastrointestinal bleeding with dabigatran: a head-to-head comparative study with rivaroxaban.

INTRODUCTION: The risk of gastrointestinal (GI) bleeding of dabigatran and rivaroxaban is relatively unexplored. The aim of our study was to compare this risk in both drugs. METHODS: We examined the medical records of patients on either dabigatran or rivaroxaban from October 2010 to April 2013 in two hospitals. RESULTS: A total of 374 patients (147 rivaroxaban vs. 227 dabigatran) were identified. GI bleeding occurred in 5.3% in the dabigatran when compared to 4.8% in the rivaroxaban group (p = 0.8215). Multivariate analysis showed that the odds of GI bleeding while on dabigatran for ≤40 days when compared to ≥40 days was 8.3 (p < 0.0001). In the rivaroxaban group, patients who were on the drug for ≤40 days had a higher incidence of bleeding when compared to those >40 days (OR = 2.8, p = 0.023). Concomitant use of antiplatelets (single or dual) or non-steroidal anti-inflammatory drugs was not associated with increased bleeding in the dabigatran group; however, the use of dual antiplatelet agents with rivaroxaban was associated with an increased risk of GI bleeding (OR = 7.4, p = 0.0378). Prior GI bleeding had a higher risk of bleeding in the rivaroxaban group (OR = 15.5, p = 0.0002). CONCLUSION: Dabigatran was not associated with a higher incidence of GI bleeding. Both drugs had a higher bleeding risk in the first 40 days.

Ischemic colitis: A forgotten entity. Results of a retrospective study in 118 patients.

OBJECTIVE: The aim of our study was to document our 6-year experiences in identifying the clinical characteristics, laboratory findings, risk factors and the outcomes of patients with ischemic colitis (IC) in a community hospital setting. METHODS: The medical records of patients who were diagnosed with IC from 2007 to 2013 in two community hospitals were retrospectively reviewed. Their clinical characteristics, laboratory results, radiological, endoscopic and histological evidence, anatomic location of the lesion, comorbidities, concomitant use of drugs, and so on, were collected. RESULTS: A total of 118 patients with IC was identified, most were elderly individuals with a female predominance. The most common symptoms were abdominal pain, rectal bleeding and diarrhea. Hypertension, hyperlipidemia, coronary artery disease and diabetes mellitus were the most common comorbidities. Erythema, edema and erosions/ulcerations were the most common endoscopic findings. Left colon was the most affected location of lesion (84.8%), and there was one case of pancolitis. The descending colon was the most common affected segment, while rectum was the least affected segment. Severe IC occurred in 12.7% of the patients. Death within 30 days from the diagnosis of the disease occurred in 4.2%. CONCLUSIONS: IC is majorly occurred in elderly with a female predominance. Cardiovascular disease and its associated risk factors are the most common comorbidities. Left colon is the most affected location of the disease and the overall mortality rate was 4.2%. Physicians should make every effort to identify these patients, especially those with high risks.

Risk factors of recurrent ischemic colitis: a multicenter retrospective study.

BACKGROUND/AIMS: Recurrence of ischemic colitis (IC) has not been studied extensively. The aim of this study was to investigate the characteristics of recurrent IC in the community setting and to identify any risk factors. METHODS: We conducted a retrospective study in two community hospitals. Medical records of patients with IC from January 2007 to January 2013 were reviewed. Demographic details, clinical features, co- morbidities, concomitant use of medications,laboratory studies, imaging findings, endoscopic and histological features, surgery, hospital stay, and death within 30 days were collected. Patients were divided into two groups (recurrent IC group, non-recurrent IC group). RESULTS: A total of 118 patients with IC were identified. IC recurred in 10 patients (8.5%) during the study period. Half of the patients in the recurrent IC group were current smokers as compared to only 18.7% of patients in the non-recurrent group. In the recurrent IC group, 20.0% of patients never smoked as compared to 61.7% in the non-recurrent group (p=0.027).Abdominal aortic aneurysm (AAA) was more frequent in the recurrent IC group (40.0% vs. 4.7%; p=0.003). No differences in other clinical symptoms, CT scan findings, comorbidities, endoscopic features, or use of concomitant medications were observed between the two groups. The need for surgical intervention, blood transfusion, intensive care unit stay, mechanical ventilation,length of hospital stay, and anatomic location of affected segments did not differ between the two groups. CONCLUSIONS: IC recurred in 8.5% of patients during the six-year study period. Current smoking status and presence of AAA were identifying risk factors for recurrence of IC.

