You are here

Screening prior to biological therapy in Crohn's disease: Adherence to guidelines and prevalence of infections. Results from a multicentre retrospective study

Digestive and Liver Disease, Volume 46, Issue 10, October 2014, Pages 881-886

Abstract

Background

Screening for opportunistic infections prior to starting biological therapy in patients with inflammatory bowel disease is recommended.

Aims

To assess adherence to screening for opportunistic infections prior to starting biological therapy in Crohn's disease patients and its yield.

Methods

A multicentre retrospective study was conducted in Crohn's disease patients in whom infliximab or adalimumab was started between 2000 and 2010. Screening included tuberculin skin test, interferon-gamma release assay or chest X-ray for tuberculosis. Extended screening included screening for tuberculosis and viral infections. Patients were followed until three months after ending treatment. Primary endpoints were opportunistic and serious infections.

Results

611 patients were included, 91% on infliximab. 463 (76%) patients were screened for tuberculosis, of whom 113 (24%) underwent extended screening. Screening for tuberculosis and hepatitis B increased to, respectively, 90–97% and 36–49% in the last two years. During a median follow-up of two years, 64/611 (9%, 3.4/100 patient-years) opportunistic infections and 26/611 (4%, 1.6/100 patient-years) serious infections were detected. Comorbidity was significantly associated with serious infections (hazard ratio 3.94).

Conclusions

Although screening rates for tuberculosis and hepatitis B increased, screening for hepatitis B was still suboptimal. More caution is required when prescribing biologicals in patients with comorbid conditions.

Keywords: Anti-TNF therapy, Inflammatory bowel disease, Opportunistic infections, Prevention.

1. Introduction

Anti-tumour necrosis factor (TNF) therapy has been an effective therapeutic strategy for the induction and maintenance of remission in inflammatory bowel disease (IBD)[1], [2], [3], and [4]. Irrespective of its benefits, safety concerns have been raised with regard to the occurrence of serious infections in patients on anti-TNF therapy[5] and [6], most notably reactivation of latent tuberculosis (TB) [7] .

As anti-TNF therapy is associated with the occurrence of serious infections in IBD and rheumatoid arthritis[8], [9], [10], [11], and [12], international guidelines advise pre-treatment screening for opportunistic infections to prevent complications[13], [14], and [15]. Screening for latent TB, mainly by using a tuberculin skin test (TST) and/or interferon-gamma release assay (IGRA), prior to anti-TNF therapy has proven to be effective in reducing the number of active TB-infections[16] and [17]. Although IGRAs may be associated with less false positive results as compared with TST, the high initial costs associated with IGRAs may impair their cost-effectiveness [18] . Screening for hepatitis B is generally considered to be useful as well [19] but data on human immunodeficiency virus (HIV), hepatitis C virus (HCV) and herpes viruses, such as varicella zoster virus (VZV), are less convincing[14] and [20].

The yield of screening is highly dependent on the adherence of healthcare providers to various pre-treatment screening strategies. To date, only one study assessed the adherence to recommended screening in IBD patients and found it to be relatively low (65% for TB and 25% for hepatitis B), but improving over time [21] . Additionally, the yield of screening is dependent on the prevalence rate of opportunistic infections in the background population, which shows great variation between countries worldwide[22] and [23].

In the present study, we aimed to assess adherence to screening prior to anti-TNF therapy and its yield in a cohort of patients with Crohn's disease (CD) treated with either infliximab or adalimumab.

2. Methods

2.1. Patient population and data collection

A multicentre retrospective cohort study was performed in three large referral centres for IBD in the Netherlands: the University Medical Centre Utrecht, the Academic Medical Centre in Amsterdam and the University Medical Centre Groningen. Anti-TNF therapy-naive patients with CD receiving their first administration of infliximab or adalimumab between 2000 and 2010 were included.

Data regarding demographic characteristics, comorbidity, disease characteristics, pre-treatment screening, anti-TNF therapy, immunosuppressive co-medication and opportunistic and serious infections during follow-up were obtained from the electronic medical record (EMR). Comorbidity was considered to be present if a patient had one or more of the following diseases: chronic renal failure, congestive heart failure, chronic obstructive pulmonary disease (COPD), diabetes mellitus, or chronic liver disease.

Patients were followed up from screening prior to first administration of infliximab or adalimumab until three months after discontinuation of anti-TNF therapy, death or end of study period (2011-06).

