Detection of latent tuberculosis infection in hemodialysis patients: Comparison between the quantiferon-tuberculosis gold test and the tuberculin skin test

Background Tuberculosis (TB) remains an important cause of morbidity and mortality in hemodialysis (HD) patients. A gold standard for the diagnosis of latent tuberculosis infection (LTBI) is lacking. Objective The aim of this study was to compare the diagnostic utility of the QuantiFERON-Tuberculosis Gold (QFT-G) test with the tuberculin skin test (TST) in identifying LTBI in patients with end-stage renal disease (ESRD) on HD. Study design The present study had a prospective design. Patients and methods A total of 74 patients with ESRD on HD without active TB and other immunosuppressive conditions were tested for LTBI by the QFT-G test and the TST. Results LTBI, as estimated by the QFT-G test and TST, was detected in 35.1 and 13.5% of the HD patients, respectively; 37.8% of patients were positive for the QFT-G test and/or the TST. There was a poor agreement between QFT-G test and TST results in patients with ESRD on HD (QFT-G test vs. TST: κ=0.25, 95% confidence interval=0.12–0.37). TST was positive in 2.7% of patients when the QFT-G test was negative, and it was negative in 24.3% of patients when the QFT-G test was positive. There was no significant difference in duration of HD or creatinine levels between QFT-G-positive and QFT-G-negative patients (P=0.08 and 0.2, respectively). TST-positive patients had a significantly shorter duration of HD and lower creatinine levels than TST-negative patients (P=0.001 and 0.01, respectively). Conclusion In patients with ESRD and on HD, LTBI cannot be simply ruled out with a negative TST result, but rather a QFT-G test is recommended. Screening and treatment of LTBI should be carried in dialysis patients, aiming to prevent progression to active TB and secondary infection of others.


Introduction
In Egypt, tuberculosis (TB) is an important health problem. Egypt is ranked among the mid-level incidence countries [1]. The reported prevalence of chronic renal failure is 225 patients per million in Egypt [2]. Individuals with latent tuberculosis infection (LTBI) are assumed to have viable TB bacilli in their body. These bacilli are dormant, but have the potential to reactivate and cause disease [3]. According to the WHO, ∼2-3 billion people in the world have LTBI, and 5-15% of them will suffer from reactivation of TB during their life. Therefore, the treatment for LTBI influences the future global prevention of TB infection [4].
Patients with end-stage renal disease (ESRD) and on dialysis are 6-25 times more likely to develop TB than the general population, mainly because of the impaired cellular immunity associated with this condition [5]. The mortality rate of TB in dialysis patients is high in comparison with the general population, ranging from 17 to 75% [6]. In these patients, the diagnosis of TB is often difficult because of nonspecific symptoms and prevailing extrapulmonary involvement [4,[7][8][9]. Diagnosis of LTBI mainly depends on the immune reaction of the host rather than that of the bacterium itself, because the amount of Mycobacterium tuberculosis is small in LTBI individuals. A gold standard for the diagnosis of the LTBI is lacking. At present, there are two screening tests for LTBI: the tuberculin skin test (TST) and interferon-γ release assays (IGRAs, including the QuantiFERON-Tuberculosis Gold (QFT-G) and the T-SPOT.TB test) [9].

Aim
The aim of the present study was to compare the diagnostic utility of QFT-G test with the TST in identifying LTBI in patients with ESRD on hemodialysis (HD).

Patients and methods
This prospective study included adult patients with ESRD on HD, who were recruited prospectively from the outpatient HD unit at the Sohag University Hospital, Sohag, Egypt, in the period between February and April 2016. All patients enrolled were required to provide their consent. The study was approved by the Ethics Committee at Sohag Faculty of Medicine. Exclusion criteria were presence of active TB disease and other immunosuppressive diseases such as HIV, diabetes mellitus, immunological disorders, long duration of corticosteroid treatment, and chronic liver diseases.
All patients were subjected to the following: (1) Complete history taking and examination, including age, sex, comorbid conditions, duration of dialysis, and history of Bacille Calmette-Guérin (BCG) vaccination (visual inspection for BCG scar). (2) Chest radiograph.
(3) Laboratory investigations included the following: erythrocytic sedimentation rate test, serum creatinine test, and sputum analysis for acid-fast bacillifor 3 successive days. (4) TST was performed by injecting 0.1 ml of tuberculin purified protein derivative (PPD) (5 U) intradermally into the skin of the forearm. The test was then read 48-72 h later. On the basis of published guidelines, induration greater than or equal to 10 mm was considered a positive TST for ESRD patients on dialysis [3].
QFT-G test in-tube: 1 ml of blood was collected by venipuncture directly into each of the collection tubes for the QFT-G test. The QFT-G system uses three specialized blood-collection tubes that contain antigens representing certain M. tuberculosis proteins (ESAT-6, CFP-10, and TB 7.7) as well as positive (mitogen) and negative (nil) controls. The mitogen tube can be used with the QFT-G test as a positive control. This may be especially warranted where there is doubt regarding the individual's immune status, and therefore it is specially used in immunocompromised patients. The following steps were performed according to the package insert of QFT-G (in-tube method), manufactured by Cellestis Limited (Hilden, Germany) [10]. The concentration of interferon-γ (IFN-γ) was determined by enzyme-linked immunosorbent assay.

