Journal of Infectious Diseases Advance Access published August 2, 2013

Transcription

1 Journal of Infectious Diseases Advance Access published August 2, Reversion and Conversion of Interferon-gamma Release Assay Results in HIV-1 Infected Individuals Maximilian C. Aichelburg 1,*, Thomas Reiberger 2, Florian Breitenecker 1, Mattias Mandorfer 2, Athanasios Makristathis 3, and Armin Rieger 1 1 Department of Dermatology, Division of Immunology, Allergy and Infectious Diseases (DIAID) 2 Department of Internal Medicine III, Division of Gastroenterology & Hepatology, Vienna HIV & Liver Study Group 3 Department of Hygiene and Medical Microbiology, Division of Clinical Microbiology, Medical University of Vienna, Vienna, Austria * Corresponding author and requests for reprints. Mailing address: Department of Dermatology, Division of Immunology, Allergy and Infectious Diseases, Vienna General Hospital, Medical University of Vienna, Waehringer Guertel 18-20, A-1090 Vienna, Austria, Phone: /4232, Fax: /4233, maximilian.aichelburg@meduniwien.ac.at The Author Published by Oxford University Press on behalf of the Infectious Diseases Society of America. All rights reserved. For Permissions, please journals.permissions@oup.com

2 2 ABSTRACT In this prospective study HIV-1 infected subjects underwent QuantiFERON-TB Gold In- Tube interferon-gamma release assay (IGRA) testing at baseline and after 24 months in a low TB-incidence country. Concordant baseline and follow-up results were observed in 86% (686/794) of subjects. IGRA conversions occurred in 9% (63/718) while IGRA reversions were seen in 33% (25/76) of individuals. Of the 10 active TB cases during follow-up 5 had concordant positive IGRA results and 2 were IGRA converters. Although the clinical significance of IGRA conversions and reversions remains to be established, repeated IGRA testing seems to be of value in HIV-1 infected individuals.

3 3 INTRODUCTION Individuals with human immunodeficiency virus type 1 (HIV-1) infection are at increased risk of latent tuberculosis infection (LTBI), subsequent active tuberculosis (TB) reactivation and develop more severe clinical disease manifestations [1] and thus represent an important target group for LTBI testing and treatment. For more than a century, the tuberculin skin test (TST) has been the most extensively used immune-based test for detecting LTBI [2]. More recently, laboratory T-cell-based interferon- release assays (IGRAs) have been developed as an alternative in vitro immunodiagnostic approach to the TST possessing logistical conveniences [3]. Despite the fact that their performance is dependent on the actual CD4 + T cell count, these tests have been demonstrated to detect LTBI with higher reliability than the TST in individuals infected with HIV-1 [4]. The interpretation of serial TST results may be complicated by non-specific variations in test results including boosting, conversions (negative to positive test), and reversions (positive to negative test) upon repetitive testing [5]. Serial testing with IGRA is attractive since it avoids subjective measurements, can be repeated without sensitization and boosting in subsequent tests, and requires only a single patient visit. However, there is only limited data on the reproducibility of IGRAs, particularly with regard to within-person variability of T-cell responses during serial testing. Thus, the updated CDC guidelines 2010 ask for more research on the serial testing performance of IGRAs, especially in individuals with HIV-1 infection [6]. The current study was performed in a cohort of HIV-1 infected individuals in a country with a low risk of TB re-infection after inclusion. The objectives of this prospective, single-center study were to determine (i) the incidence of IGRA reversions and conversions in a cohort of HIV-1-infected individuals, and to identify (ii) socio-epidemiological and clinical risk factors associated with reverted or converted IGRA results.