The metoclopramide black box warning for tardive dyskinesia: effect on clinical practice, adverse event reporting, and prescription drug lawsuits.

OBJECTIVES: We examined the effects of the black box warning about the risk of tardive dyskinesia (TD) with chronic use of metoclopramide on management of gastroparesis within a single clinical practice, and on reporting of adverse events. METHODS: Medical records of gastroparesis patients were evaluated for physician management choices. The FDA Adverse Event Reporting System (FAERS) was analyzed for event reports, and for lawyer-initiated reports, with metoclopramide from 2004 to 2010. Google Scholar was searched for court opinions against metoclopramide manufacturers. RESULTS: Before the black box warning, 69.8% of patients received metoclopramide for gastroparesis, compared with 23.7% after the warning. Gastroenterologists prescribed domperidone more often after than before the warning. Metoclopramide prescriptions decreased after 2008. Adverse event reporting increased after the warning. Only 3.6% of all FAERS reports but 70% of TD reports were filed by lawyers, suggesting a distortion in signal. Forty-seven legal opinions were identified, 33 from 2009-2010. CONCLUSIONS: The black box warning for metoclopramide has decreased its usage and increased its rate of adverse event reporting. Lawyer-initiated reports of TD hinder pharmacovigilance.

Clinical and endoscopic features of angiosarcoma of the colon: two case reports and a review of the literature.

BACKGROUND: Angiosarcoma of the gastrointestinal tract is an extremely rare tumor. We present two cases of angiosarcoma of the colon. One patient developed syncope and anemia and the other had rectal bleeding. DISCUSSION: Endoscopic findings of the colon were different; the first case demonstrated a reddened colonic fold without an obvious mass and the other had an ulcerated friable mass. Both of the tumors had positive immunohistochemical stains for CD31 and CD34 (one was also positive for CD117). Surgery was not performed in either case. Treatment included chemotherapy with imatinib for the first patient and paclitaxel for the second. Prognosis for this tumor is generally poor; in our cases, one patient was still alive after 5 months of follow-up and the other died within 2 months of diagnosis. After presenting our cases, we reviewed the current literature on angiosarcoma of the colon.

Lubiprostone induced ischemic colitis.

Ischemic colitis accounts for 6%-18% of the causes of acute lower gastrointestinal bleeding. It is often multifactorial and more commonly encountered in the elderly. Several medications have been implicated in the development of colonic ischemia. We report a case of a 54-year old woman who presented with a two-hour history of nausea, vomiting, abdominal pain, and bloody stool. The patient had recently used lubiprostone with close temporal relationship between the increase in the dose and her symptoms of rectal bleeding. The radiologic, colonoscopic and histopathologic findings were all consistent with ischemic colitis. Her condition improved without any serious complications after the cessation of lubiprostone. This is the first reported case of ischemic colitis with a clear relationship with lubiprostone (Naranjo score of 10). Clinical vigilance for ischemic colitis is recommended for patients receiving lubiprostone who are presenting with abdominal pain and rectal bleeding.

Relationship between mesenteric abnormalities on computed tomography and malignancy: clinical findings and outcomes of 359 patients.

