Anti-tumor necrosis factor (anti-TNF) therapy has been shown to be effective and safe for treatment of pediatric inflammatory bowel disease (IBD) (1–4). One of the key cytokines in the immune response against tuberculosis (TB) infection is TNF (5). Individuals treated with anti-TNF therapy have an increased risk of reactivation of latent TB infection (LTBI) (6); however, there is no publication showing these observations in children with IBD. Both TB skin test (TST) and interferon-γ release assay (IGRA) have been recommended for LTBI screening before initiation of anti-TNF therapy in children (7). The risk of TB in individuals undergoing anti-TNF therapies can be dramatically reduced by 9 months of isoniazid (INH) preventive therapy (8). Children with IBD often require immunosuppressive therapy including corticosteroids, which may influence the results of LTBI testing in patients with IBD (9). In addition, malnutrition, one of the prominent features of pediatric IBD (10), extensively deteriorates immune function (11,12) and thus may be another factor influencing these results. Significant intraindividual IGRA variability and definition of conversions and reversions have been recently reported (13). Screening for LTBI is obligatory before the treatment is started (14). A few data have been published describing continual monitoring of patients treated with anti-TNF (15,16). Data on regular monitoring during therapy and factors influencing IGRA results are scarce, especially in children.
The aim of the study was to determine factors influencing the results of IGRA in pediatric patients with IBD before and during anti-TNF therapy.
All of the pediatric patients with IBD treated in University Hospital Motol by anti-TNF therapy between the years 2008 and 2012 were consecutively retested using IGRA in approximately 1-year interval. Demographic and clinical characteristics of patients according to the Montreal classification (17) are described in Table 1. Patients who started anti-TNF treatment before April 2008 had been initially tested by “PCR TB” in blood and urine, x-ray imaging, personal history, and TST. All of the children who started anti-TNF treatment after April 2008 were tested for TB by IGRA, x-ray imaging, personal history, and TST. The mean follow-up of 1-year retesting was 3 years (1–5 years). All of the included participants (72 patients) were previously vaccinated against Bacillus Calmette–Guérin in infancy. None of the patients was an immigrant or had reported a contact with a high-risk population with TB. The patients were diagnosed as having IBD based on the Porto criteria (59 Crohn disease [CD], 11 ulcerative colitis [UC], 2 IBD-unclassified [IBD-U]) (18). The disease activity has been assessed by the Pediatric Crohn's Disease Activity Index (19) or the Pediatric Ulcerative Colitis Activity Index (20) and by routine laboratory parameters on the same date as blood for IGRA was collected. The nutritional status was assessed using the standard deviation of weight-for-height ratio (z score) and height-for-age ratio during the visit. All of the medications have been recorded. At least 2 months of treatment before IGRA collection, with azathioprine and 6-mercaptopurine at doses more than 1 and 0.5 mg · kg−1 · day−1, respectively, was assessed as considerable. Corticosteroid use was assessed as relevant when given at least 1 week before IGRA collection in doses more than 0.5 mg · kg−1 · day−1. Fifty-nine percent (98/165), 11% (18/165), and 6% (10/165) of the patients were tested with IGRA while taking immunomodulators, corticosteroids, and combined immunomodulators and corticosteroid therapy, respectively. Among 72 included patients, 70 were treated with infliximab and 2 with adalimumab.
The QuantiFERON—TB Gold in tube (QFT, Cellestis, Valencia, CA) test was performed at the Department of Clinical Microbiology, Na Bulovce Hospital and at the National Reference Laboratory for Mycobacteria. The blood samples for QFT were gathered into sample tubes and transported to the laboratory where they were processed immediately. The QFT test was performed according to the manufacturer's instructions. The threshold value of TB antigen minus NIL for discrimination between positivity and negativity was set to 0.35 IU/mL, according to the manufacturer's instructions. The interpretation was done by software supplied by the manufacturer using a complete algorithm for interpretation.
Conversion and Reversion
IGRA conversion was defined as a positive result under anti-TNF therapy after a negative baseline test. IGRA reversion was defined as a negative test result subsequent to a positive result in previous tests.
The data analysis was performed using the R software program (version 3.0.1). We used the Mann-Whitney U test for comparison of the 2 groups. Association of treatment modalities and activity of disease with indeterminate results were tested using the Fisher exact test and described by odds ratios (OR) with 95% confidence intervals (95% CI). The linear mix model was constructed using R-project package lmerTest (version 1.2-1). For analysis of factors that may influence IGRA, all of the obtained results were used (collected before initiating and during anti-TNF treatment).
The study was approved by the ethics committee of the authors’ institution, and written informed consent was obtained from children and their parents.
During the observation interval, 165 results of IGRA in 72 patients were recorded. Among these tests, 5 (3%) were positive, 144 (87%) negative, and 16 (10%) indeterminate.
