What Is Known
- Extraintestinal manifestations are frequently observed in adult patients with inflammatory bowel disease.
- Little is known about extraintestinal manifestation in the pediatric inflammatory bowel disease population.
What Is New
- Extraintestinal manifestations are common in pediatric patients with inflammatory bowel disease.
- Epidemiology and presentation of extraintestinal manifestation seem to be similar in the pediatric and adult inflammatory bowel disease population.
- In up to 30% of those patients presenting with extraintestinal manifestation, extraintestinal manifestation appear before inflammatory bowel disease diagnosis.
- Type of extraintestinal manifestation may affect the responses to anti–tumor necrosis factor. Best rates were seen for peripheral arthritis and uveitis.
Inflammatory bowel diseases (IBDs) are chronic inflammatory disorders of the gastrointestinal tract and can be classified into the 2 main subtypes Crohn disease (CD) and ulcerative colitis (UC) (1). As IBDs are systemic diseases, which can involve multiple organ systems, extraintestinal manifestations (EIM) are frequently observed affecting up to 50% of the adult IBD population (2–6). In a non-negligible proportion, EIM appear even before IBD diagnosis is established (7). Although some reviews broaden the concept of EIM to non–IBD-specific autoimmune disorders such as thyroid disease or vitiligo and IBD-related complications such as osteopathy or nephrolithiasis, the typical EIM involve the following 4 organs: skin (erythema nodosum [EN], pyoderma gangrenosum [PG], psoriasis, aphthous stomatitis), joints (peripheral arthritis, axial arthropathy), biliary tract (primary sclerosing cholangitis [PSC]), and eyes (uveitis) (7). Most of EIM parallel intestinal disease activity (8–12), are more common in CD than in UC, and are more frequently observed with longer IBD duration (13). In addition, up to one quarter presents with more than 1 EIM (8). Morbidity and mortality are considerably affected (14,15). Although pathogenesis remains mostly elusive, intestinal and extraintestinal IBDs seem to share tumor necrosis factor (TNF)-dependent mechanisms (16) and several studies and case reports were able to demonstrate beneficial effect of anti-TNF treatments (17–23). EIM (such as arthritis and paradoxical psoriasiform reactions) can, however, also resemble adverse effects of anti-TNF therapy (24).
Although knowledge of prevalence, appearance, and possible treatment options for EIM in adults is increasing, it is still limited in the pediatric IBD population. Recent studies have shown higher rates of EIM at IBD onset compared with adults with peripheral arthritis and aphthous stomatitis being the most prevalent (25,26). To our knowledge, so far no study has addressed the chronological order of EIM appearance in respect to IBD diagnosis or the influence of anti-TNF treatment on the evolution and development of EIM in a pediatric IBD cohort. Given this paucity of data, we aimed to assess the frequency and type of EIM, chronological order of appearance of EIM, and the use of and response to anti-TNF treatment in the Pediatric Swiss IBD Cohort Study (PSIBDCS).
The PSIBDCS is a nationwide substudy of the Swiss IBD Cohort Study (SIBDCS) including all regions of Switzerland (27). Enrollment started in 2008. The SIBDCS and its substudy are supported by the Swiss National Science Foundation and are approved by the local ethics committees of the participating centers. Patients are included in the PSIBDCS if they are 18 years or younger. Additional inclusion criteria and regular assessment scheme (baseline and annual follow-up questionnaires) have been discussed elsewhere (28). Patients were recruited at University Hospitals, community hospitals, and large private practices throughout Switzerland. A total of 329 pediatric patients are currently included. All 329 patients were retrospectively analyzed for the purpose of the present study.
Definition of Extraintestinal Manifestation and Anti–tumor Necrosis Factor Outcome
All EIM had to be diagnosed by clinical experts: diagnosis of skin manifestations was established by a dermatologist, joint affections by a rheumatologist, eye manifestations by an ophthalmologist, and PSC by a gastroenterologist. We analyzed the following EIM: peripheral arthritis, uveitis, PG, EN, aphthous stomatitis, axial arthropathy, psoriasis, and PSC. Diagnosis of EIM relied on previously published criteria (7,25,29). We did not consider anemia, glaucoma, and pancreatic involvement as EIM because it may also be considered as complication of IBD therapy (25). Evolution of EIM under anti-TNF treatment was judged according to the physician's global assessment, which was based on patient history and clinical findings. This anti-TNF response was classified into the following 3 categories: clinical improvement, stable disease course unaffected by anti-TNF treatment, and clinical worsening.