BACKGROUND: Mesenteric abnormalities are detected on abdominal computed tomography (CT) performed for various indications. GOALS: Determine the risk of malignancy on follow-up of patients with these abnormalities without a preexisting malignancy. STUDY: Data were collected on all patients at NorthShore University HealthSystem with abdominal CT scan reports of mesenteric abnormalities labeled as "panniculitis" from January 2005 to April 2010. RESULTS: Three hundred fifty-nine patients were identified, 81 (22.6%) had a known malignancy at the time of the index abdominal CT scan. Nineteen (6.8%) of the 278 had a new diagnosis of malignancy on evaluation of the findings of the index CT scan. Among the 240 (86.33%) that did not have a notation of the abnormality in their medical record, 11 (4.58%) developed a malignancy during the study period. Sixty-eight of the 248 (24.46%) without a known malignancy had diseases associated with mesenteric abnormalities. The presence of these were associated with a reduction in the likelihood that the abnormalities are associated with new or delayed diagnosis of a malignancy (odds ratio, 0.197; 95% confidence interval, 0.0045-0.8501; P=0.013). Progression of underlying malignancy was unlikely in those where the mesenteric abnormalities did not worsen in appearance on follow-up CT scans (odds ratio, 0.03268; 95% confidence interval, 0.0028-0.3761; P=0.0061). CONCLUSIONS: In the presence of an underlying disease associated with these findings, the subsequent finding of a malignancy is less likely. In addition, neglect of these findings may result in delayed diagnosis of cancer.

T-cell non-Hodgkin's lymphomas reported to the FDA AERS with tumor necrosis factor-alpha (TNF-α) inhibitors: results of the REFURBISH study.

OBJECTIVES: The risk of non-Hodgkin's lymphoma (NHL) with tumor necrosis factor alpha (TNF-α) inhibitors is unclear, whether related to concomitant thiopurines usage or due to the underlying inflammatory disease. We sought to review all cases of T-cell NHL reported to the Food and Drug Administration (FDA) in patients receiving TNF-α inhibitors for all approved indications and examine the risk of T-cell NHL with TNF-α inhibitors in comparison with the use of thiopurines in inflammatory bowel disease (IBD). METHODS: The FDA Adverse Event Reporting System (AERS) was queried for all lymphomas following treatment with the following TNF-α inhibitors: infliximab, adalimumab, certolizumab, etanercept, and their trade names. Full reports for T-cell NHL cases were identified using the Freedom of Information Act. In addition, T-cell NHL reported in patients IBD with the use of the thiopurines-azathioprine, 6-mercaptopurine, and their trade names were also collected. A search of MEDLINE was performed for additional T-cell NHL with TNF-α inhibitors or thiopurines, not reported to the FDA but available in published literature. The histological subtypes of T-cell NHL reported with TNF-α inhibitors were compared with reported subtypes in Surveillance Epidemiology and End Results (SEER) -17 registry. Reported risk of T-cell NHL in IBD with TNF-α inhibitors, thiopurines, or concomitant use was calculated using Fisher's exact test using 5-aminosalicylates as control drugs. RESULTS: A total of 3,130,267 reports were downloaded from the FDA AERS (2003-2010). Ninety-one cases of T-cell NHL with TNF-α inhibitors were identified in the FDA AERS and nine additional cases were identified on MEDLINE search. A total of 38 patients had rheumatoid arthritis, 36 cases had Crohn's disease, 11 had psoriasis, 9 had ulcerative colitis, and 6 had ankylosing spondylitis. Sixty-eight of the cases (68%) involved exposure to both a TNF-α inhibitor and an immunomodulator (azathioprine, 6-mercaptopurine, methotrexate, leflunomide, or cyclosporine). Hepatosplenic T-cell lymphoma (HSTCL) was the most common reported subtype, whereas mycosis fungoides/Sezary syndrome and HSTCL were identified as more common with TNF-α-inhibitor exposure compared with SEER-17 registry. Nineteen cases of T-cell NHL with thiopurines were identified in the FDA AERS and one additional case on MEDLINE. Reported risk of T-cell NHL was higher with TNF-α inhibitor use in combination with thiopurines (95% confidence interval (CI) 4.98-354.09; P<0.0001) and thiopurines alone (95% CI 8.32-945.38; P<0.0001) but not with TNF-α inhibitor use alone (95% CI 0.13-10.61; P=1.00). CONCLUSIONS: Risk of T-cell NHL is increased with TNF-α inhibitor use in combination with thiopurines but not with TNF-α inhibitors alone.