Repeated tuberculin skin testing following therapy with TNF-alpha inhibitors

Transcription

1 Clin Rheumatol (2009) 28: DOI /s ORIGINAL ARTICLE Repeated tuberculin skin testing following therapy with TNF-alpha inhibitors Inbal Fuchs & Lone Avnon & Tamar Freud & Mahmoud Abu-Shakra Received: 14 May 2008 /Revised: 14 July 2008 /Accepted: 27 August 2008 / Published online: 16 September 2008 # Clinical Rheumatology 2008 Abstract To determine the rate of true tuberculin skin test (TST) response in a cohort of patients with rheumatic disease treated with tumor necrosis factor inhibitors (TNFi). The study population included consecutive patients with rheumatoid arthritis (RA), ankylosing spondylitis (AS), and psoriatic arthritis (PsA) treated with TNFi for at least 3 months. Patients with a positive TST at screening who began Tb prophylaxis before the beginning of TNFi therapy were excluded. All patients underwent a second TST. True TST response was defined as an increase of 6 mm of induration between the screening test and the second test. Forty patients (12 men and 28 women) were included. Mean age was 51.2 years. Of them, 27 (67.5%) had RA, I. Fuchs Department of Family Medicine, Faculty of Health Sciences, Ben-Gurion University, L. Avnon Pulmonology Unit, Soroka Medical Center and Faculty of Health Sciences, Ben-Gurion University, T. Freud Siaal Research Center for Family Medicine and Primary Care, Faculty of Health Sciences, Ben-Gurion University, M. Abu-Shakra Rheumatic Diseases Unit, Soroka Medical Center and Faculty of Health Sciences, Ben-Gurion University, M. Abu-Shakra (*) Department of Medicine D, Soroka University Medical Center, POB 151, Beer-sheva, Israel Mahmoud@bgu.ac.il eight (20%) had PsA, and five patients (12.5%) had AS. At pre-treatment TST, 15 patients had a TST 5 mm. A significantly higher percent of patients with TST 5 mm was seen among men compared with women (75% vs. 21%, p=0.012) and patients with PsA compared with patients with RA (75% vs. 22%, p=0.014). At the second test, eight (20%) had an increase of 6 mm between readings with four having an increase of 10 mm or more. Four patients received infliximab and the other four were treated with etanercept. Seven of these eight patients had RA and one was a patient with PsA. Patients with true TST response were significantly older and non-smokers with elevated sedimentation rate and a higher rate of anemia. Nationality, comorbid conditions, treatment with immunosuppressives, and BCG vaccination status had no significant influence on the TST response. Serial TST testing in patients receiving TNFi is indicated to identify patients with reactivation of latent tuberculosis infection or those exposed to mycobacterium. Keywords Conversion. TNF inhibitors. Tuberculin Several studies have reported a high incidence of active tuberculosis (Tb) in patients with rheumatic diseases treated with anti-tumor necrosis factor-alpha inhibitors (TNFi) relative to background incidence of Tb. In a recent study the estimated incidence of Tb in patients treated with TNFi was 172 cases/100,000 patient years in a country where incidence in rheumatic patients cohort not receiving TNFi was 95 cases/100,000 [1]. Observations in animal models have indicated that the anti-inflammatory activity of TNFi results in the fragmentation of TNF-dependent granulomas [2, 3]. This phenomenon could lead to loss of control in the balance existing

2 168 Clin Rheumatol (2009) 28: between the immune system and the mycobacterium, and explain the presumed association between reactivation of latent Tb infection (LTBI) and treatment with TNFi [1]. As a result of these reports, individual medical societies have introduced guidelines for pre-treatment screening LTBI [4]. These guidelines consist of obtaining a detailed medical history focusing on risk factors for Tb, performing a tuberculin skin test (TST) and performing a chest radiograph. Cutoff for TST positivity varies between 5 mm and 10 mm among the different guidelines. Lack of adherence to the guidelines is associated with a sevenfold increased risk of developing Tb following therapy with TNFi [1]. Certain recent publications have also recommended periodic follow-up skin testing as a monitoring tool for possible conversion of TST during TNFi therapy [5]. Interpretation of TST and particularly of repeated tests