Interferon Gamma Release Assays in Patients with Respiratory Isolates of Non-Tuberculous Mycobacteria - a Preliminary Study.

Interferon gamma releasing assays (IGRAs) are extensively used in the diagnosis of latent tuberculosis infections. Comparing to tuberculin skin test (TST) they lack false positive results in the populations vaccinated with BCG, and in most non-tuberculous mycobacteria (NTM) infections. Nevertheless, Mycobacterium kansasii, Mycobacterium marinum, and Mycobacterium szulgai may induce positive IGRAs due to RD1 homology with Mycobacterium tuberculosis. The aim of the study was to investigate the possible influence of NTM respiratory isolates on the results of IGRAs. 39 patients (23 females and 16 males) of median age 61 years, with negative medical history concerning tuberculosis, entered the study. Identification of NTM was performed using the niacin test and molecular method GenoType CM test (Hain Lifescience). QFT-Plus was performed in 17 patients, T-SPOT-Tb - in 23 patients. Chest X-rays and a high-resolution computed tomography of the chest have been reviewed by the experienced radiologist blinded to the results of IGRAs, in search of past tuberculosis signs. Positive IGRAs results were obtained in three out of 39 patients (8%): 22% of patients with M. kansasii isolates and 18% of patients with radiological signs on HRCT that might be suggestive of past tuberculosis. Positive IGRAs correlated with radiological signs suggestive of past tuberculosis (r = 0.32, p = 0.04), and on the borderline with isolation of M. kansasii (r = 0.29, p = 0.06). These findings may suggest that a positive IGRAs result, in our material, could depend mostly on asymptomatic past Tb infection. The cross-reactivity of M. kansasii isolates with IGRAs was less probable; nevertheless, it requires further investigations. Interferon gamma releasing assays (IGRAs) are extensively used in the diagnosis of latent tuberculosis infections. Comparing to tuberculin skin test (TST) they lack false positive results in the populations vaccinated with BCG, and in most non-tuberculous mycobacteria (NTM) infections. Nevertheless, Mycobacterium kansasii, Mycobacterium marinum, and Mycobacterium szulgai may induce positive IGRAs due to RD1 homology with Mycobacterium tuberculosis. The aim of the study was to investigate the possible influence of NTM respiratory isolates on the results of IGRAs. 39 patients (23 females and 16 males) of median age 61 years, with negative medical history concerning tuberculosis, entered the study. Identification of NTM was performed using the niacin test and molecular method GenoType CM test (Hain Lifescience). QFT-Plus was performed in 17 patients, T-SPOT-Tb – in 23 patients. Chest X-rays and a high-resolution computed tomography of the chest have been reviewed by the experienced radiologist blinded to the results of IGRAs, in search of past tuberculosis signs. Positive IGRAs results were obtained in three out of 39 patients (8%): 22% of patients with M. kansasii isolates and 18% of patients with radiological signs on HRCT that might be suggestive of past tuberculosis. Positive IGRAs correlated with radiological signs suggestive of past tuberculosis (r = 0.32, p = 0.04), and on the borderline with isolation of M. kansasii (r = 0.29, p = 0.06). These findings may suggest that a positive IGRAs result, in our material, could depend mostly on asymptomatic past Tb infection. The cross-reactivity of M. kansasii isolates with IGRAs was less probable; nevertheless, it requires further investigations.

Introduction

The introduction of IGRAs to clinical practice enabled to improve the diagnostic accuracy of latent tuberculosis infection (LTBI). The assessment of LTBI by IGRAs is based on response to specific antigens: early secreted antigenic target 6 kDa (ESAT-6) and culture filtrate protein 10 kDa (CFP-10), localized in a specific genomic area of Mycobacterium tuberculosis, called the region of difference (RD1) (Borkowska 2011; Demkow 2011).

Subsequently, IGRAs have been extensively used to diagnose LTBI in susceptible populations, among others, persons after active tuberculosis contact, immunocompromised hosts, and the candidates to immunosuppressive therapy, especially to anti-TNF alfa treatment (Borkowska et al. 2011; Demkow 2011).