2.2. Definitions and outcomes

Screening for hepatitis B was not yet adopted in national and international guidelines during most of the study period. Therefore, recommended screening was defined as performing at least a TST, IGRA or chest X-ray for detection of latent TB, within three months prior to the start of treatment. TST was positive in case of an induration >10 mm, or >5 mm if immunosuppressing co-medication was used. If a patient had been vaccinated with Bacille Calmette Gúerin (BCG), a cut-off of 15 mm was used and a chest X-ray was routinely performed. Latent TB was defined as a positive result on TST (induration >10 mm, or >5 mm if immunosuppressive co-medication was used), IGRA or chest radiograph (calcification >5 mm, pleural thickening).

Extended screening was defined as TB-screening plus one or more serological tests for the detection of viral pathogens. Virological tests consisted of HBsAg and anti-HBc, anti-HCV, p24 antigen and HIV-antibodies for HIV and IgM and IgG antibodies and PCR for detection of Epstein–Barr virus (EBV), cytomegalovirus (CMV) and VZV. Virological tests of EBV, CMV and VZV were considered positive in the case of active infection (detectable IgM-antibodies or DNA).

Primary outcomes were opportunistic and serious infections. Opportunistic infections were defined as infections caused by micro-organisms that have no or limited pathogenic capacity under normal circumstances, but can cause disease in case of decreased immunity [14] . Serious infections were defined as infections requiring or prolonging hospital admission, leading to permanent disability or death [11] . Both opportunistic and serious infections had to occur during or within three months after anti-TNF treatment to be considered treatment-related [12] . Perianal and (postoperative) CD-related intra-abdominal abscesses were not taken into account.

2.3. Statistical analysis

Descriptive statistics were used to characterize the studied patient population. Means and medians were reported with a standard deviation (SD) and interquartile range (IQR), respectively. Linear regression analysis was used to assess the adherence to screening over time. Cox proportional hazard analysis was used to assess associations (expressed as hazard ratios, HR) between demographic and clinical characteristics, and serious infections. Different types of immunomodulating co-medications were included as time-dependent variables. If a variable appeared to be associated with serious infections in the univariate analysis (p < 0.10), it was further evaluated in the multivariate analysis. The primary outcome measures were reported as absolute risk and incidence rate (IR) for both opportunistic and serious infections, stated as the number of events per 100 person years of follow up.

3. Results

3.1. Patient characteristics

A total of 611 patients with CD starting infliximab or adalimumab between 2000 and 2010 were included. Of the included patients, 396 (64.8%) patients were female with a median age of 33.0 years (range 23.8–42.0 years) ( Table 1 ). Forty-seven (7.7%) patients had a comorbid condition, including 12 (25.5%) with COPD, 8 (17.0%) with chronic liver disease, and 7 (14.9%) with diabetes mellitus. The majority of patients were treated with infliximab (91%) and 80% of patients used one (63%) or two (18%) immunosuppressive co-medications, mostly thiopurines (59%).

Table 1 Baseline characteristics of Crohn's disease patients (n = 611).

  N = 611 (%)
Male gender 215 (35.2)
Median age, years (range) 33.0 (23.8–42.0)
Median disease duration, years (range) 6.5 (2.4–13.4)
Montréal A a
 A1 125 (20.5)
 A2 423 (69.2)
 A3 63 (10.3)
Montréal L b
 L1 147 (24.1)
 L2 188 (30.8)
 L3 265 (43.4)
 +L4 c 71 (11.6)
Montréal B d
 B1 241 (39.4)
 B2 157 (25.7)
 B3 163 (26.7)
 B2 and 3 8 (1.3)
 +Bp e 217 (35.5)
Comorbidity 47 (7.7)
 Chronic obstructive pulmonary disease 12 (25.5)
 Chronic liver disease 8 (17.0)
 Diabetes mellitus 7 (14.9)
 Chronic renal disease 4 (8.5)
 Congestive heart failure 2 (4.3)
Smoking f 189 (30.9)
Alcohol use g 194 (31.8)
Previous abdominal surgery h 256 (41.9)
Adalimumab 54 (8.8)
Comedication i 489 (80.0)
 Methotrexate use 101 (16.5)
 Thiopurine 360 (58.9)
 Steroid use j 135 (22.1)

a Age at diagnosis. A1 = <17 years, A2 = 17–40 years, A3 = >40 years.

b Disease localization: L1 = ileal, L2 = colonic, L3 = ileocolonic. Unknown in 1.4% and 1.9% of unscreened and screened patients, respectively.

c Affection of the upper gastrointestinal tract.

d Disease behaviour: B1 = non-stricturing, non-penetrating, B2 = stricturing, B3 = penetrating. Unknown in 1.4% and 4.3% of unscreened and screened patients, respectively.

e Perianal disease. Data were not available for one of three referral centres.

f Smoking at baseline, irrespective of amount.

g Alcohol use at baseline, irrespective of amount.

h Bowel surgery prior to baseline.

i Use of immunomodulating comedication during anti-TNF therapy, irrespective of type and number.

j Oral or intravenous glucocorticosteroid.