remains controversial. Random variability in test results occurs as a result of inter- and intra-reader variability [6]. It has been suggested that differences of reading greater than 6 mm indicate true biologic phenomena which could be either conversion or boosting. Conversion is defined as the development of new delayed type hypersensitivity to mycobacterial antigens following new infection with Mycobacterium tuberculosis, nontuberculous mycobacteria, or BCG vaccination. Booster phenomenon is an increased tuberculin reaction upon retesting in the absence of new infection and is believed to represent an anamnestic serologic response [6]. A significantly attenuated response to tuberculin was reported among patients with rheumatoid arthritis (RA) in comparison with controls in a Tb-endemic country [7]. The pattern of delayed type hypersensitivity response in response to TST had yet to be established with the introduction of TNFi treatment. Recently, Hatemi et al. demonstrated that treatment with infliximab does not suppress TST reactivity [8]. Another study showed that TST positive individuals remain positive and have no increased anergy following treatment with infliximab [5]. The aim of this study was to perform a second TST at least 3 months after beginning of TNFi therapy and identify variables that may be associated with change in TST reaction. Materials and methods Patients The study population included consecutive patients with RA, ankylosing spondylitis (AS), and psoriatic arthritis (PsA) who were treated with TNFi agents for at least 12 weeks. All patients fulfilled the American College of Rheumatology classification criteria for RA, AS, and PsA. Patients with a positive TST at screening who began Tb prophylaxis before the beginning of TNFi therapy were excluded. In Israel it is a matter of debate when to diagnose LTBI among patients treated with TNFi. While for some a TST> 5 mm is diagnostic for LTBI, it is widely accepted to diagnose LTBI only if TST reading is greater than 10 mm [9, 10]. As a result, some patients with TST between 5 10 mm had not received therapy for LTBI at screening and were included in the study. The screening visit was performed 1 2 weeks before beginning of TNFi therapy. At screening, a complete history was obtained and physical examination was completed. In addition, a pre-treatment TST (first test), inspection for BCG scar, a laboratory evaluation, and chest radiograph were performed and the results duly recorded. At recruitment to the study, a history of previous personal and/or occupational Tb exposure, travel and immigration history, and smoking status was recorded. Patients were also questioned about the following tuberculosis-related symptoms: night sweats, weight loss, cough, and/or bloody sputum. A second TST was performed after at least 12 weeks of TNFi therapy. All data were recorded in a standard protocol. Therapy for LTBI None of the patients received therapy for LTBI between the two TST. At the second test, all patients with TST readings>5 mm were referred to a pulmonologist and ultimately all began prophylaxis for LTBI. The study received local Ethics Committee approval in accordance with local requirements. All patients signed informed consent before entering the study. Tuberculin skin test The TST using five international units (IU) was performed on the volar side of the forearm according to the Mantoux method and was measured 72 h later [6, 11] by one of four experienced personnel. True TST response was defined as an increase of 6 mm of induration between the screening test and the second test [6, 11]. Patients were considered TST anergic if the test result was <2 mm, negative if result was 2 4 mm, and TST positive if skin test reading was 5 mm. Patients were classified as reverters if they were TST positive at screening and became TST negative at the second test [6]. Exposure definition Possible professional exposure to M. tuberculosis was considered positive if the patient was a health care worker, or worked specifically with immigrants, prisoners, or intravenous drug users. Possible personal exposure to Tb was considered positive if the patient immigrated from or lived in a Tb endemic country, spent more than 2 weeks in a Tb endemic country, or had close contact with known Tb patients, prisoners, immigrants, or intravenous drug users.