None of the 72 patients had positive x-ray imaging, personal history, or TST before anti-TNF treatment. Before initiation of anti-TNF treatment, IGRA had been tested in 46 patients. Positivity of IGRA had been documented before anti-TNF treatment in 1 patient who subsequently received INH and had postponed anti-TNF for 2 months. No evidence of active TB was observed. We found 8 indeterminate IGRA results before anti-TNF treatment had begun. These patients were assessed by a pulmonologist (x-ray imaging, TST, family and personal history were negative) and anti-TNF treatment was initiated. Consequently, they were closely followed up by a pulmonologist for approximately 3 months.
During follow-up, we performed 119 IGRA with 107 (90%) negative, 4 (3%) positive, and 8 (7%) indeterminate results. We identified conversion of negative IGRA to positive in 4 patients (3%), and reversion in 5 patients (4%), all with CD.
All of the patients with positive IGRA had temporarily discontinued biological treatment. They received INH and were followed up by a pulmonologist for approximately 3 months, with no signs of TB on chest x-ray and history/physical examination. One patient experienced a relapse during the discontinuation of anti-TNF treatment. In all of the patients, anti-TNF therapy was restarted after the discontinuation period. All of them subsequently had a negative QFT test during the follow-up.
Factors Influencing Indeterminate IGRA Results
Patients at the time of IGRA testing with indeterminate results had significantly lower actual weight-for-height z score (P = 0.022), higher platelet count (P = 0.00017), and lower level of serum albumin (P = 0.015) compared with patients with positive or negative IGRA at the time of testing. The height-for-age z score did not differ significantly between these groups (P = 0.24). We observed profound differences in the activity of CD (expressed by Pediatric Crohn's Disease Activity Index [PCDAI]) in patients with indeterminate results compared to patients with positive or negative IGRA (P = 0.000043) (Fig. 1A). The difference between Pediatric Ulcerative Colitis Activity Index (PUCAI) results was not significant in patients with determinate and indeterminate IGRA results (Fig. 1B).
When tested in linear mixed-model indeterminate, IGRA results were associated with corticosteroid treatment (P = 0.00028), anti-TNF therapy (P = 0.038), and disease activity of IBD (PCDAI > 10 or PUCAI > 10) (P = 0.014), but not with immunomodulators (Table 2). In a subanalysis of factors influencing the indeterminate results of IGRA in patients with CD only, all except 1 predictor (immunomodulators) were associated with the linear mixed model (corticosteroids with P = 0.011, anti-TNF therapy with P = 0.046, PCDAI with P = 1.13 × 10−7, platelet count with P = 0.00022, albumin with P = 0.0012, weight-to-height with P = 0.018). PCDAI was, however, identified as a single independent risk factor for indeterminate results in the mixed model adjusted for predictors, with P < 0.1 in the unadjusted model. (The model was constructed using corticosteroids, anti-TNF therapy, PCDAI, platelet count, serum albumin, and weight-for-height z score as a predictor and indeterminate IGRA as outcome.)
According to European Crohn's and Colitis Organisation guidelines, screening for LTBI must be performed before starting anti-TNF treatment (21). Both IGRA and TST have been recommended for LTBI screening before anti-TNF therapy is initiated in children (7). Adult patients with IBD treated with anti-TNF therapy have an increased risk of reactivation of LTBI (6); however, there is no publication showing these observations in children. In the literature, there is a note on the development of active TB in 2 children with IBD, from a personal communication of the authors (22). No data have been published on regular monitoring of IGRA in patients with IBD with anti-TNF treatment. Recently, a study from Turkey (the risk of TB is considered moderate and Bacillus Calmette–Guérin vaccination is included in the country's immunization schedule) has shown the importance of testing for LTBI in children with juvenile idiopathic arthritis (23). Personal history, physical examination, TST, and chest x-ray were performed every 6 months in that study. During follow-up, 4.8% of patients had positive TST and were given INH prophylaxis, but TB symptoms were not evident in these patients (23).