3.2. Pre-treatment screening and prevalent infections

Of 611 CD patients, 463 (75.8%) were screened for latent TB-infection prior to start of anti-TNF therapy: 177 (29.0%) with either TST or chest X-ray, 183 (30.0%) with both TST and chest X-ray and 103 (16.8%) employing an IGRA ( Table 2 ). The latter was always used in combination with chest X-ray, with (2%) or without (15%) TST.

Table 2 Screening practice, outcomes and treatment.

Screening test Patients (%) Outcomes a Treatment
    Positive Non-specific b  
Chest X-ray 388 (63.5) 2 (0.5) 1 (0.3) 2 INH
TST 348 (57.0) 12 (3.4) 3 (0.9) 6 INH
IGRA 103 (16.9) 0 (0.0) 5 (4.9) No treatment c
Hepatitis B 80 (13.1) 1 (1.3) 0 (0.0) No treatment d
Hepatitis C 78 (12.8) 0 (0.0) 0 (0.0) Not applicable
HIV 57 (9.3) 0 (0.0) 0 (0.0) Not applicable
EBV 41 (6.7) 4 (10.0) 3 (7.3) 1 valganciclovir e
CMV 40 (6.5) 1 (2.5) 1 (2.5) 1 valganciclovir
VZV 9 (1.5) 0 (0.0) 0 (0.0) Not applicable f
HSV 3 (0.5) 0 (0.0) 0 (0.0) Not applicable

a Only outcomes that were not negative are displayed.

b Outcome was neither negative nor positive.

c In the patients with non-interpretable outcomes, there was no suspicion for latent TB based on chest X-ray and/or TST.

d No treatment was indicated: the patient had a cleared infection.

e One patient with an EBV-CMV co-infection was treated with valganciclovir.

f A positive outcome for VZV was defined as negative IgG-serology. Treatment would have consisted of vaccination.

INH, isoniazid; TST, tuberculin skin test; IGRA, interferon-gamma release assay; HIV, human immunodeficiency virus; EBV, Epstein–Barr virus; CMV, cytomegalovirus; VZV, varicella zoster virus; HSV, herpes simplex virus.

Of 463 patients screened for tuberculosis, 15 (3.2%) were found to have a probable latent TB-infection, all detected by TST. Among patients tested with TST the proportion of positive outcomes was significantly lower in the group of patients using any co-medicationvs.no co-medication (9/282, 3.2%vs.6/66, 9.1%,p = 0.03).

All 15 patients with positive TST outcomes also had a chest X-ray. Two of them showed abnormalities on chest X-ray and were treated with isoniazid (INH) prior to the start of TNF inhibitor treatment. Six of the remaining 13 patients with positive TST results, but normal chest X-rays, were treated or had previously been treated with INH. The other seven patients with positive TSTs were not treated. One of the untreated patients was followed-up by regularly scheduled chest X-rays. No positive results were seen in patients tested with IGRA. In five cases the test was repeatedly reported as non-interpretable. In all five patients TST and chest X-ray were negative and thus latent TB was considered to be not present.

Virological screening was performed less often than TB-screening. Tests for hepatitis B, hepatitis C, and HIV were performed in 80 (13.1%), 78 (12.8%), and 57 (9.3%) patients, respectively. The only detected abnormality was one case of a cleared hepatitis B infection. Testing for herpes viruses, including CMV, EBV, VZV, was performed in up to 41 (6.7%) patients. Three active EBV-infections were found, but not treated with antiviral drugs, and one patient had an EBV-CMV co-infection and treated with valganciclovir. Of the four patients with an active EBV-infection, three received co-treatment with either azathioprine or methotrexate, and one patient received co-treatment with both azathioprine and corticosteroids.

The proportion of patients screened for TB increased over time, with an annual increase of 6.4% (95% CI 1.8–6.8%,p < 0.01). Similarly, both screening for hepatitis B and overall virological screening were more frequently performed over time, with annual increases of 3.7% (95% CI 1.1–6.3%,p = 0.01) and 4.3% (95% CI 1.8–6.8%,p < 0.01), respectively. Fig. 1 shows adherence to screening for hepatitis B, one or more viral infections and latent tuberculosis over time.

gr1

Fig. 1 Percentages of patients screened for hepatitis B, one or more viral infections (‘virological’) and latent tuberculosis over time.