Data Collection and Management
Completed patient and physician questionnaires (baseline and annual follow-up) were sent to the datacenter of the SIBDCS located at the Institute of Social and Preventive Medicine, University of Lausanne, Switzerland, where data were validated by the responsible data manager and finally entered into a Microsoft Access database (Access 2000; Microsoft Switzerland Ltd. Liab. Co., Wallisellen, Switzerland). Baseline questionnaire at enrollment and annual follow-up questionnaires included demographic data, IBD subtype and diagnosis, disease localization according to international guidelines, prior and current medications, and past and present EIM. Electronic and written charts of those patients, who reported past or present EIM, were additionally reviewed by clinical experts to extract exact appearance of EIM in relation to IBD diagnosis, EIM subtype, chronological order of appearance of multiple EIM, and evolution of EIM in response to anti-TNF treatment. All 329 patients in the PSIBDCS were eligible. For detailed analysis, only those patients with 1 or more EIM were included. For a comparison between the pediatric and adult population, data from the SIBDCS were used according to a recent study conducted by Vavricka et al (7).
For all statistical analyses, IBM software SPSS version 23.0.0 (2014 SPSS Science, Inc., Armonk, NY) was used. Metric data are shown as medians with their total range. Categorical data are depicted as percentage of the group total. Comparison between categorical variables was performed by using Chi-square test or Fisher exact test, if sample size was low (n < 10). A 2-sided P value of <0.05 was considered statistically significant. The association between potential predicting factors and positive anti-TNF outcome was analyzed by means of logistic regression.
Of the 329 pediatric patients with IBD included in the study, 173 (52.6%) had CD and 156 (47.4%) had UC/IBD-unclassified (IBD-U), 148 patients were girls (45.0%). Median age at enrollment was 14 years (0–17) and median age at IBD diagnosis was 12 years (0–16). The patients had been experiencing IBD for a median of 3 years (0–16). Patient demographics according to IBD subtype (including disease localization and received medications) are summarized in Table 1. Of the 329 patients, 55 presented with 1 or more EIM (16.7%). The characteristics of those patients are depicted in Table 2.
Frequency and Types of Extraintestinal Manifestation
Of the 55 patients with at least 1 EIM, 39 patients had CD (39/55, 70.9%), 12 patients had UC (12/55, 21.8%), and 4 patients had IBD-U (4/55, 7.3%) as their underlying condition. Of these 55 patients with EIM, 39 (70.9%), 12 (21.8%), 3 (5.5%), and 1 patient (1.8%) reported 1, 2, 3, and 4 EIM, respectively. At IBD onset, presence of EIM was more frequent than that in the adult population (28/329, 8.5% vs 62/1249, 5.0%, P = 0.014). EIM were more frequently observed in patients with CD (39/173, 22.5%) when compared to patients with UC/IBD-U (16/156, 10.3%, P = 0.003). The most prevalent EIM were peripheral arthritis (26/329, 7.9%) and aphthous stomatitis (24/329, 7.3%), followed by uveitis (6/329, 1.8%), EN (5/329, 1.5%), axial arthropathy (5/329, 1.5%), psoriasis (4/329, 1.2%), PSC (4/329, 1.2%), and PG (2/329, 0.6%). Peripheral arthritis, axial arthropathy, and EN were less frequently encountered among pediatric patients with IBD compared to adult patients (7.9% vs 20.5%, P < 0.001; 1.5% vs 4.8%, P = 0.008; and 1.5% vs 3.7%, P = 0.048). Frequency of other EIM was comparable between the 2 populations. For a detailed overview see Figure 1. Although 27.6% of all EIM (21/76) appeared before the diagnosis of IBD, the majority of EIM appeared once IBD diagnosis was established (42/76, 55.3%). The remaining proportion of EIM (12/76, 15.8%) was first observed at the time of establishment of IBD diagnosis. Data from 1 EIM were missing. So, at the time of IBD diagnosis, 28 of the 55 patients presented with a total of 33 EIM (observed before or at the time of IBD diagnosis); aphthous stomatitis and peripheral arthritis were again the most frequently observed EIM (15/28, 53.6% and 10/28, 35.7%, respectively). For a synopsis over the phenotypic features at IBD diagnosis see Supplemental Digital Content 1, Table 1 (http://links.lww.com/MPG/A827). Details about frequency and type of EIM according to IBD subtype are shown in Table 2.