4 4 MATERIAL AND METHODS Study setting and recruitment. This study is an extension of a project initiated in 2006 to evaluate the QuantiFERON-TB Gold In-Tube IGRA for routine use in HIV-1 infected individuals in a low-incidence country [7]. Briefly, HIV-1 infected patients aged 18 years attending the HIV-outpatient clinic of the Vienna General Hospital who voluntarily underwent IGRA-testing were prospectively enrolled. Written informed consent for inclusion was provided by all participants. Patients with current active TB disease were excluded. The study protocol was approved by the ethics committee of the Medical University of Vienna, Austria. Laboratory assays. Venous blood samples were drawn in three QuantiFERON-TB Gold In-Tube evacuated tubes, precoated with either M. tuberculosis-specific antigens (ESAT-6, CFP-10, TB7.7), phytohemaglutinin for the positive control or no antigen (nil) for the negative control. Within 4 hours of drawing specimens were processed according to the manufacturer s instructions as described elsewhere [8]. The IGRA was done at baseline and repeated after approximately 2 years using identical protocols. In accordance with CDC guidelines IGRA conversion was defined as baseline IFN- < 0.35 IU/ml and follow-up IFN IU/ml whereas IGRA reversion was defined as baseline IFN IU/ml and follow-up IFN- < 0.35 IU/ml [6]. All subjects underwent chest x-ray screening for TB at least once a year and were clinically monitored for TB-related clinical symptoms as described elsewhere [7]. In case of IGRA positivity subjects were recommended to start both ART and TB chemoprevention. Statistical analyses. Statistical analyses were conducted employing Statistica version 6.0 for Windows (Statsoft, Hamburg, Germany). Median values were compared using the non-parametric Wilcoxon-Mann-Whitney U rank-sum test. Fisher s exact test was used to evaluate differences in proportions. The inter-test agreement of serial dichotomized IGRA testing was assessed by Cohen s kappa ( ) statistics. All p-values (P) were two-tailed and P < 0.05 was considered to denote statistical significance.

5 5 RESULTS Baseline characteristics of the study participants. A total of 1525 HIV-1 infected individuals were asked to participate in this study. Seven patients declined participation, 13 were diagnosed with active TB at baseline and in 659 individuals no repeat IGRA was performed. Finally, a total of 846 subjects (representative of the entire study population) with both baseline and repeat IGRA results were included in this study. The mean age was 39.6 ±11.5 years, 71% (597/846) of the subjects were male and 29% (249/846) were female. The majority of individuals were of white ethnicity (86%, 730/846). At enrollment, 53% (446/846) of the patients were on ART. In 46% (387/846) of the subjects HIV-1 RNA levels were undetectable (< 50 copies/ml). The mean actual and nadir CD4 + T cell counts were 413 ± 248 and 227 ± 184 cells/mm 3, respectively. Previous TB disease had been diagnosed in 4% (30/846) of individuals. Concordance of baseline and follow-up IGRA results. At baseline, the IGRA yielded positive results in 9% (76/846), negative results in 85% (718/846) and indeterminate results in 6% (52/846) of subjects, respectively. After a mean of days the repeat IGRA was performed. Of the 76 patients initially tested positive, 58% (44/76) were concordantly positive, and 9% (7/76) were indeterminate on serial testing, respectively. In 33% (25/76) of patients the serial IGRA result was negative and thus, a reversion of IGRA result was observed. At follow-up testing of the 718 patients with a negative IGRA at baseline, 89% (642/718) individuals remained IGRA negative whereas 9% (63/718) converted to a positive IGRA. Among the 52 subjects with indeterminate baseline IGRA results, the second IGRA was positive in 3, negative in 43 and indeterminate in 6 individuals. The inter-test agreement of serial QFT testing was moderate as indicated by a Cohan κ-coefficient of (95%CI: ). The observed inter-test agreement was (95% CI: ) while the probability of random agreement of serial QFT testing was calculated as (95% CI: ) (Table 2). When applying a more stringent criteria for IGRA conversion (follow-up IFN- response > 0.70 IU/ml) 30% (19/63) of converters would have tested negative at follow-up. Changes in quantitative IFN- responses for converters and reverters are shown in supplementary table 1. Although not part of the initial study protocol a third IGRA was performed in 60% (38/63) of converters showing negative IGRA results in 55% (21/38) thereof. These 21 patients had a lower IFN- response than those who remained IGRA