3.3. Incident opportunistic and serious infections

During a median follow-up of two (range 0.3–10.9) years, 64 opportunistic infections were found in 56 (9.2%) patients and 26 serious infections in 26 patients (4.3%) ( Table 3 ). Eight of 64 (12.5%) opportunistic infections were serious. Incidence rates were 3.4 per 100 patient-years for opportunistic infections and 1.6 per 100 patient-years for serious infections.

Table 3 Incidence of opportunistic a and serious infections b in Crohn's disease patients (n = 611).

  N = 611 (%)
Patients with opportunistic infections (%) 56 (9.2)
Patients with serious infections (%) 26 (4.3)
Sum of follow-up, years (range) 1624
Median follow-up, years (range) 2.0 (0.8–3.8)
Opportunistic infections/100 PY 3.4
Serious infections/100 PY 1.6
Lethal infections (%) 2 (0.3)

a Opportunistic infections = infections caused by micro-organisms that have no or limited pathogenic capacity under normal circumstances, but can cause disease in case of decreased immunity [14] .

b Serious infections = infections requiring or prolonging hospital admission, leading to permanent disability or death [11] .

PY, patient-years of follow up.

Three patients died during or within three months after discontinuation of anti-TNF therapy. Two female patients with severe comorbidity both died of pneumosepsis: one with spastic tetraplegia and one with an EBV-associated B-cell non-Hodgkin lymphoma, for which she was treated with chemotherapy. One patient with a history of a not radically resected adenocarcinoma of the jejunum died of small bowel perforation.

3.4. Infection types

3.4.1. Opportunistic infections

The majority of opportunistic infections consisted of herpes virus-infections, mostly varicella zoster and herpes simplex, and fungal infections, mainly candidiasis ( Table 4 ).

Table 4 Type of opportunistic and serious infections of 611 Crohn's disease patients (n = 611).

Infection type a Opportunistic infections d (%) Serious infections e (%)
Herpes simplex 16 (25.0) 1 (3.8)
Herpes zoster 16 (25.0) 1 (3.8)
Candidiasis or other fungal 16 (25.0) 0 (0.0)
EBV 9 (14.1) 0 (0.0)
CMV 4 (6.3) 1 (3.8)
Tuberculosis b 3 (4.7) 3 (11.5)
Pneumonia c 0 (0.0) 11 (42.3)
Complicated UTI 0 (0.0) 5 (19.2)
Bacteraemia 0 (0.0) 4 (15.4)
Total 64 26

a Infections can be opportunistic, serious, or both. In case the numbers in both columns are ≥1, overlap exists.

b Including one case of miliary tuberculosis.

c Including one case of Coxiella burnetii infection.

d Opportunistic infections = infections caused by micro-organisms that have no or limited pathogenic capacity under normal circumstances, but can cause disease in case of decreased immunity [14] .

e Serious infections = infections requiring or prolonging hospital admission, leading to permanent disability or death [11] .

EBV, Epstein–Barr virus; CMV, cytomegalovirus; UTI, urinary tract infection.

3.4.2. Serious infections

Three patients were admitted because of viral infections: one patient with genital herpes simplex, one patient with extremely painful herpes zoster and one febrile patient with general illness due to CMV reactivation. Three cases of active TB occurred during follow-up. All three had been screened for latent TB-infection prior to anti-TNF therapy. One 34-year old male patient, who had negative outcomes of all three tests at screening, developed pulmonary and cerebral miliary tuberculosis during adalimumab treatment, successfully treated with quadruple therapy. The two other patients, both suffering from pulmonary TB, were screened at baseline using TST and chest X-ray. One of them had a positive TST, but was nevertheless started on infliximab therapy without TB-prophylaxis. The other one had negative outcomes for both tests. Both patients were effectively treated with combination therapy. Serious infections compromised mostly pneumonias caused by different and often unknown pathogens, including only one case of confirmed pneumococcal infection. Other serious infections were complicated urinary tract infections and bacteremias of various origins.