The overall prevalence of LTBI in different countries is closely related to tuberculosis (Tb) burden; thus, the countries with high Tb burden would have more LTBI cases diagnosed with IGRAs and those with low Tb burden – less such cases (Kuś et al. 2011).

The superiority of IGRAs over tuberculin skin test (TST) in LTBI diagnostic pathway, is related to its higher specificity, e.g. lack of false positive results in the populations vaccinated against M. tuberculosis and those infected with most of non-tuberculous mycobacteria (NTM) (Demkow 2011; Kuś et al. 2011; Mancuso et al. 2012). Nevertheless, some NTM, such as Mycobacterium kansasii, Mycobacterium marinum and Mycobacterium szulgai share RD1 with M. tuberculosis and could induce false positive immunological response assessed by IGRAs (Demkow 2011). This observation may be important, especially in the countries with low Tb burden, defined by European Centre for Disease Prevention and Control experts as notification rate lower than 20 per 100 000 population (ECDC/WHO 2018), and with increased incidence of diseases caused by NTM (Prevots et al. 2010). In Poland, as well as in Slovakia and in the United Kingdom, a large proportion of NTM infections have been caused by M. kansasii, possibly influencing the specificity of IGRAs in LTBI diagnostics (Słupek et al. 1997; Hoefsloot et al. 2013; van der Werf et al. 2014; Wilińska et al. 2014; Bakuła et al. 2018).

Thus, the aim of the present study was to investigate the influence of NTM isolation from respiratory specimens on the results of IGRAs in patients with no medical history of tuberculosis.

Experimental

Materials and Methods

Patients. Overall, 39 patients (23 females and 16 males) of median age 61 years (27–85 years), from whom NTM was cultured from respiratory specimens (sputum and/or bronchial washings) in the period of 2010–2017, and IGRA test was performed simultaneously, entered the study. Patients, who had been diagnosed and treated for tuberculosis, were excluded from the study.

Non-tuberculous mycobacterial lung disease (NTMLD) was recognized in 16 patients according to American Thoracic Society/Infectious Diseases Society of America (ATS/IDSA) and recent British Thoracic Society recommendations (Griffith et al. 2007; Haworth et al. 2017). In 23 patients, respiratory isolates did not cause the disease. The characteristic of the population is summarized in Table I.

Table I

Characteristics of the population of patients, from whom NTM was isolated from respiratory specimens.

Sex	No of pts	Age Median (range)	BMI Median (range)	Number of patients with certain coexisting disease
				COPD	ILD or GPA	CTD	CF	Npl	Others[*]
Males	16	58.5 (28–75)	24.9 (19.5–38.9)	5	3	0	0	3	5
Females	23	62 (29–85)	23.8 (16.2–37.3)	5	6	4	2	2	4
Total	39	61 (27–85)	24.3 (15–38.9)	10	9	4	2	5	9

COPD – chronic obstructive pulmonary disease; ILD – interstitial lung disease; GPA – granulomatosis with polyangiitis; CTD – connective tissue disease; CF – cystic fibrosis; npl – neoplasm; BMI – body mass index

Diabetes (2), hypothyreosis (1), lung aspergilloma (1), renal insufficiency (1), bronchiectasis (2), actinomycosis (1), lung cirrhosis (1), lobar pulmonary artery agenesis (1), trombofilia (1)

Chest X-rays and high resolution computed tomography (HRCT) of the chest were reviewed by the experienced radiologist blinded to the results of IGRAs in search of fibrotic foci localized in the upper lobes or upper parts of lower lobes, as well as parenchymal and lymph nodes’ calcifications.

Methods of NTM culture and identification. The specimens were digested with the sodium hydroxide and N-acetyl-L-cysteine (NaOH/NALC) method. After decontamination, the sample was neutralized with sterile phosphate buffer (pH 6.8) and centrifuged at 3000 × g for 15 min. The pellet was suspended in 2 ml of phosphate buffer. The strains were cultured on solid media: egg-based L-J medium, Stonebrink medium and in automated system MGIT (Becton Dickinson) (Klatt et al. 2015).