Statistical analysis
Statistical analyses were performed using SAS software (SAS Institute Inc., Cary, NC, USA). Values are reported as mean±SD. A P value (two-tailed) less than 0.05 was considered statistically significant. Agreement between the results of the QFT-G test and the results of TST was evaluated using κ statistics, where κ less than 0.4 represents poor agreement, κ values from 0.4 to 0.75 represent fair to good agreement, and κ greater than 0.75 represents an excellent agreement [11].

Results
According to inclusion and exclusion criteria, the study included 74 adult patients with ESRD on HD. The mean±SD age of the patients was 45±18 years. Fortyfive (60.8%) participants were males and 29 (39.2%) were females. Forty-seven (63.5) patients were BCG vaccinated. The mean±SD duration of HD was 28±19 months, as shown in Table 1.
Of the 74 participants in the present study, 26 (35.1%) patients had a positive QFT-G test. There were no significant differences in duration of HD and creatinine levels between QFT-G-positive and QFT-G-negative patients (P=0.08 and 0.2, respectively), as shown in Table 2.
Of the 74 participants in the present study, 10 (13.5%) had positive TST results. Patients with positive TST results had a significantly shorter duration of HD and lower creatinine levels than patients with negative TST (P=0.001 and 0.01, respectively), as shown in Table 3.
Of the 74 participants, 28 (37.8%) patients had positive QFT-G and/or positive TST results. Twenty-six (35.1%) patients had positive QFT-G results, whereas 10 (13.5%) patients had positive TST results. There was  a poor agreement between QFT-G and TST results in patients with ESRD on HD (QFT-G vs. TST: κ=0.25, 95% confidence interval=0.12-0.37). TST results were positive in two (2.7%) patients when QFT-G results were negative, and the results were negative in 18 (24.3%) patients when QFT-G results were positive, as shown in Table 4.