3 Clin Rheumatol (2009) 28: Statistical analysis Analyses were carried out using SPSS 14.0 for Windows software package. Continuous variables were compared using Student s t test. Chi-square or Fischer s exact test where appropriate were used for comparing qualitative variables. A p value<0.05 was considered significant for all statistical tests. Results Forty patients (12 men and 28 women) were included in the study. Their mean age was 51.2 years (SD=13.6 yrs). Of them, 27 (67.5%) had RA, eight (20%) had PsA, and five patients (12.5%) had AS. Fifteen (37.5%) patients were smokers and four (10%) were ex-smokers. Comorbid conditions included diabetes mellitus (four patients (10%)), hypertension (seven patients (17.5%)), coronary heart disease, congestive heart failure, and renal failure, (one patient each) and two patients (5%) had asthma. None of the patients had a history of active Tb in the past. Thirtyfive (87.5%) patients had BCG vaccination in childhood, all but one more than 15 years before TST. Of the 40 patients, 25 (62.5%) received infliximab and 15 (37.5%) received etanercept. No patient was treated with adalimumab. TST results The mean interval between the first and second TST was months (range 3 53 months). The results of TST reading at the two assessments are shown in Table 1. At the screening test 50% of the patients were anergic and 15 patients had a TST reading>5 mm. Table 2 compares characteristics of patients with TST 5 mm with those with reading<5 mm, at the screening test. A significantly higher percentage of patients with TST 5 mm was seen among men compared with women (75% vs. 21%, p=0.012) and patients with PsA compared with patients with RA (75% vs. 22%, p=0.014). At the second assessment, 20 (50%) patients had a reading>5 mm. None of the patients was diagnosed with active Tb. Repeated TST reading Compared with induration at screening, eight (20%) patients had an increase equal to or greater than 6 mm Table 1 Tuberculin skin testing reading at the 2 assessments TST result (mm) 1st test No. (%) 2nd test No. (%) True TST response No. (%) Reverter at 2nd test No. (%) 2< 20 (50) 16 (40) 3 (7.5) 2 5< 5 (12.5) 4 (10) 2 (5) 5 15 (37.5) 20 (50) 3 (7.5) 2 (5) Table 2 Characteristics of patients with TST<5 mm compared with those with reading 5 mm at screening (1st test) <5 mm at screening (N=25) 5 mmat screening (N=15) p value Age, mean years (SD) (15.43) (10.38) Gender No. (%) Male 3 (12) 9 (60) Female 22 (88) 6 (40) Country of birth, No. (%) Israel-Jewish 10 (40) 11 (73.3) Israel-Bedouin 4 (16) Eastern Europe 7 (28) 3 (20) Others 4 (16) 1 (6.7) Years in Israel (immigrants), 27.7 (19.48) 33.5 (17.99) Smoking status, No. (%) Current smoker 7 (28) 8 (53.3) Ex-smoker 4 (16) 0 No. of adults/no. of rooms, 0.75 (0.36) 0.68 (0.44) Diagnosis No. (%) Rheumatoid arthritis 21 (84) 6 (40) Psoriatic arthritis 2 (8) 6 (40) Ankylosing spondylitis 2 (8) 3 (20) (true response) at the second testing. Of those eight patients, four had an increment equal to or greater than 10 mm. Four patients received infliximab and the other four were treated with etanercept. Seven of the patients had RA and one was a patient with PsA. Two patients with PsA were TST positive at screening and became reverters at the second reading. Variables associated with true TST response Table 3 compares the demographics of the patients who had true TST response with the patients without conversion after therapy with TNFi. Patients with true TST response were significantly older and non-smokers. There was a higher rate of true response among immigrants. Twenty-five percent of the true responders are Israeli-born (Israeli Jewish and Bedouin) in comparison with 72% Israeli-born among the non-responders (p=0.03). Medical therapies and laboratory data of the RA patients are shown in Table 4. Seven of the true TST responders received treatment with prednisone and all received methotrexate therapy. All responders were more likely to be anemic and with a higher erythrocyte sedimentation rate (ESR) prior to TNFi treatment (p= 0.01and 0.001, respectively). Table 5 shows the number of patients with possible occupational and personal Tb exposure and percentage of patients who presented with Tb-related symptoms.