Identification of culture was performed using the niacin test and a molecular method GenoType CM test (Hain Lifescience). The GenoType CM test, using the DNA-STRIP method allowed the identification of M. tuberculosis complex strains and 14 clinically relevant NTM within a single procedure. The procedure for identifying strains consists of three steps: isolation of DNA, amplification using primers labeled with biotin and a reverse hybridization. This hybridization reaction includes the consecutive steps: chemical denaturation of the amplification products, hybridization of single-stranded amplicons labeled with biotin on a membrane coated with probes, washing, adding streptavidin/alkaline phosphatase conjugate, staining reaction using alkaline phosphatase (Zabost and Augustynowicz-Kopeć 2015).

Molecular detection of Identification using the BD ProbeTec ET system (Becton Dickinson Diagnostic Instruments) was performed according to the manufacturer’s instructions. The instrument reported amplification signals > 3 500 method-otherthan-acceleration (MOTA) units as positive (Klatt 2015). It was performed in all cases with a positive result of direct bacterioscopy.

Interferon gamma release assays. The IGRAs assays identified cellular immune responses to M. tuberculosis by measuring interferon-gamma (IFN-γ) after stimulation of T cells with M. tuberculosis-specific antigens. Two tests were available: T-SPOT.TB based on the Elispot-enzyme-linked immuno spot and QuantiFeron TB Gold Plus (QFT-Plus) based on the enzymelinked immunosorbent assays (ELISA) technique. The T-SPOT.TB test is based on measurement of the number of peripheral mononuclear cells that produce IFN-γ after stimulation with two antigens: ESAT-6 and CFP 10. The antigens used in QFT-Plus consisted of a peptide cocktail simulating the ESAT-6 and CFP 10 (Borkowska et al. 2017).

QFT-Plus was performed in 17 patients, T-SPOT-Tb – in 23 (including one patient, in whom both IGRAs have been performed).

Statistical analysis. The data were presented as medians and ranges or as number and percentage of positive cases. The differences between categorical variables were analyzed with the chi-square test, for quantitative variables – ANOVA test was used. The correlation was assessed with Spearman rank order test. P < 0.05 was considered statistically significant.

Results

Positive IGRAs results were obtained in three out of 39 patients (8%): positive QFT Plus – in two cases, T-SPOT-TB – in one case. In one patient, in whom both tests were performed, T-SPOT-Tb was negative but QFT was positive.

IGRAs results according to the identified type of NTM were shown in Table II. Positive results were obtained in 2/9 (22%) of patients with M. kansasii and one patient with M. fortuitum isolate. IGRAs positivity was thus found in 2/10 (20%) of NTM sharing the RD1 region with M. tuberculosis (M. kansasii, M. szulgai) and 1/29 (3%) of those without RD1 sharing (p = 0.31).

Table II

IGRAs results according to the NTM species.

Species	IGRA (+)	IGRA (−)	Total
M. kansasii	2 (22%)	7	9
M. avium	0	7	7
M. gordonae	0	6	6
M. chimaera	0	5	5
M. xenopi	0	5	5
M. fortuitum	1 (33%)	2	3
M. szulgai	0	1	1
M. abscessus	0	1	1
M. mucogenicum	0	1	1
M. smegmatis	0	1	1
Total	3 (8%)	36	39

Positive IGRAs were obtained in 2/23 females (9%) and 1/16 males (6%), (p = 0.91).

IGRAs results according to the patients’ age (Table III) revealed that positive results were found only in the patients above 60 years of age; nevertheless, the age-related differences were not significant (p = 0.49).

Table III

IGRAs result according to patients’ age.

Age (years)	≤ 24	25–44	45–59	≥ 60	Total
IGRAs (+)	0	0	0	3 (14%)	3
IGRAs (−)	0	11	8	17	36
Total	0	11	8	20	39

HRCT analysis revealed the presence of lesions suggesting the possibility of the previous infection with M. tuberculosis in 17/39 (44%) of patients.