Discussion
The TST is the classic method for the diagnosis of LTBI, which is based on cell-mediated immune response induced by LTBI. It measures the delayedtype hypersensitivity response to intradermal inoculation of tuberculin PPD [3,12]. QFT-G is another diagnostic method for the diagnosis of TB. This test uses two proteins encoded by a unique genomic segment termed 'Region of Difference 1', which is present in M. tuberculosis but is absent from all strains of BCG, most nontuberculous mycobacteria, and Mycobacterium bovis [13]. These proteins are major targets of Thelper type 1 cells in infected individuals with M. tuberculosis. Therefore, a T cell response to these antigens could serve as a specific marker of M. tuberculosis infection, avoiding the antigenic cross-reactivity of PPD, the main cause of poor specificity of the TST [14].
In our study, 35.1% of patients with ESRD on HD were positive for QFT-G. There was no significant difference in the duration of HD or creatinine levels between QFT-G-positive and QFT-G-negative patients. These results are in agreement with those of Abdel-Nabia et al. [15]. In addition, Ates et al. [16] and Hoffmann et al. [17] recorded that there was no difference between QFT-G-positive and QFT-Gnegative patients regarding HD duration, and suggested that patients with ESRD on HD were still able to produce IFN-γ. However, Inoue et al.
[18] recorded a significant increase in indeterminate QFT-G results with increased duration of dialysis.
We found that 13.5% of ESRD patients on HD were positive for TST. TST-positive patients had a significantly shorter duration of HD and lower creatinine levels than TST-negative patients. These results are in agreement with those of Abdel-Nabia et al. [15], but in disagreement with those of Sagheb et al. [19] and Ates et al. [16], who found that there was no significant relationship between TST results and duration of HD.
In our study, 37.8% of HD patients had LTBI, and they were positive for the QFT-G test and/or TST; 35.1% of the patients were positive for QFT-G, whereas 13.5% of patients were positive for TST. Lee et al. [20] studied LTBI in patients with ESRD on dialysis, and reported that 62.5% of patients were positive with TST, and 40% of the patients were positive with the QFT-G test. In addition, Lee et al. [21] studied LTBI in patients with ESRD on HD, and found that 34.4% of patients were positive by the QFT-G test and 10.8% were indeterminate, whereas by using the TST 53.9% of patents were positive. Abdel-Nabia et al. [15] reported that 25% of the HD patients were positive for the QFT-G test and/or TST, 20% were QFT-G positive, and 15% were TST positive. Dialysis patients are not only at higher risk for reactivation of TB disease but also at higher risk for nosocomial transmission of TB within dialysis centers [22].   Winthrop et al. [25] found that the concordance between results was better, ranging from 71% (TST vs. T-SPOT.TB) to 79% (TST vs. QFT-G) to 87% (QFT-G vs. T-SPOT.TB). Lee et al. [21] recorded a poor correlation between TST and QFT-G for any TST cutoff criteria in patients with ESRD on HD. Abdel-Nabia et al. [15] found 85% concordance between QFT-G and TST results in HD patients. The poor agreement between QFT-G results and TST results could be explained by multiple limitations affecting the role of TST in the diagnosis of LTBI and active TB in patients with ESRD and on dialysis.
There are multiple limitations to the use of TST in the diagnosis of LTBI or active TB in patients with ESRD. The first limitation is the poor sensitivity of TST in patients with ESRD. There is a higher prevalence of anergy to TST in patients with ESRD than in the general population (44 vs. 16%) [26]. The second limitation of TST is its low specificity. Individuals vaccinated with BCG but not infected with M. tuberculosis can show false-positive results with TST [27]. Farhat et al. [28] found that the effect of BCG on TST received in infancy is minimal, especially greater than or equal to 10 years after vaccination.
In a comparative study between using the QFT-G test and the TST for the diagnosis of M. tuberculosis infection, the authors reported that the QFT-G test has excellent sensitivity and specificity and is unaffected by BCG vaccination. The specificity of the TST is high in non-BCG-vaccinated populations, but low and variable in BCGvaccinated populations. In addition, they found that there was a good agreement between the clinical findings and the QFT-G test results [29].
Helmya et al. [30] studied the value of QFT-G assays in monitoring the efficacy of antituberculosis therapy, and recorded that there was a correlation between treatment outcome and changes in IFN-γ response to M. tuberculosis-specific antigens. The low sensitivity and specificity of the TST may explain the reported poor correlation of its results with history, chest radiograph, and QFT-G in the diagnosis of TB in HD patients [8].
Multiple authors studied LTBI in immunocompromised populations and found the superiority of IGRAs in the diagnosis of LTBI in comparison with the TST, which has low sensitivity in these settings [31][32][33][34].
Other advantages of the IGRAs compared with the TST include the shorter time required for obtaining results (16-24 vs. 48-72 h), no need for return visits, objective (instrument-based) interpretation of the test, and lack of boosting effect in repeated tests. Higher sensitivity of IGRAs would identify more infected persons among those with a false-negative TST result. On the other hand, higher specificity will reduce false-positive test results, thus avoiding unnecessary chemoprophylaxis. More true-positive results in infected individuals would increase the rate of diagnosis and treatment of LTBI before progression to active TB [35]. Two meta-analyses have been previously conducted, and both of them reported little value for the prediction of active TB with either IGRAs or TST [36,37].
Limitations of the IGRAs include lack of distinction between LTBI and active TB, and another important issue is the limited availability of IGRAs (in contrast with the worldwide accessible TST). However, the important disadvantage of IGRAs is their high cost [38]. In patients with ESRD and on dialysis, active TB or LTBI cannot be simply ruled out with a negative TST result, but rather IGRA tests and more invasive investigations are recommended [4].
In several reports and guidelines, screening and prophylaxis of LTBI in ESRD patients are recommended [4,39,40].
There are multiple limitations to our study. First, a small number of ESRD patients were included in our study. Second, our study was not a follow-up study to record values for the prediction of active TB with either the QFT-G test or the TST. Further studies are recommended including larger number of patients with ESRD with a good follow-up period to record values for predicting active TB with either the QFT-G test or the TST.

Conclusion
In patients with ESRD and on HD, LTBI cannot be simply ruled out with a negative TST result, but rather IGRA tests (QFT-G) are recommended. Screening and treatment for LTBI should be carried in dialysis patients, aiming to prevent progression to active TB and secondary infection of others.

Financial support and sponsorship
Nil.

Conflicts of interest
There are no conflicts of interest.