Distribution of Different Extraintestinal Manifestation According to Their Chronological Appearance
Aphthous stomatitis was the most prevalent EIM (21/55, 38.2%), which appeared as first EIM (n = 55), followed by peripheral arthritis (19/55, 34.5%) and uveitis (4/55, 7.3%). If the patient was diagnosed with a second EIM (n = 16), peripheral arthritis (5/16, 31.3%), EN (3/16, 18.8%), aphthous stomatitis (2/16, 12.5%), and axial arthropathy (2/16, 12.5%) were most frequent. In those patients presenting with a third EIM (n = 4), occurrence of the following EIM were reported: peripheral arthritis (1/4, 25.0%), uveitis (1/4, 25.0%), aphthous stomatitis (1/4, 25.0%), and psoriasis (1/4, 25.0%). Median time between IBD diagnosis and occurrence of first EIM was 1 month (range –37.5–149.0). Figure 2 illustrates the chronological order of appearance of each individual EIM in relation to the time of IBD diagnosis (in months). Peripheral arthritis appeared before IBD diagnosis in 28.0% (7/25, exact appearance in 1 case unknown), uveitis in 16.7% (1/6), EN in 20% (1/5), axial arthropathy in 40.0% (2/5), and aphthous stomatitis in 29.2% (7/24). All cases of psoriasis (4/4, 100%) and PG (2/2, 100%) appeared after IBD diagnosis were established. In addition, peripheral arthritis (64.0% vs 28.0%, P = 0.011) and uveitis (83.3% vs 16.7%, P = 0.021) were significantly more likely to appear after diagnosis of IBD than before establishment of IBD diagnosis. Median lag of time of appearance before IBD diagnosis in the group of patients in whom EIM preceded IBD diagnosis was −5.0 months (range −37.5 to −0.4).
Type of Anti–tumor Necrosis Factor Treatment and Treated Extraintestinal Manifestation
Anti-TNF therapy was initiated in 31 of the 55 patients with EIM (56.4%). Most of them were treated with a single anti-TNF agent (23/31, 74.2%), whereas 5 patients were treated with 2 and 3 patients with 3 different anti-TNF agents (5/31, 16.1% and 3/31, 9.7%, respectively). So, a total of 42 treatment courses were initiated. In 78.6% (33/42), anti-TNF treatment was started for the purpose of treating underlying IBD activity. In 3 cases (3/42, 7.1%), anti-TNF was solely initiated for the purpose of treating EIM; infliximab was started for peripheral arthritis (n = 1) and axial arthropathy (n = 1); and adalimumab was initiated for peripheral arthritis (n = 1). In 5 cases (5/42, 11.9%), anti-TNF was started for the purpose of treating both intestinal disease activity and EIM. In 1 case, exact indication for anti-TNF treatment was not known. IBD patients presenting with EIM were more likely to be treated with anti-TNF compared to those without EIM (31/55, 56.4% vs 96/274, 35.0%, P = 0.003). The most frequently treated EIM were peripheral arthritis (16/42, 38.1%) and aphthous stomatitis (13/42, 31.0%), followed by axial arthropathy, uveitis, and EN. For details see Table 3. Under anti-TNF treatment, 23 EIM appeared in 19 of the 31 treated patients (19/31, 61.3%). Among those 23 EIM, peripheral arthritis (6 cases) and aphthous stomatitis (5 cases) were the most frequently reported. Three cases of psoriasis occurred under anti-TNF therapy, which can be considered as anti-TNF-induced psoriasiform skin lesions. Further reported EIM, which appeared under anti-TNF, were: uveitis (2 cases), PG (1 case), EN (2 cases), and axial arthropathy (1 case). In remaining 3 cases, the exact EIM subtype was unknown.