3.5. Risk factor analysis

In univariate analysis, factors associated with the occurrence of serious infections were longer disease duration (HR 1.05, 95% CI 1.02–1.05), ileal disease (HR 3.22, 95% CI 1.25–8.33) and presence of comorbidity (HR 4.93, 95% CI 1.95–12.50) (Supplementary Table S1). The hazard ratios (HRs) for the use of steroids and the use of two co-medications at a time were 1.86 (95% CI 0.54–6.37) and 1.38 (95% CI 0.29–6.59), respectively. The only factors that were independently associated with serious infections in multivariate analysis were comorbidity, with a HR of 3.94 (95% CI 1.40–11.14) and ileal disease (HR 2.96, 95% CI 1.12–7.82) ( Table 5 ).

Table 5 Multivariate analysis of factors associated with serious infections. a

  Hazard Ratio 95%-confidence interval p value
Age at inclusion 0.99 0.95–1.03 0.67
Disease duration 1.04 0.99–1.10 0.09
Montreal L b
 L1 2.96 1.12–7.82 0.03
 L2 1.82 0.66–5.08 0.25
 L3 Reference Reference 0.19
Comorbidity 3.94 1.40–11.14 0.01

a Serious infections = infections requiring or prolonging hospital admission, leading to permanent disability or death [11] .

b Disease localization. L1 = ileal, L2 = ileocolic, L3 = ileocolonic.

4. Discussion

This study demonstrates that screening for TB prior to the start of anti-TNF therapy in CD was performed in 75% of patients, of whom 25% was additionally tested for at least one viral infection. Seventeen percent of TB-screened patients were screened for hepatitis B. Screening rates have clearly increased over time. The incidence rates of serious and opportunistic infections were, respectively, 9% and 6%. Comorbidity was the strongest predictor for the occurrence of serious infections.

Three percent of patients screened for latent TB-infection had positive test results necessitating treatment prior to starting infliximab or adalimumab. Approximately half of these patients were treated with isoniazid. One of the untreated patients developed active TB-infection during infliximab treatment. In line with previous data [11] , negative screening results did not preclude the occurrence of TB-infection during anti-TNF treatment as two patients with negative outcomes for TST and chest X-ray developed active TB during treatment.

The negative screening results in our cohort might be explained by the fact that active TB represented ade novoinfection instead of a reactivation of latent TB. Yet, none of the patients who developed active TB had a history of contact with TB patients or prolonged stay in countries where TB is endemic. A more likely explanation is that patients with CD were anergic, due to the disease itself [24] or the use of immunosuppressive medication, resulting in false-negative TST results [25] . This is underlined by the lower rate of positive outcomes of TST among patients using immunosuppressive co-medication in our study. In order to lower the rate of false-negative TST results, some authors advocate TB-screening at the initial consultation of IBD, thus, before the start of immunomodulator or corticosteroid treatment[25] and [26]. The outcome of TST is also influenced by previous vaccination with BCG, which can lead to false-positive results. In the Netherlands the BCG coverage is low, because since the 1950s universal BCG vaccination has been replaced by BCG vaccination targeted at high-risk groups [27] . Because of the low BCG coverage in our cohort, the outcome of TST can hardly been affected by previous BCG vaccinations. Although BCG vaccination status was not routinely recorded during screening, it was always assessed in case of positive TST results.

When evaluating the risk of latent TB, it is of major importance not to solely rely on TST or chest X-ray outcomes, but to carefully take a medical history as well [28] . Unfortunately, this could not be completely ascertained from the electronic medical records. IGRAs may be of additional value. These assays measure the T-cell production of interferon gamma in response toin vitrostimulation of antigens specific for Mycobacterium tuberculosis and absent in BCG and atypical mycobacteria. IGRAs may therefore result in less false-positive results as compared with TST. Yet, the cost-effectiveness of IGRAs are affected by their high initial costs, in particular in countries with a low prevalence of latent TB [18] .

Adherence to screening was consistent with the results of a previous study [21] . This study demonstrated that respectively 65% and 25% of 287 IBD-patients starting anti-TNF agents between 1998 and 2010 were appropriately screened for TB and hepatitis B. Additionally, a similar trend of increased screening over time was found. Most important factors associated with non-adherence to screening for latent TB and hepatitis B included initiation of anti-TNF therapy before 2006 and 2010, respectively. At these time-points recommendations regarding screening for TB and hepatitis B were elaborated in international guidelines[13] and [15]. Obviously, in our cohort adherence to screening was significantly higher in patients using adalimumab, which was registered in the Netherlands for the treatment of CD in 2007.