As previously published, in adults, the indeterminate results of IGRA represent a clinically relevant limitation in the management of patients with IBD in whom treatment with anti-TNF agents is indicated (9,24). For the first time, the present study shows the effect of actual activity of IBD and anti-inflammatory agents on the outcome of testing for LTBI in children. We confirmed a previous observation in adults showing that corticosteroid treatment increases the risk for indeterminate IGRA results (9,15). The present study shows 10% of indeterminate IGRA results with a substantial proportion of patients on immunosuppressive therapy—corticosteroids (12%) and immunomodulators (60%), respectively. The rates of indeterminate IGRA results and the usage of therapy in recently published studies are shown in Table 3(9,16,24–26). Higher rate of indeterminate IGRA results published by Helwig et al (9) corresponded with the high percentage of patients undergoing corticosteroid therapy; however, the numbers of tests in that study are relatively small. On the contrary, studies with high rates of usage of immunomodulators or anti-TNF therapy had relatively low numbers of indeterminate results (16,25). This also suggests that corticosteroids may play a more important role than immunomodulators and anti-TNF therapy. The rate of indeterminate IGRA results seems to be much higher in the present study than published in the meta-analysis for immunosuppressed individuals (4.4%) (27). Corticosteroids were widely used for induction of remission when disease activity was high. The effect of disease activity has not been sufficiently studied in previously published articles. Papay et al tested some laboratory parameters (total serum protein, serum albumin, C-reactive protein, white blood cell count, lymphocyte count, CD4+ cell count) and found hypoalbuminemia and low lymphocyte count as independent predictors for indeterminate results (24). Clinically relevant data have not been tested. In the present study, we found the activity of disease expressed as PCDAI as the only independent risk factor for indeterminate results in children with CD. Platelet count has been shown to be a marker of activity of IBD (28,29). We found that children with indeterminate results of IGRA have significantly higher platelet counts; however, when tested only in patients with CD and adjusted for PCDAI, only PCDAI was associated independently, showing that the level of platelet count is not sufficient to describe disease activity in pediatric CD. Moreover, malnutrition as immunocompromising status (11,12) accompanying relapse may play an important role in influencing IGRA results. The present study shows difference of nutritional status expressed as z score weight-for-height between patients with indeterminate and determinate results of IGRA; however, when tested in the logistic regression model, the z score of weight-for-height was not an independent risk factor. Serum albumin level has been used as a marker of malnutrition (30) and as an acute-phase reactant. Albumin is part of the PCDAI score, and thus it is not surprising that the association of serum albumin level with indeterminate IGRA is dependent on PCDAI.
High disease activity is the most profound risk factor for indeterminate IGRA results, and it is more important than corticosteroid treatment in pediatric patients with CD. Immunomodulators do not influence the results of this test. Based on these observations, modification of the recommendations on when to test for LTBI will be problematic. Because exclusive enteral nutrition has become widely used as a first-line therapy for children with CD in Europe, and azathioprine does not seem to be a risk factor for indeterminate results, a possible solution for the prevention of these situations would be testing for LTBI when first remission is achieved.
There were several limitations to our study. First, the number of pediatric patients treated by anti-TNF therapy is generally lower compared with adult patients. Accordingly, the number of included pediatric patients in our study was lower when compared with studies in adults and the numbers are extremely low to draw any firm conclusions. Hence, replication of these results, preferably in larger cohorts, is necessary. Because the disease activity of CD and UC must be judged separately (PCDAI and PUCAI), an appropriate cutoff for PCDAI and PUCAI must be established when evaluating disease activity in the whole group, and categorical data must be conducted. Because this may cause potential bias, the factors that may influence IGRA results in patients with CD were tested separately; however, the same could not be done for patients with UC owing to the low number of patients in this group. The present study does not have enough power to detect the effect of treatment modality or disease activity on conversion or reversion rates. A multicenter study with a large number of children treated with anti-TNF therapy would be necessary to identify these factors, with the potential to reduce prophylactic treatment with INH.
This is the first study describing regular monitoring by IGRA in patients receiving anti-TNF treatment and testing predictive factors for indeterminate IGRA results in pediatric patients with IBD. Although corticosteroid treatment was traditionally considered the main risk factor for indeterminate results of IGRA in adults, we identified a new important predictor of indeterminate IGRA: disease activity of pediatric IBD. The PCDAI seems to be the most important predictor for indeterminate results in patients with CD.
