Potential Role for IL-2 ELISpot in Differentiating Recent and Remote Infection in Tuberculosis Contact Tracing Authors: Benjamin Krummel, Alan Strassburg, Martin Ernst, Norbert Reiling, Barbara Eker, Heidrun Rath, Robert Hoerster, Waltraud Wappler, Andrea Glaewe, Volker Schoellhorn, Giovanni Sotgiu, Christoph Lange Published: 2010, PLOS ONE DOI: 10.1371/journal.pone.0011670 ABSTRACT Interferon (IFN)-γ release assays (IGRA) have improved tuberculosis contact tracing, but discrimination of recent from remote Mycobacterium tuberculosis contacts is not possible by IGRA alone. We present results of a tuberculosis contact investigation with a new early-secretory-antigenic-target (ESAT)-6 and culture-filtrate-protein (CFP)-10 specific interleukin (IL)-2 ELISpot in addition to ESAT-6 and CFP-10 specific IFN-γ ELISpot and tuberculin skin testing (TST). Results of the TST, IFN-γ ELISpot and IL-2 ELISpot were positive in 6/172 (3.4%), 7/167 (4.2%) and 6/196 (3.1%) of contacts, respectively. Close contact (≥100 hours) to the index case increased the risk of positive results in the IFN-γ ELISpot, TST, and IL-2 ELISpot by 40.8, 19.3, and 2.5 times, respectively. Individuals with a positive IFN-γ ELISpot/negative IL-2 ELISpot result had a median (IQR) duration of index case exposure of 568 hours (133_1000) compared to individuals with a positive IFN-γ ELISpot/positive IL-2 ELISpot result (median = 24 hours; 20_130; p-value = 0.047). Combination of a M. tuberculosis specific IFN-γ ELISpot with a M. tuberculosis specific IL-2 ELISpot significantly improved the identification of individuals with the highest risk of recent M. tuberculosis infection and is a promising method that should be explored to target tuberculosis preventive chemotherapy. INTRODUCTION Identification and treatment of individuals with lasting Mycobacterium tuberculosis specific immune responses (latent tuberculosis infection -, LTBI) following recent contact to a contagious index case with pulmonary tuberculosis are crucial to tuberculosis control [1] . The diagnosis and management of LTBI has evolved over the last decade with improvements in the immunodiagnosis by M. tuberculosis specific interferon-γ release assays (IGRA) in addition to the tuberculin skin test [2] . In principle, IGRA detect interferon-γ release by mononuclear cells in the ELISpot (T-SPOT. TB – Oxford Immunotec, Abingdon, UK) or in whole blood in the ELISA (QuantiFERON-TB GOLD in tube; QFT-G-IT; Cellestis, Carnegie, Australia) following ex vivo contact with antigens that are naturally encoded in the M. tuberculosis genomic region of difference (RD)-1 and RD11 (only QFT-G-IT). The genes encoding for these proteins are not present in most non tuberculous mycobacteria and are deleted from the genome of M. bovis Bacille Calmette Guerin (BCG) strains that are used for tuberculosis vaccinations [3] , [4] . The population of T lymphocytes responding rapidly with the production of IFN-γ following ex vivo stimulation with these antigens belongs predominantly to the pool of CD4+ CD45RA-CCR7- effector memory cells that likely had recent antigen exposure [5] . In contrast, long-lived central memory T-cells may persist even after successful treatment of tuberculosis [6] , [7] and are less likely to release IFN-γ after short incubation with antigen while they are more likely to produce IL-2 upon antigen exposure than effector cells [7] . Thus, assaying M. tuberculosis -specific T cell function defined solely by the quantification of IFN-γ secretion after short term ex vivo incubation with M. tuberculosis -specific antigen may prove to be an insufficient indicator for recent inhalation exposure to M. tuberculosis . In contrast, simultaneous measures of other T cell function, such as M. tuberculosis -specific stimulation of IL-2 secretion may improve the discrimination of active versus remote LTBI [7] . A 28-year-old schoolteacher was referred to our hospital for diagnostic evaluations for presumptive tuberculosis. Between March and October 2007 he had consulted seven different general practitioners complaining of progressive cough and weight loss until finally in November 2007 a chest x-ray was performed and a cavitating lesion in the left upper lobe of the lung was demonstrated. In our hospital, alcohol acid fast bacilli (AAFB) were readily identified on sputum smears. The bacilli were identified as M. tuberculosis by culture and susceptibility testing revealed antibiotic drug resistance to isoniacid and streptomycin. We estimated by the presence of his symptoms that he had been infectious for a period of seven months prior to hospital admission. During this period, he had exposure to a large number of close contacts including pupils and teachers from two different schools and members of a church parish. We describe the results of tuberculosis contact tracing among well-defined contacts of a highly symptomatic sputum smear positive index case, comparing a novel M. tuberculosis antigen specific IL-2 ELISpot assay with a M. tuberculosis antigen specific IFN-γ ELISpot and the TST for the diagnosis of LTBI. This outbreak provided an opportunity to test the hypothesis that simultaneous measures of M. tuberculosis -specific IFN-γ and IL-2 secretion may improve identification of individuals with the highest risk of recent infection with M. tuberculosis . MATERIALS AND METHODS Individuals with