The results of IGRAs according to the results of radiological analysis are shown in Table IV. Positive IGRAs were found in 3/17 (18%) patients with the above-mentioned radiological signs and none of the remaining patients (p = 0.15).

Table IV

The IGRAs results according to radiologic signs of past tuberculosis.

Chest CT	IGRA (+)	IGRA (−)	Total
Past tb signs	3 (18%)	14 (82%)	17
No past tb signs	0 (0%)	22	22
Total	3	36	39

CT – computed tomography, tb – tuberculosis

Positive IGRAs results correlated with radiological signs suggestive of possibility of infection with M. tuberculosis in the past (r = 0.32, p = 0.04); the correlation of positive IGRA with the isolation of M. kansasii was borderline (r = 0.29, p = 0.06).

Discussion

Positive IGRAs have been found in three (8%) of patients with NTM cultured from respiratory specimens. Two positive IGRAs results concerned the patients with M. kansasii isolates (22%), the species sharing RD1 with M. tuberculosis, one – the patient with M. fortuitum isolate, the species not sharing RD1. The Japanese studies revealed IGRAs positivity in 19–52% of M. kansasii isolates (Kobashi et al. 2009; Sato et al. 2016). The European experience is scarce; nevertheless, Hermansen et al. (2014) found positive IGRAs results in 2/2 M. kansasii isolates (100%).

The discrepancies concerning IGRAs positivity were found also in M. avium/M. intracellulare (MAC) infection: from 0–8% positive results in the studies conducted in Northern Europe and Japan (Adams et al. 2008; Kobashi et al. 2009; Hermansen et al. 2014) to 34% in the study conducted in South Korea (Ra et al. 2011). In our study, all seven cases of M. avium isolates were IGRAs negative. As MAC doesn’t share the RD1 region with M. tuberculosis, positive IGRAs obtained for the patients with MAC isolates by other authors might reflect the background influence of past tuberculosis, especially in countries with high Tb burden (Ra et al. 2011; Wang et al. 2016).

Poland belongs to low Tb burden countries, with the incidence rate of tuberculosis calculated as 19.7/100 000 in 2010 and 15.1/100 000 in 2017 (Korzeniewska-Koseła 2017, 2018). The analysis of LTBI prevalence assessed with QFT, performed in 2010 on 621 healthy subjects from Mazowieckie province, revealed the positive results in 23.3% of them, more frequently in older people compared to younger ones (Kuś et al. 2011). Positive IGRAs in the present study were found exclusively in patients > 60 years of age, indicating possible influence of background non-recognized Tb infection in the past on the results obtained.

Retrospective analysis of HRCT scans by an experienced radiologist blinded to IGRAs results, revealed the lesions suggestive of possibility of infection with M. tuberculosis in the past in 17 out of 39 patients (44%), despite lack of Tb in anamnesis. Positive IGRAs were noted in three out of 17 patients with radiological signs suggestive of past tuberculosis (18%) and none of the patients without such findings on chest CT scan.

These findings may suggest that positive IGRAs result, in our material, could depend mostly on asymptomatic past Tb infection. The cross-reactivity of M. kansasii isolates with IGRAs is less probable, because it was present only in two out of nine patients with M. kansasii isolates.

The same type of analysis has been performed by Sato et al. who found T-SPOT positivity in 33% of patients with M. kansasii isolates, but after exclusion of those with a history of tuberculosis (defined as either Tb diagnosis and treatment in the past or chest X-ray features suggesting previous tuberculosis), the percentage of T-SPOT positive cases decreased to 19%. They concluded that features of previous Tb are the only risk factor for positive IGRAs in the patients with M. kansasii respiratory isolates.

Since the group of patients was small in the current study, further studies are required to answer the question, whether the positive IGRAs in patients with M. kansasii isolates is caused by RD1 cross-reactivity or rather the background of asymptomatic tuberculosis in the past.