4 170 Clin Rheumatol (2009) 28: Table 3 Demographic characteristics and diagnoses of true TST responders and non-responders Six (75%) of the patients with true TST response had symptoms of cough, weight loss, or night sweats compared with 13 (40.6%) among patients without true TST response (p=0.089). All patients with true TST response were vaccinated with BCG in childhood compared with 27 (84.4%) among non-responders (p=0.31). A positive history of possible occupational or other personal exposure to Tb had no significant influence on the probability to true response to TST, but the responders lived in significantly more crowded living conditions (p=0.003). Discussion Responders n=8 Non-responders n=32 p value Age, mean years (SD) (10.6) (12.3) Gender No. (%) Women 7 (87.5) 21 (65.6) Men 1 (12.5) 11 (34.4) Country of birth, No. (%) Israel-Jewish 2 (25) 19 (59.4) Israel-Bedouin 0 4 (12.5) Eastern Europe 2 (25) 8 (25) NS Others 4 (50) 1 (3.1) Smoking status, No. (%) Current smoker 0 15 (46.9) Ex-smoker 1 (12.5) 3 (9.4) Years in Israel (immigrants), 40.3 (16.7) a 21.8 (17) b No of adults/no. of rooms, 1.09 (0.52) 0.64 (0.43) Diagnosis No. (%) RA 7 (87.5) 20 (62.5) PsA 1 (12.5) 7 (21.9) AS 0 5 (15.6) NS Not significant a Six immigrants b Nine immigrants Table 4 Medical therapy and laboratory data of patients treated with TNF inhibitors True TST responders n=8 Nonresponder n=32 p value Medical therapy, No. (%) Prednisone 7 (87.5%) (56.3%) Mean dosage in mg (SD) 11 (10.26) 8 (4.25) Anti-malaria 6 (75%) (59.4%) Azathioprine 2 (25%) 9 (28.1%) Methotrexate 8 (100%) (90.6%) Mean dosage in mg (SD) 7.81 (1.6) (3.38) Mean duration of TNFi therapy months Lab values (median values) Hemoglobin, mg% WBC, Creatinine, mg% ESR, mm/hr CRP mean, mg% Chest radiograph Normal No (%) 5 (62.5) 21 (75) Interstitial changes 3 (37.5) 3 (10.7) No (%) Other No. (%) 0 (0) 8 (25.6) Criteria for diagnosing LTBI in patients treated with TNFi remain a matter of debate. A variety of guidelines have been proposed and they differ mainly in the TST reading cutoff used for diagnosis of LTBI and in the duration of prophylaxis administration prior to initiation of TNFi. Limitations of TST include: (a) a high occurrence of attenuated response to tuberculin or anergy as a result of impairment in cellular immune response seen in patients with RA, other immune mediated diseases and chronic diseases and (b) a positive reading in patients sensitized by BCG vaccination less than 15 years before TST and/or as a result of advanced age [6, 12, 13].Despite those limitations, TST remains useful in diagnosing of LTBI in chronic rheumatic patients, inflammatory bowel diseases, dialysis patients, and patients with human immunodeficiency virus [14]. It has been suggested that once TST is positive it is no longer useful; therefore, repeated TST is not useful for patients diagnosed with LTBI and who received prophylaxis therapy [6]. The role of repeated TST in patients treated with TNFi who had not received therapy for LTBI at screening is unclear. Annual TST is performed among high risk populations, including mycobacteriology lab personnel, and residents and employees of high risk institutions (e.g. hospitals, detention facilities). In light of recent studies, it appears also that TST reading remains informative during TNFi therapy [5, 8]. True TST response could be due to either conversion or boosting [6]. The booster phenomenon, a false positive increment in the TST reading, is maximal if the interval between the first and second TST is between 1 5 weeks, and rarely seen if the interval is greater than 90 days [6]. Conversion occurs following new infection with M. tuberculosis, nontuberculous mycobacteria, or BCG vaccination, and it is associated with 8 15% risk of developing active Tb within 1 4 years [6]. Distinguishing boosting from conversion is difficult and is based on interval