Incidence rates of opportunistic and serious infections were in line with previously reported data[10], [11], [12], and [29]. The most important risk factors for serious infections found in previous cohort studies[11] and [12]and one case-control study [30] were the use of corticosteroids or more than one co-medication. Although the HR was also increased for the use of co-medication in our study, concomitant treatment with thiopurines or methotrexate was not significantly associated with serious infections (1.86; 95% CI 0.54–6.37). Nevertheless, concomitant immunomodulator treatment was associated with an up to six-fold increased risk of serious infections, which might be of clinical importance. The most important independent predictor for the occurrence of serious infections was comorbidity. Comorbidities, in particular COPD and diabetes mellitus, are clearly associated with serious infections in rheumatic arthritis patients on anti-TNF therapy[31] and [32]. Although this association has not previously been observed in IBD, results from a large hospitalization database indicate that co-morbidity is associated with infection-related hospitalization [33] . In contrast to previous observational studies[29], [30], and [34], no independent association was observed between advanced age and the occurrence of serious infections. These contradicting results might be caused by the relatively small number of older patients in our cohort. Alternatively, results of other observational studies may be confounded as comorbidity was not included in the multivariate analysis. Therefore, age may serve as a proxy for comorbidity. With the growth of the elderly IBD population – currently representing 10–30% of the total IBD population [33] – clinicians will increasingly be confronted with IBD patients with comorbid conditions. Thus, clinicians should exercise caution when prescribing TNF-α inhibitors in (elderly) IBD patients with comorbid conditions.

Another finding of our study is that 20 out of 26 (77%) serious infections were caused by bacterial pathogens. To prevent bacterial infections it has been recommended to perform an extensive work-up prior to initiating immunomodulator or anti-TNF therapy, encompassing a history of bacterial infections, blood tests and urine analysis. Additionally, all patients should receive pneumococcal vaccination[19] and [31]. Of note, 9 out of 20 (45%) serious bacterial infections were not caused by pneumococcal bacteria. As these bacterial infections are very likely to representde novoinfections, vigilant monitoring for signs and symptoms of urinary, pulmonary, or gastrointestinal infections during anti-TNF therapy is warranted. Empirically, such an approach may be more effective instead of testing all patients for latent bacterial infections prior to initiating anti-TNF therapy.

The strengths of our study are the large number of included patients and the multicentric design, thereby reliably reflecting screening practice as performed in academic centres.

Our study has several limitations. First, the small number of patients screened for tuberculosis and hepatitis B precludes definitive conclusions regarding the protective effect of screening. However, universal screening for hepatitis B did not seem to be very useful in our cohort, probably because of the low prevalence of hepatitis B in the Dutch background population (i.e., 0.2% [34] ). Second, vaccinations were not taken into account in this study. The most recent multidisciplinary Dutch guideline on the use of biologicals was published in 2011 and recommends vaccinating patients for influenza and optionallyStreptococcus pneumoniae. This guideline was implemented after closure of our study. Vaccination against influenza would probably not have affected our outcomes. Conversely, pneumococcal vaccination might have prevented one or more of the serious infections observed in our cohort. Third, the retrospective design may have led to an underestimation of the actual adherence to TB-screening, as TST-outcomes were mostly reported non-electronically. Since it is very unlikely for positive results of TST not to be reported in the EMR, we do not feel that this has had a major influence on our data. Fourth, according to our data, IBD patients with comorbidity have an increased risk of serious infections, irrespective of age, disease duration and disease localization. Yet, it is difficult to quantify how much of this increased risk is due to anti-TNF or underlying comorbid conditions.

In conclusion, at least three quarters of CD-patients were screened for TB prior to the start of anti-TNF therapy, of whom a quarter were additionally screened for one or more viral infections. Although screening rates clearly increased over time, adherence to screening for hepatitis B was still suboptimal at the end of the study period. Only three percent of patients screened for TB were shown to have a latent TB-infection, but a negative TST did not always rule out the possibility of an active TB-infection later on. This study extends our knowledge on the safety of anti-TNF therapy, by showing that comorbidity is an important risk factor for the occurrence of serious infections. Therefore, caution is required when prescribing anti-TNF compounds in patients with comorbidity. This is especially relevant since the elderly population with IBD is growing fast.

Conflict of interest

HH has acted as a consultant for Abbvie. CY has acted as a consultant for Abbvie and received payments for lectures from Ferring and MSD. BO has received payments for lectures from Ferring and acted as a consultant for Abbvie and MSD. MGH has been a consultant and recipient of an unrestricted research grant from Abbvie.

Appendix A. Supplementary data

The following are the supplementary data to this article:

Download file

Supplementary Table S1 Univariate analysis of factors associated with serious infections.