6 6 positive (median IFN- value 0.75 (IQR: ) IU/ml vs (IQR: ) IU/ml). Independent predictors for IGRA conversion and reversion. In multivariate regression analysis there were no statistical significant differences between converters and non-converters in terms of sex, age and HIV-1-related parameters including actual and nadir CD4 + T cell counts, HIV-1 RNA level, ART status and prior AIDS manifestation. 63.5% (40/63) of patients with a converted IGRA compared to 21.6% (169/783) of non-converters originated from high TB incidence country according to the WHO (OR, 2.3; P = 0.015). Patients with a converted IGRA were more likely having acquired HIV-1 infection by intravenous drug abuse (28.6% vs. 17.9%; OR, 2.5; P = 0.012) (Table 2). Adjusted analysis revealed the following risk factors of reverted IGRA results: male sex (OR, 15.9; P = 0.004) and origin from a high TB-incidence country (OR, 1.9; P = 0.019). Age, mode of infection with HIV-1, actual and nadir CD4 + T cell count and ART were not associated with a reverted IGRA result (Table 2). There was no difference in terms of intra-person variability of CD4 + T cell count at baseline and follow-up testing among reverters and converters. Cases of active tuberculosis during follow-up. During the observational period 10 individuals (1.2%; 10/846) were diagnosed active TB with a median delay of 435 days (IQR= ) from inclusion. None of these subjects reported to have had contact to an active TB case or traveled to a country with high TB-prevalence between study inclusion and TB diagnosis. Five patients suffered from pulmonary TB, 3 from extrapulmonary TB and 2 from miliary TB. Seven of the 10 active TB cases had culture-confirmed disease. The median baseline IFN- release was significantly higher in the 7 IGRA positive patients with subsequent active TB as compared to the remaining 69 IGRA positive individuals who had not developed active TB (7 IU/ml vs. 2 IU/ml; P = ). Active TB was observed in 11% (5/44) of subjects with concordant positive IGRA results. Only one patient progressed to active TB despite showing negative IGRA results both at baseline and at follow-up. Two subjects with TB were IGRA converters whereas one had a reverted IGRA. DISCUSSION HIV-1-infected subjects represent a high-risk group for active TB development for which an accurate detection of LTBI is of paramount importance. Although broadly recommended and increasingly used, data on the interpretation of IGRA results in serial testing is scarce. In this

7 7 prospective, longitudinal study on HIV-1 infected individuals in a low TB-incidence country followed for 3 years, repeated IGRA testing showed promising results. Although inter-test agreement of serial IGRA testing was moderate, more than 86% of patients had concordant IGRA results at baseline and at follow-up. These results are consistent with previously published data in HIV-uninfected populations [9]. In terms of risk factors for converted IGRA results, we found that origin from a high TB-incidence country according to WHO and injection drug abuse were associated with IGRA conversion. Interestingly, origin from a TB incidence country was also associated with IGRA reversion. However, a rationale with respect to biological plausibility is lacking. Only a limited number of longitudinal studies assessing the value of IGRAs in serial testing are available. The studies already published were predominantly performed in either contact [10] or health-care worker investigations [11, 12]. Among contacts of active TB cases, IGRA conversions and reversions occurred with conversion rates depending on the test and definitions used [10]. Conversions, reversions, and nonspecific variations leading to discordance were observed with serial IGRA testing just as they do with TST in a study on health-care workers in India [11]. A recent study by Fong et al. showed that the clinical significance of IGRA conversions in serial testing remains a challenging task for clinicians and that the use of single cut-off point criteria for IGRA may lead to a significant number of false-positive results and overdiagnosis of LTBI [12]. So far, no study prospectively evaluated the performance of an IGRA in serial testing of HIV- 1-infected patients in a TB low-incidence country. To the best of our knowledge, this extension of our single-center study is the first to assess the serial testing performance of IGRA in a large cohort of individuals with HIV-1 infection. Gray et al. recently identified false-positive IGRA results by repeat testing (within 40 days) in HIV-1-infected subjects in a low TB incidence setting (Denver, USA). However, the retrospective study design and the fact that only individuals with a positive IGRA result were presented in this study may have introduced a bias [13]. Although the reversion rate observed in our study was not as high as the 72% observed in the US [13] still over one third of initially IGRA positive patients tested negative at follow-up. Consistently, IGRA reversion of unknown origin has been described in healthcare workers with rates of IGRA reversion ranging from 25 to 41% [14]. Thus, it might be conceivable to retest HIV-1-infected patients with positive IGRA results with no tuberculosis exposure risks