5 Clin Rheumatol (2009) 28: Table 5 History of possible exposure to tuberculosis, tuberculosis related symptoms and BCG vaccination of the patients TST responders (N=8) Non-responders (N=32) p value 0 4 mm at screening (N=25) 5+ mm at screening (N=15) p value Possible Tb exposure No. (%) Occupational 0 (0) 17 (53) (24) 11 (73.3) Personal 4 (50) 23 (72) (72) 9 (60) Tb related symptoms a No. (%) Yes 6 (75%) 13 (40.6) (44) 8 (53) No 2 (25) (56) 7 (46.7) BCG scar No. (%) Yes 8 (100) 27 (84.4) (80) 15 (100) No 0 (0) 5 (15.6) 5 (20) 0 a Cough, weight loss, night sweats between tests, possibility of Tb exposure, absolute size of second test or the difference between two tests [6]. An absolute size>10 mm and more than 6 mm increase in the second test suggests conversion. In this study, the mean interval between the two skin tests was months and the absolute size of TST in all of the true TST responders was >10 mm, suggesting boosting is unlikely. There was no significant difference between rate of BCG vaccination among true TST responders and non-responders. Although 35 patients (87.5%) received BCG, 34 of them were vaccinated more than 15 years before the study was conducted. It is widely accepted that most of the BCG reactivity wanes over 10 years and therefore TST performed more than 15 years after BCG should be interpreted as for unvaccinated persons [6]. This observation suggested that in our patients, true TST response is not the result of previous BCG vaccination but due to true conversion reaction. The possible association between true TST response and reactivation of LTBI in this group of patients is a plausible theory that has yet be proved. TNFi-induced immune aberrations and breakdown of granuloma could trigger the TST response [2 3]. Alternatively, the true TST response may be the result of activity of TNFi on T-cells. Berg et al. [15] reported that treatment of RA patients with etanercept may lead to increased peripheral T-cell reactivity to tuberculin and other microbial antigens, suggesting that TNF alpha blockade may stimulate certain aspects of antimicrobial immune defense. Re-exposure to mycobacterium is yet another alternative explanation. We did not identify a greater exposure risk to Tb among the true TST responders. However, we found that the responders abide in denser living conditions. While the prevalence of Tb in Israel is low, it is possible that TNFi-treated patients were exposed to mycobacterium from other patients during their frequent visits to the clinic. In light of the multiple plausible mechanisms, the lack of overt explanation for the true TST response and the absence of studies that assessed the risk of active Tb among true TST responders, we suggest adopting the more vigilant course of action: diagnosing and treating the true TST responders as LTBI. We recognize that as a result of this approach, more TNFi-treated patients will become candidates for LTBI prophylaxis and thus vulnerable to increased risk of liver toxicity due to concomitant therapy of isoniazid with other drugs, particularly methotrexate and non-steroidal antiinflammatory drugs [16]. Recently, the QuantiFERON TB Gold (QTG) test for diagnosing both active and LTBI has been introduced [17 19]. Advantages of the assay include: lack of anamnestic response and absence of cross reactivity with commonly encountered mycobacterium other than tuberculosis except for Mycobacterium Kansasii and Mycobacterium Marinum, results in 24 h, and no need for a repeat visit. While the QTG may be more specific than TST for detecting M. tuberculosis infection, its sensitivity still needs to be determined. Recent studies reported that the test is less sensitive among immuno-suppressed patients, and in more than 10% of cases the results can not be interpreted as either positive or negative [20]. The Center for Disease Control (CDC) does not recommend its use among immuno-suppressed patients with increased risk for progression LTBI to active Tb [21]. In summary, this study suggests that true TST response is common following TNFi therapy as a result of several plausible mechanisms. We encourage treating LTBI with anti-tuberculosis therapy when TST 5 mm at screening, performing periodic TST and treating LTBI when the patient becomes a true responder. Acknowledgement The authors would like to thank Prof. Camil Fuchs for his help with statistical analysis and Ms. Maria Rubinstein and Ms. Ruthie Hadar for their help in conducting the skin tests. Disclosures None.