References

  • [1] A.K. Akobeng, M. Zachos. Tumor necrosis factor-alpha antibody for induction of remission in Crohn's disease. Cochrane Database of Systematic Reviews. 2004;1:CD003574
  • [2] B.W. Behm, S.J. Bickston. Tumor necrosis factor-alpha antibody for maintenance of remission in Crohn's disease. Cochrane Database of Systematic Reviews. 2008;1:CD006893
  • [3] M.M. Lawson, A.G. Thomas, A.K. Akobeng. Tumour necrosis factor alpha blocking agents for induction of remission in ulcerative colitis. Cochrane Database of Systematic Reviews. 2006;3:CD005112
  • [4] W. Reinisch, W.J. Sandborn, P. Rutgeerts, et al. Long-term infliximab maintenance therapy for ulcerative colitis: the ACT-1 and -2 extension studies. Inflammatory Bowel Diseases. 2012;18:201-211 Crossref
  • [5] T. Bongartz, A.J. Sutton, M.J. Sweeting, et al. Anti-TNF antibody therapy in rheumatoid arthritis and the risk of serious infections and malignancies: systematic review and meta-analysis of rare harmful effects in randomized controlled trials. Journal of the American Medical Association. 2006;295:2275-2285 Crossref
  • [6] J.A. Singh, G.A. Wells, R. Christensen, et al. Adverse effects of biologics: a network meta-analysis and Cochrane overview. Cochrane Database of Systematic Reviews. 2011;2:CD008794
  • [7] J. Keane, S. Gershon, R.P. Wise, et al. Tuberculosis associated with infliximab, a tumor necrosis factor alpha-neutralizing agent. New England Journal of Medicine. 2001;345:1098-1104 Crossref
  • [8] L. Peyrin-Biroulet, P. Deltenre, N. de Suray, et al. Efficacy and safety of tumor necrosis factor antagonists in Crohn's disease: meta-analysis of placebo-controlled trials. Clinical Gastroenterology and Hepatology. 2008;6:644-653 Crossref
  • [9] G.R. Lichtenstein, B.G. Feagan, R.D. Cohen, et al. Serious infections and mortality in association with therapies for Crohn's disease: TREAT registry. Clinical Gastroenterology and Hepatology. 2006;4:621-630 Crossref
  • [10] J.F. Colombel, E.V. Loftus Jr., W.J. Tremaine, et al. Gastroenterology. 2004;126:19-31 Crossref
  • [11] H. Fidder, F. Schnitzler, M. Ferrante, et al. Long-term safety of infliximab for the treatment of inflammatory bowel disease: a single-centre cohort study. Gut. 2009;58:501-508 Crossref
  • [12] I.C. Lawrance, G.L. Radford-Smith, P.A. Bampton, et al. Serious infections in patients with inflammatory bowel disease receiving anti-tumor-necrosis-factor-alpha therapy: an Australian and New Zealand experience. Journal of Gastroenterology and Hepatology. 2010;25:1732-1738 Crossref
  • [13] G.R. Lichtenstein, M.T. Abreu, R. Cohen, et al. American Gastroenterological Association Institute technical review on corticosteroids, immunomodulators, and infliximab in inflammatory bowel disease. Gastroenterology. 2006;130:940-987 Crossref
  • [14] J.F. Rahier, S. Ben-Horin, Y. Chowers, et al. European evidence-based Consensus on the prevention, diagnosis and management of opportunistic infections in inflammatory bowel disease. Journal of Crohn's and Colitis. 2009;3:47-91 Crossref
  • [15] A. Kornbluth, D.B. Sachar. Ulcerative colitis practice guidelines in adults: American College Of Gastroenterology, Practice Parameters Committee. American Journal of Gastroenterology. 2010;105:501-523 Crossref
  • [16] L. Carmona, J.J. Gómez-Reino, V. Rodríguez-Valverde, et al. Effectiveness of recommendations to prevent reactivation of latent tuberculosis infection in patients treated with tumor necrosis factor antagonists. Arthritis and Rheumatism. 2005;52:1766-1772 Crossref
  • [17] J.J. Gomez-Reino, L. Carmona, D.M. Angel. Risk of tuberculosis in patients treated with tumor necrosis factor antagonists due to incomplete prevention of reactivation of latent infection. Arthritis and Rheumatism. 2007;57:756-761 Crossref
  • [18] M. Van der Have, B. Oldenburg, H.H. Fidder, et al. Optimizing screening for tuberculosis and hepatitis B prior to starting tumor necrosis factor-α inhibitors in Crohn's disease. Digestive Diseases and Sciences. 2013; [Epub ahead of print]