Clinical Evolution of Extraintestinal Manifestation Under Anti–tumor Necrosis Factor Treatment
Data on clinical outcome of anti-TNF treatment were available for 37 of the 53 treated EIM (69.8%). In the majority, EIM showed improvement (17/37, 45.9%) or stable disease course (13/37, 35.1%), whereas clinical worsening was observed in only 7 cases (7/37, 18.9%). Peripheral arthritis and uveitis showed good clinical response rates to anti-TNF (61.5% and 66.7%, respectively), whereas those of PG, axial arthropathy, aphthous stomatitis, and EN were ≤50% (50.0%, 50.0%, 33.3%, and 33.3%, respectively, Table 3). In a multivariate regression model adjusted for age and sex, appearance of EIM before IBD diagnosis was the only independent predictor for positive anti-TNF outcome (odds ratio 9.70, 95% confidence interval 1.04–90.04, P = 0.046). Details about the multivariate analysis can be found in the Supplemental Digital Content 2, Table 2 (http://links.lww.com/MPG/A828).
In this analysis of the PSIBDCS, we report on frequency and chronological order of appearance of EIM, the use of anti-TNF agents, and the disease course under such therapy. EIM are common in pediatric patients with IBD and can appear in nearly 30% before IBD diagnosis is established. Anti-TNF are used frequently in those patients, although they are started for the purpose of treating EIM in only a minority. Type of EIM may affect anti-TNF outcome. Best response rates were seen for peripheral arthritis and uveitis.
EIM were frequently encountered in pediatric IBD, the prevalence of 16.7% is comparable to prior data from Guariso et al (26). Prevalence is, however, considerably lower compared to the studies conducted by Dotson et al (29) and Jose et al (25), which is mainly due to our more stringent definition of EIM, because we did not include nonspecific arthralgia or other nonspecific EIM such as anemia, hepatitis, pancreatitis, or osteoporosis. In accordance to the findings of Guariso et al (26), presence of EIM at IBD onset was more frequent in the PSIBDCS compared to the adult population (8.5% vs 5%), although difference was considerably smaller (14.3% vs 7.3% in the study of Guariso). Our numbers fit well within the range of 6% to 47% EIM prevalence typically reported (2,4–6,13,30). Our finding that EIM were more frequently observed in CD compared with UC patients is consistent with prior studies from adult IBD cohorts (7,8,13). Moreover, the order of frequency is consistent throughout the adult EIM literature and our data are in agreement with published observations: musculoskeletal symptoms (such as peripheral arthritis) are followed by stomatitis, ophthalmological problems (such as uveitis), and skin changes (8,25). Although the proportion of patients presenting with arthritis is higher among adult patients, order of frequency of EIM did not show a difference between the adult and pediatric patients with IBD. EIM were more likely to appear after IBD diagnosis compared to before. Nonetheless, a non-negligible proportion (up to 30%) of EIM appeared before IBD diagnosis was established. Axial arthropathy appeared before IBD diagnosis in an even higher proportion. Both findings are consistent with data from adult cohorts (7). So, clinicians should be aware of those EIM manifesting before intestinal symptoms to decrease the diagnostic delay. Our group has recently shown that such diagnostic delay is a concern (31) and that a delayed diagnosis is associated with a complicated disease course in patients with CD (32). Appearance of >1 EIM was observed only infrequently. Thus, the fact that 1 EIM seems to increase the susceptibility of developing other EIM does not seem to be true for pediatric patients (33) or latency between first and second EIM may have been too long to allow detection in children. Taken together, our data suggest that epidemiology and presentation of EIM is quite similar in the pediatric and the adult population, suggesting similar disease mechanisms.
Anti-TNF were frequently used in pediatric patients with EIM. Those agents were, however, specifically initiated for treating EIM in only a minority. So, clinical practice seems to be in accordance to current guidelines, which recommend to treat underlying IBD activity rather than EIM themselves. Nonetheless, patients with EIM were significantly more often treated with anti-TNF than those without EIM. The latter may be explained by prior findings from Vavricka et al (8) who showed that active disease is an independent risk factor for EIM in both UC and CD. Upon anti-TNF treatment, EIM showed overall response rates of nearly 50%, which depended on the underlying EIM with the best rates for peripheral arthritis and uveitis. Although direct comparisons are not possible, higher response rates to anti-TNF treatment were described in the adult population (71.8%) (34). New onset of EIM under anti-TNF treatment also has been encountered in our study. Diagnosis of psoriasis was established in 3 of the 31 anti–TNF-treated patients, which can be interpreted retrospectively as anti–TNF-induced psoriasiform skin lesion. Prevalence seems to be in accordance to that reported in the literature (35).