6 172 Clin Rheumatol (2009) 28: References 1. Gomez-Reino J, Carmona L, Descalzo M (2007) Risk of tuberculosis in patients treated with tumor necrosis factor antagonists due to incomplete prevention of reactivation of latent infection. Arthritis Rheum 57: Bean AG, Roach DR, Briscoe H et al (1999) Structural deficiencies in granuloma formation in TNF gene-targeted mice underlie the heightened susceptibility to aerosol Mycobacterium tuberculosis infection which is not compensated for by lymphotoxin. J Immunol 162: Flynn JL, Goldstein MM, Chan J et al (1995) Tumor necrosis factor-alpha is required in the protective immune response against Mycobacterium tuberculosis in mice. Immunity 2: Keane J (2001) Tuberculosis associated with infliximab, a tumor necrosis factor α-neutralizing agent. N Engl J Med 345: Joven BE, Almodovar R, Galindo M, Mateo I, Pablos JL (2006) Does anti-tumour necrosis factor a treatment modify the tuberculin PPD response? Ann Rheum Dis 65: Menzies D (1999) Interpretation of repeated tuberculin tests. Boosting, conversion, and reversion. Am J Respir Crit Care Med 159: Ponce de Leon D, Acevedo-Vasquez E, Sanchez-Torres A et al (2005) Attenuated response to purified protein derivative in patients with rheumatoid arthritis: study in a population with a high prevalence of tuberculosis. Ann Rheum Dis 64: Hatemi G, Melikoglu M, Fresko I, Masatlioglu S, Tascilar K, Yazici H (2006) Infliximab does not suppress the tuberculin skin test (purified protein derivative. J Rheumatol 34: Fuchs-Polishuk I, Avnon LS, Abu shakra M (2007) Tuberculosis among patients treated with TNF-alpha blockers. Harefuah 146: Elkayam O, Balbir-gurman A, Lidgi M, Rahav G, Weiler-Ravel D (2007) Guidelines of the Israeli association of rheumatology for prevention of tuberculosis in patients treated with TNF-alpha blockers. Harefuah 146: Al Zahrani K, Al Jahdali H, Menzies D (2000) Does size matter? Utility of size of tuberculin reactions for the diagnosis of mycobacterial disease. Am J Respir Crit Care Med 162: Mow SW, Abreu-martin MT (2004) High incidence of anergy in inflammatory bowel disease patients limits the usefulness of PPD screening before infliximab therapy. Clin Gastroenterol and Hepatol 2: Wang L, Turner MO, Elwood RK et al (2002) A meta-analysis of the effect of Bacille Calmette Guerin vaccination on tuberculin skin test measurement. Thorax 57: American Thoracic Society (2000) Targeted tuberculin testing and treatment of latent tuberculosis infection. Am J Resp Crit Care Med 161:S Berg L, Lampa J, Rogberg S, van Vollenhoven R, Klareskog L (2001) Increased peripheral T cell reactivity to microbial antigens and collagen type II in rheumatoid arthritis after treatment with soluble TNFalpha receptors. Ann Rheum Dis 60: Vanhoof J, Landewe S, Van Wijngaerden Geusens P (2003) High incidence of hepatotoxicity of isoniazid treatment for tuberculosis chemoprophylaxis in patients with rheumatoid arthritis treated with methotrexate or sulfalazine and anti-tumour necrosis factor inhibitors. Ann Rheum Dis 62: Andersen P, Munk ME, Pollock JM, Doherty TM (2000) Specific immune-based diagnosis of tuberculosis. Lancet 356: Mazurek G, Jereb J, Lobue P, Iademerco M, Metchock B, Vernon A (2005) Guidelines or using the QuantiFERON-TB test for detecting Mycobacterium tuberculosis infection, United States. MMWR Morb mortal Wkly Recomm Rep 54 (RR15): Ferrara G, Losi M, D Amico R et al (2006) Use in routine clinical practice of two commercial blood tests for diagnosis of infection with Mycobacterium tuberculosis: a prospective study. Lancet 367: Pratt A, Nicholl K, Kay L (2007) Use of QuantiFERON TB Gold test as part of a screening program in patients with RA under consideration for treatment with anti-tnf-alpha agents: the Newcastle experience. Rheumatology 46: Centers for Diseases Control and Prevention. Guidelines for Using the QuantiFERON -TB Test for Diagnosing Latent Mycobacterium tuberculosis Infection. mmwrhtml/rr5202a2.htm