8 8 to prevent unnecessary TB chemoprevention and possible drug-drug interactions. This strategy is in accordance with Centers for Disease Control and Prevention guidelines [6]. We have to acknowledge the potential limitations of this study including its single-center nature. Furthermore, due to the fact that this study was no contact investigation, anamnesis regarding previous contact to an active TB index case was not part of the initial study protocol and is therefore missing. In conclusion, we provide further evidence that IGRAs is of clinical value both in immunocompetent individuals and in patients with moderate immunodeficiency. Due to objective read out and independence on patient compliance, IGRAs have the potential to replace the TST for the identification of HIV-1-infected individuals at risk for the development of active TB. However, IFN- variability should be kept in mind when interpreting repeated IGRA results. Acknowledgements The authors would like to thank Regina Aichwalder, Silvia Reichholf and Alexander Riegler for excellent technical assistance and the patients who participated in this study. M.C.A., F.B., and A.R. designed the study, enrolled patients and contributed to data collection. M.C.A., T.R. and M.M. performed data analysis. M.C.A., T.R., F.B., M.M., A.M., and A.R. participated in data interpretation. M.C.A. wrote the article. All authors have read and approved the final manuscript. Potential conflicts of interest All authors: no conflicts. Financial support None.

9 9 FOOTNOTE PAGE Potential conflicts of interest All authors: no conflicts. Financial support None. Presented in part 52 th Interscience Conference on Antimicrobial Agents and Chemotherapy, 9-12 September 2012, San Francisco, USA abstract H Corresponding author and requests for reprints Dr. Maximilian C. Aichelburg Department of Dermatology, Division of Immunology, Allergy and Infectious Diseases, Medical University of Vienna Waehringer Guertel A-1090 Vienna Austria Phone: /7705 Fax: / Alternate corresponding author Dr. Armin Rieger Department of Dermatology, Division of Immunology, Allergy and Infectious Diseases, Medical University of Vienna Waehringer Guertel A-1090 Vienna Austria Phone: /7705 Fax: /

10 10 REFERENCES 1. Kaufmann SH, McMichael AJ. Annulling a dangerous liaison: vaccination strategies against AIDS and tuberculosis. Nat Med 2005; 11:S Huebner RE, Schein MF and Bass JB, Jr. The tuberculin skin test. Clin Infect Dis 1993; 17: Andersen P, Munk ME, Pollock JM and Doherty TM. Specific immune-based diagnosis of tuberculosis. Lancet 2000; 356: Rangaka MX, Wilkinson KA, Seldon R, et al. Effect of HIV-1 infection on T-Cell-based and skin test detection of tuberculosis infection. Am J Respir Crit Care Med 2007; 175: Menzies D. Interpretation of repeated tuberculin tests. Boosting, conversion, and reversion. Am J Respir Crit Care Med 1999; 159: Mazurek GH, Jereb J, Vernon A, LoBue P, Goldberg S and Castro K. Updated guidelines for using Interferon Gamma Release Assays to detect Mycobacterium tuberculosis infection - United States, MMWR Recomm Rep 2010; 59: Aichelburg MC, Rieger A, Breitenecker F, et al. Detection and prediction of active tuberculosis disease by a whole-blood interferon-gamma release assay in HIV-1-infected individuals. Clin Infect Dis 2009; 48: Aichelburg MC, Tittes J, Breitenecker F, Reiberger T, Kohrgruber N and Rieger A. Prognostic value of indeterminate IFN-gamma release assay results in HIV-1 infection. J Clin Microbiol 2012; 50: Baker CA, Thomas W, Stauffer WM, Peterson PK and Tsukayama DT. Serial testing of refugees for latent tuberculosis using the QuantiFERON-gold in-tube: effects of an antecedent tuberculin skin test. Am J Trop Med Hyg 2009; 80: Pai M, Joshi R, Dogra S, et al. T-cell assay conversions and reversions among household contacts of tuberculosis patients in rural India. Int J Tuberc Lung Dis 2009; 13: Pai M, Joshi R, Dogra S, et al. Serial testing of health care workers for tuberculosis using interferon-gamma assay. Am J Respir Crit Care Med 2006; 174: Fong KS, Tomford JW, Teixeira L, et al. Challenges of interferon-gamma release assay conversions in serial testing of health-care workers in a TB control program. Chest 2012; 142:55-62.