Our study has several strengths and some limitations as well. So far, it is one of the largest analyses of collected data evaluating frequency and occurrence of EIM in the pediatric IBD population. Furthermore, a detailed chart review revealed information about chronological appearance of EIM according to EIM subtype, the use of anti-TNF treatment and clinical response to such a therapy. The combination of physician- and patient-based questionnaire may have prevented the underreporting of EIM in our study population. We, however, used an uncontrolled, noninterventional study design, which limits interpretation of anti-TNF treatment outcome. With annual follow-up visits, important details occurring during this period of time may have been missed. This potential recall bias has, however, been limited by the combination of physician- and patient-based questionnaires. Given the fact that the Pediatric Swiss IBD cohort is not strictly population based, a selection bias may be present. For instance, patients included by private practices and community hospitals were underrepresented compared with those patients included by university hospitals (30.9% of the patients presenting with EIM were recruited in private practices or community hospitals, whereas 69.1% were recruited by physicians working in a university hospital). A clear limitation of our analysis is that—at the time of data analysis—we did not clearly differentiate between psoriasis and psoriasiform anti–TNF-associated skin lesions, which is an increasingly reported phenomenon (35). With 4 cases, prevalence of psoriasis was, however, extremely low, and 3 out of 4 cases can be retrospectively considered as anti-TNF induced as they presented under anti-TNF treatment for the first time. Because of the retrospective nature of the study, a non-negligible proportion of patients received concomitant immunomodulators and/or steroids in addition to anti-TNF, which may have led to an overestimation of anti-TNF response rates.
In summary, in a cohort of pediatric patients with IBD, EIM were frequently encountered. In up to 30% of patients, EIM appeared before IBD was diagnosed. Knowledge of these findings may result in increased awareness of underlying IBD, thereby decreasing potential diagnostic delay. Anti-TNF for the treatment of certain EIM is effective, although a substantial proportion of new EIM may present despite ongoing anti-TNF therapy. Randomized controlled trials are, however, needed in the pediatric IBD population.
The authors thank all patients and the staff of the SIBDCS for their commitment.
1. Abraham C, Cho JH. Inflammatory bowel disease. N Engl J Med
2. Bernstein CN, Blanchard JF, Rawsthorne P, et al. The prevalence of extraintestinal diseases in inflammatory bowel disease: a population-based study. Am J Gastroenterol
3. Bernstein CN, Wajda A, Blanchard JF. The clustering of other chronic inflammatory diseases in inflammatory bowel disease: a population-based study. Gastroenterology
4. Mendoza JL, Lana R, Taxonera C, et al. Extraintestinal manifestations in inflammatory bowel disease: differences between Crohn's disease and ulcerative colitis. Med Clin (Barc)
5. Ricart E, Panaccione R, Loftus EV, et al. Autoimmune disorders and extraintestinal manifestations in first-degree familial and sporadic inflammatory bowel disease: a case-control study. Inflamm Bowel Dis
6. Rankin GB, Watts HD, Melnyk CS, et al. National Cooperative Crohn's Disease Study: extraintestinal manifestations and perianal complications. Gastroenterology
1979; 77 (4 pt 2):914–920.
7. Vavricka SR, Rogler G, Gantenbein C, et al. Chronological order of appearance of extraintestinal manifestations relative to the time of IBD diagnosis in the Swiss Inflammatory Bowel Disease Cohort. Inflamm Bowel Dis
8. Vavricka SR, Brun L, Ballabeni P, et al. Frequency and risk factors for extraintestinal manifestations in the Swiss Inflammatory Bowel Disease Cohort. Am J Gastroenterol
9. Trost LB, McDonnell JK. Important cutaneous manifestations of inflammatory bowel disease. Postgrad Med J
10. Barrie A, Regueiro M. Biologic therapy in the management of extraintestinal manifestations of inflammatory bowel disease. Inflamm Bowel Dis
11. Levitt MD, Ritchie JK, Lennard-Jones JE, et al. Pyoderma gangrenosum in inflammatory bowel disease. Br J Surg
12. Mir-Madjlessi SH, Taylor JS, Farmer RG. Clinical course and evolution of erythema nodosum and pyoderma gangrenosum in chronic ulcerative colitis: a study of 42 patients. Am J Gastroenterol
13. Veloso FT, Carvalho J, Magro F. Immune-related systemic manifestations of inflammatory bowel disease. A prospective study of 792 patients. J Clin Gastroenterol
14. Das KM. Relationship of extraintestinal involvements in inflammatory bowel disease: new insights into autoimmune pathogenesis. Dig Dis Sci
15. Monsén U, Sorstad J, Hellers G, et al. Extracolonic diagnoses in ulcerative colitis: an epidemiological study. Am J Gastroenterol
16. Vavricka SR, Scharl M, Gubler M, et al. Biologics for extraintestinal manifestations of IBD. Curr Drug Targets
17. Caspersen S, Elkjaer M, Riis L, et al. Infliximab for inflammatory bowel disease in Denmark 1999–2005: clinical outcome and follow-up evaluation of malignancy and mortality. Clin Gastroenterol Hepatol
2008; 6:1212–1217. quiz 176.