  • [19] J.P. Gisbert, M. Chaparro, M. Esteve. Review article: prevention and management of hepatitis B and C infection in patients with inflammatory bowel disease. Alimentary Pharmacology and Therapeutics. 2011;33:619-633 Crossref
  • [20] N. Viget, G. Vernier-Massouille, D. Salmon-Ceron, et al. Opportunistic infections in patients with inflammatory bowel disease: prevention and diagnosis. Gut. 2008;57:549-558
  • [21] B.P. Vaughn, G.A. Doherty, S. Gautam, et al. Screening for tuberculosis and hepatitis B prior to the initiation of anti-tumor necrosis therapy. Inflammatory Bowel Diseases. 2012;18:1057-1063 Crossref
  • [22] C. Dye, S. Scheele, P. Dolin, et al. Consensus statement. Global burden of tuberculosis: estimated incidence, prevalence, and mortality by country. WHO Global Surveillance and Monitoring Project. Journal of the American Medical Association. 1999;282:677-686 Crossref
  • [23] M.J. Alter. Epidemiology of hepatitis B in Europe and in worldwide. Journal of Hepatology. 2003;39(Suppl. II):S64-S69
  • [24] W.S. Mow, M.T. Abreu-Martin, K.A. Papadakis, et al. High incidence of anergy in inflammatory bowel disease patients limits the usefulness of PPD screening before infliximab therapy. Clinical Gastroenterology and Hepatology. 2004;2:309-313 Crossref
  • [25] A.M. Schoepfer, B. Flogerzi, S. Fallegger, et al. Comparison of interferon-gamma release assay vs. tuberculin skin test for tuberculosis screening in inflammatory bowel disease. American Journal of Gastroenterology. 2008;103:2799-2806
  • [26] P. Papay, A. Eser, S. Winkler, et al. Factors impacting the results of interferon-γ release assay and tuberculin skin test in routine screening for latent tuberculosis in patients with inflammatory bowel diseases. Inflammatory Bowel Diseases. 2011;17:84-90 Crossref
  • [27] Health Council of the Netherlands. Vaccination of young children against tuberculosis. (Health Council of the Netherlands, The Hague, 2011) publication no. 2011/04
  • [28] BTS recommendations for assessing risk and for managing Mycobacterium tuberculosis infection and disease in patients due to start anti-TNF-alpha treatment. Thorax. 2005;60:800-805
  • [29] M. Naganuma, R. Kunisaki, N. Yoshimura, et al. Prospective analysis of the incidence of and risk factors for opportunistic infections in patients with inflammatory bowel disease. Journal of Gastroenterology. 2013;48:595-600 Crossref
  • [30] M. Toruner, E.V. Loftus Jr., W.S. Harmsen, et al. Risk factors for opportunistic infections in patients with inflammatory bowel disease. Gastroenterology. 2008;134:929-936 Crossref
  • [31] S.A. Van Dartel, J. Fransen, W. Kievit, et al. Predictors for the 5-year risk of serious infections in patients with rheumatoid arthritis treated with anti-tumour necrosis factor therapy: a cohort study in the Dutch Rheumatoid Arthritis Monitoring (DREAM) registry. Rheumatology. 2013;52:1052-1057 Crossref
  • [32] M.F. Doran, C.S. Crowson, G.R. Pond, et al. Predictors of infection in rheumatoid arthritis. Arthritis and Rheumatism. 2002;46:2294-2300 Crossref
  • [33] A.N. Ananthakrishnan, E.L. McGinley. Infection-related hospitalizations are associated with increased mortality in patients with inflammatory bowel diseases. Journal of Crohn's and Colitis. 2013;7:107-112 Crossref
  • [34] M. Cottone, A. Kohn, M. Daperno, et al. Advanced age is an independent risk factor for severe infections and mortality in patients given anti-tumor necrosis factor therapy for inflammatory bowel disease. Clinical Gastroenterology and Hepatology. 2011;9:30-35 Crossref

Footnotes

a Department of Gastroenterology and Hepatology, University Medical Center Utrecht, Utrecht, Netherlands

b Department of Gastroenterology and Hepatology, University Medical Center Groningen, University of Groningen, Groningen, Netherlands

c Department of Gastroenterology and Hepatology, Academic Medical Center, University of Amsterdam, Amsterdam, Netherlands

lowast Corresponding author at: Heidelberglaan 100, 3508 GA Utrecht, Netherlands. Tel.: +31 887557021; fax: +31 887555533.

1 These authors contributed equally to the work.