11 Gray J, Reves R, Johnson S and Belknap R. Identification of false-positive QuantiFERON-TB Gold In-Tube assays by repeat testing in HIV-infected patients at low risk for tuberculosis. Clin Infect Dis 2012; 54:e Zwerling A, van den Hof S, Scholten J, Cobelens F, Menzies D and Pai M. Interferongamma release assays for tuberculosis screening of healthcare workers: a systematic review. Thorax 2012; 67:62-70.

12 TABLE 1. AGREEMENT OF BASELINE AND FOLLOW-UP IGRA RESULTS IGRA 1 IGRA 2 N (%) Negative Negative 642 (82.9%) Negative Positive 63 (8.1%) Positive Positive 44 (5.8%) Positive Negative 25 (3.2%) inter-test agreement (95% CI: ) Cohan κ-coefficient (95% CI: ) probability of random agreement (95% CI: ) NOTE. CI = confidence interval; IGRA = interferon-gamma release assay 12

13 TABLE 2. RISK FACTORS ASSOCIATED WITH CONVERSION AND REVERSION OF IGRA RESULTS BY MULTIVARIATE REGRESSION ANALYSIS Patients All Non-Converters Converters Non-Reverters Reverters P N=846 N=783 N=63 N=821 N=25 P Sex (m/f, % male) 597/249 (70.6%) 556/227 (71.0%) 41/22 (65.1%) /248 (69.8%) 24/1 (96.0%) (OR: 15.9; 95% CI ) Age (years) 39.6 ± ± ± ± ± Born in a country with the indicated TB burden according WHO 25 cases per 100,000 population (n; %) 209 (24.7%) 169 (21.6%) 40 (63.5%) (OR: 2.3; 95% CI ) 194 (23.6%) 15 (60.0%) (OR: 1.9; 95% CI ) Mode of infection with HIV-1 Heterosexual contact (n; %) 334 (39.5%) 312 (39.8%) 22 (34.9%) (39.1%) 13 (52%) Homosexual contact (n; %) 329 (38.9%) 307 (39.2%) 22 (34.9%) (39.0%) 9 (36%) Injection drug use (n; %) 158 (17.1%) 140 (17.9%) 18 (28.6%) (19.0%) 2 (8.0%) (OR: 2.5; 95% CI ) CD4 + nadir (cells/µl) 227 ± ± ± ± ± Enrollment CD4 + (cells/µl) 413 ± ± ± ± ± Enrollment HIV-RNA (log copies/ml) 3.16 ± ± ± ± ± On ART (n; %) 446 (52.7%) 415 (53.0%) 31 (49.2%) (53.0%) 11 (44.0%) AIDS (n; %) 182 (21.5% 172 (22.0%) 10 (15.9%) (19.0%) 26 (28.0%) Previous TB (n; %) 30 (3.5%) 27 (3.4%) 3 (4.7%) (3.5%) 1 (4.0%) NOTE. ART = antiretroviral therapy; CI = confidence interval; IGRA = interferon-gamma release assay; OR = odds ratio; TB = tuberculosis Downloaded from at Pennsylvania State University on May 9, 2016