18. Generini S, Giacomelli R, Fedi R, et al. Infliximab in spondyloarthropathy associated with Crohn's disease: an open study on the efficacy of inducing and maintaining remission of musculoskeletal and gut manifestations. Ann Rheum Dis
19. Kaufman I, Caspi D, Yeshurun D, et al. The effect of infliximab on extraintestinal manifestations of Crohn's disease. Rheumatol Int
20. Kugathasan S, Miranda A, Nocton J, et al. Dermatologic manifestations of Crohn disease in children: response to infliximab. J Pediatr Gastroenterol Nutr
21. Rispo A, Scarpa R, Di Girolamo E, et al. Infliximab in the treatment of extra-intestinal manifestations of Crohn's disease. Scand J Rheumatol
22. Barreiro-de-Acosta M, Lorenzo A, Domínguez-Muñoz JE. Efficacy of adalimumab for the treatment of extraintestinal manifestations of Crohn's disease. Rev Esp Enferm Dig
23. Herfarth H, Obermeier F, Andus T, et al. Improvement of arthritis and arthralgia after treatment with infliximab (Remicade) in a German prospective, open-label, multicenter trial in refractory Crohn's disease. Am J Gastroenterol
24. Vavricka SR, Schoepfer A, Scharl M, et al. Extraintestinal manifestations of inflammatory bowel disease. Inflamm Bowel Dis
25. Jose FA, Garnett EA, Vittinghoff E, et al. Development of extraintestinal manifestations in pediatric patients with inflammatory bowel disease. Inflamm Bowel Dis
26. Guariso G, Gasparetto M, Visonà Dalla Pozza L, et al. Inflammatory bowel disease developing in paediatric and adult age. J Pediatr Gastroenterol Nutr
27. Levine A, Koletzko S, Turner D, et al. ESPGHAN revised porto criteria for the diagnosis of inflammatory bowel disease in children and adolescents. J Pediatr Gastroenterol Nutr
28. Pittet V, Juillerat P, Mottet C, et al. Cohort profile: the Swiss Inflammatory Bowel Disease Cohort Study (SIBDCS). Int J Epidemiol
29. Dotson JL, Hyams JS, Markowitz J, et al. Extraintestinal manifestations of pediatric inflammatory bowel disease and their relation to disease type and severity. J Pediatr Gastroenterol Nutr
30. Su CG, Judge TA, Lichtenstein GR. Extraintestinal manifestations of inflammatory bowel disease. Gastroenterol Clin North Am
31. Vavricka SR, Spigaglia SM, Rogler G, et al. Systematic evaluation of risk factors for diagnostic delay in inflammatory bowel disease. Inflamm Bowel Dis
32. Schoepfer AM, Dehlavi MA, Fournier N, et al. Diagnostic delay in Crohn's disease is associated with a complicated disease course and increased operation rate. Am J Gastroenterol
33. Ardizzone S, Puttini PS, Cassinotti A, et al. Extraintestinal manifestations of inflammatory bowel disease. Dig Liver Dis
2008; 40 (suppl 2):S253–S259.
34. Vavricka SR, Biedermann L, Rogler G, et al. Anti-TNF treatment for extraintestinal manifestations of inflammatory bowel disease in the Swiss IBD Cohort Study. Gastroenterology
35. Fiorino G, Danese S, Pariente B, et al. Paradoxical immune-mediated inflammation in inflammatory bowel disease patients receiving anti-TNF-α agents. Autoimmun Rev