close contact to the symptomatic index case during the period of March 2007 until October 2007 were identified by the municipal health care department of the city of Luebeck, Germany. Contacts were asked to estimate their exposure time to the index case in hours per week. We identified a group of close (with a total exposure of 100 hours or more) and another of occasional contacts (exposure of less than 100 hours). For 76 contacts who did not provide any information about exposure to the index case, the time was estimated according to the average exposure of classmates. In addition information about BCG vaccination status, possible former contact to tuberculosis, and pre-existing illness and medication were obtained. Data were recorded on a standardized questionnaire. All adults and children who were enrolled agreed by written informed consent to participate in the study. In addition, the agreement of both legal guardians was obtained for the children who were enrolled. The study was approved by the ethical committee of the University of Luebeck, Germany. TUBERCULIN SKIN TEST (TST) Tuberculin skin testing was performed by intracutaneous injection of 2U/0.1ml of tuberculin RT23 (Statens Serum Institut, Copenhagen, Denmark) on the volar aspect of the forearm. The TST was administered and read by experienced staff from the Municipal Health Authority. Indurations were measured at 48–72 hours by two independent readers [8] . Parallel to TST testing blood was taken from contacts after informed consent following an eight week period after the last exposure to the index case [9] . Chest radiographs of the TST-positive contacts were performed, and clinical and radiographic findings were reviewed for evidence of active tuberculosis. ELISPOT ASSAY ELISpot assays for IFN-γ and separately for IL-2 release by mononuclear cells were performed using test-plates from AID, Strassberg, Germany. Briefly, 250,000 PBMCs were plated overnight on 96-well plates, which had been pre-coated with a mouse anti-human IFN-γ or IL-2 antibody, in 200 µl culture medium per well. The cells were left unstimulated (negative control), were stimulated with 10 ng/ml anti-CD3 monoclonal antibody (clone X35, Beckman-Coulter, Krefeld, Germany; positive control) and with 5 µg/ml of ESAT-6 and CFP-10 peptides (kindly provided by AID, Strassberg, Germany). Culture of the plates, washing, counterstaining, visualisation and analysis of the spots was performed according to the manufacturer guidelines. The response of stimulated cultures was considered positive when the test well contained >5 spots and had at least twice the number of spots than the negative control well. The background number of spots in negative control wells was always less than five spots per well. The response of stimulated cultures was considered negative when the test well contained ≤5 spots or less than twice the number of spots in the unstimulated control well and when the positive control well contained at least 50 spots at the same time. ANALYSIS OF HOST ACTIVATING POTENTIAL OF THE INDEX CASES M. TUBERCULOSIS STRAIN Human THP1 cells were differentiated for 96h in the presence of 5 ng/ml PMA. Human monocyte-derived macrophages (MDM) were generated as previously described [10] . Mycobacterium tuberculosis strains H37Rv (ATCC 27294) and the index cases M. tuberculosis strain (9729/07) were grown at 37°C in Middlebrook 7H9 broth supplemented with Middlebrook OADC-Enrichment (BD Biosciences) to an OD of 0.2–0.4. Cultures were harvested, aliquoted, and frozen at −80°C. Viable cell counts were determined by plating serial dilutions of cultures on Middlebrook 7H10 agar plates (BD Biosciences). For experimental in vitro infections, aliquots were diluted in culture media and the preparation was passed six times through a 27-gauge needle to ensure proper dispersion of mycobacteria. THP-1 (3×10 5 /ml) cells and MDM (5×10 5 ) cells were stimulated in 500 µl cultures for 4h in the absence or presence of Mtb strains H37Rv or 9729/07 at the indicated multiplicity of infection (MOI). Bacterial lipopolysaccharide (LPS) of Salmonella enterica , serotype friedenau H909, was kindly provided by Prof. Dr. H. Brade (Research Center Borstel, Borstel, Germany) and used as positive control. Supernatants were frozen at −80°C until analysis. The formation of TNF-α was analyzed by ELISA (TNF Duoset, R&D Systems), as recommended by the manufacturer. STATISTICAL ANALYSIS Data were analysed using Stata 9.0 (StataCorp, Stata Statistical Software Release 9, College Station, TX, USA, 2005). Categorical data were compared using the Chi-squared test (or Fisher's exact test, when expected cell sizes were smaller than five). The Wilcoxon-Mann-Whitney test was used for continuous measurements to test relationships in unpaired analysis, when assumed that the dependent variable is a not normally distributed interval variable. Since there is no gold standard for the determination of LTBI, the analysis was focussed on estimating the strength of association between the degree of exposure (the duration and proximity of contact with the infectious case) and the test results. Logistic regression analysis was performed in order to evaluate which covariates have had an association with a positive test result. Test concordance was assessed by Cohen's κ-statistics with agreement considered slight’for k≤0.2, ‘fair’ for 0.2