1. Rosh JR, Lerer T, Markowitz J, et al. Retrospective Evaluation of the Safety and Effect of Adalimumab Therapy (RESEAT) in pediatric Crohn's disease. Am J Gastroenterol
2. Hyams J, Crandall W, Kugathasan S, et al. Induction and maintenance infliximab therapy for the treatment of moderate-to-severe Crohn's disease in children. Gastroenterology
3. Hyams JS, Griffiths A, Markowitz J, et al. Safety and efficacy of adalimumab for moderate to severe Crohn's disease in children. Gastroenterology
4. Russell RK, Wilson ML, Loganathan S, et al. A British Society of Paediatric Gastroenterology, Hepatology and Nutrition survey of the effectiveness and safety of adalimumab in children with inflammatory bowel disease. Aliment Pharmacol Ther
5. Roach DR, Bean AG, Demangel C, et al. TNF regulates chemokine induction essential for cell recruitment, granuloma formation, and clearance of mycobacterial infection. J Immunol
6. Keane J, Gershon S, Wise RP, et al. Tuberculosis associated with infliximab, a tumor necrosis factor alpha-neutralizing agent. N Engl J Med
7. Solovic I, Sester M, Gomez-Reino JJ, et al. The risk of tuberculosis related to tumour necrosis factor antagonist therapies: a TBNET consensus statement. Eur Respir J
8. Gomez-Reino JJ, Carmona L, Angel Descalzo M. Risk of tuberculosis in patients treated with tumor necrosis factor antagonists due to incomplete prevention of reactivation of latent infection. Arthritis Rheum
9. Helwig U, Muller M, Hedderich J, et al. Corticosteroids and immunosuppressive therapy influence the result of QuantiFERON TB Gold testing in inflammatory bowel disease patients. J Crohns Colitis
10. Hartman C, Eliakim R, Shamir R. Nutritional status and nutritional therapy in inflammatory bowel diseases. World J Gastroenterol
11. Cunningham-Rundles S, McNeeley DF, Moon A. Mechanisms of nutrient modulation of the immune response. J Allergy Clin Immunol
12. Schaible UE, Kaufmann SH. Malnutrition and infection: complex mechanisms and global impacts. PLoS Med
13. van Zyl-Smit RN, Zwerling A, Dheda K, et al. Within-subject variability of interferon-g assay results for tuberculosis and boosting effect of tuberculin skin testing: a systematic review. PLoS ONE
14. Travis SP, Stange EF, Lemann M, et al. European evidence based consensus on the diagnosis and management of Crohn's disease: current management. Gut
2006; 55 (suppl 1):i16–i35.
15. Papay P, Primas C, Eser A, et al. Retesting for latent tuberculosis in patients with inflammatory bowel disease treated with TNF-alpha inhibitors. Aliment Pharmacol Ther
16. Qumseya BJ, Ananthakrishnan AN, Skaros S, et al. QuantiFERON TB gold testing for tuberculosis screening in an inflammatory bowel disease cohort in the United States. Inflamm Bowel Dis
17. Silverberg MS, Satsangi J, Ahmad T, et al. Toward an integrated clinical, molecular and serological classification of inflammatory bowel disease: report of a Working Party of the 2005 Montreal World Congress of Gastroenterology. Can J Gastroenterol
2005; 19 (suppl A):5–36.
18. IBD Working Group of the European Society for Paediatric Gastroenterology, Hepatology and NutritionInflammatory bowel disease in children and adolescents: recommendations for diagnosis—the Porto criteria. J Pediatr Gastroenterol Nutr
19. Hyams JS, Ferry GD, Mandel FS, et al. Development and validation of a pediatric Crohn's disease activity index. J Pediatr Gastroenterol Nutr
20. Turner D, Otley AR, Mack D, et al. Development, validation, and evaluation of a pediatric ulcerative colitis activity index: a prospective multicenter study. Gastroenterology
21. Rahier JF, Ben-Horin S, Chowers Y, et al. European evidence-based consensus on the prevention, diagnosis and management of opportunistic infections in inflammatory bowel disease. J Crohns Colitis
22. Veereman-Wauters G, de Ridder L, Veres G, et al. Risk of infection and prevention in pediatric patients with IBD: ESPGHAN IBD Porto Group Commentary. J Pediatr Gastroenterol Nutr
23. Kilic O, Kasapcopur O, Camcioglu Y, et al. Is it safe to use anti-TNF-alpha agents for tuberculosis in children suffering with chronic rheumatic disease? Rheumatol Int
24. Papay P, Eser A, Winkler S, et al. Predictors of indeterminate IFN-gamma release assay in screening for latent TB in inflammatory bowel diseases. Eur J Clin Invest
25. Bélard E, Semb S, Ruhwald M, et al. Prednisolone treatment affects the performance of the QuantiFERON gold in-tube test and the tuberculin skin test in patients with autoimmune disorders screened for latent tuberculosis infection. Inflamm Bowel Dis
26. Schoepfer AM, Flogerzi B, Fallegger S, et al. Comparison of interferon-gamma release assay versus tuberculin skin test for tuberculosis screening in inflammatory bowel disease. Am J Gastroenterol
27. Diel R, Loddenkemper R, Nienhaus A. Evidence-based comparison of commercial interferon-gamma release assays for detecting active TB: a metaanalysis. Chest
28. Harries AD, Fitzsimons E, Fifield R, et al. Platelet count: a simple measure of activity in Crohn's disease. Br Med J (Clin Res Ed)
29. Talstad I, Rootwelt K, Gjone E. Thrombocytosis in ulcerative colitis and Crohn's disease. Scand J Gastroenterol
30. Mijac DD, Jankovic GL, Jorga J, et al. Nutritional status in patients with active inflammatory bowel disease: prevalence of malnutrition and methods for routine nutritional assessment. Eur J Intern Med