Tuberculin skin test for the diagnosis of latent tuberculosis during renal replacement therapy in an endemic area: A single center study

Introduction

Screening patients at high risk for development of tuberculosis (TB) and use of therapy for latent tuberculosis (LTB) infection to prevent the development of active TB is a well-accepted strategy for TB control. Chronic kidney disease (CKD) patients on maintenance hemodialysis (MHD) are a high-risk group for TB.[1–5] These patients are approximately 10 times more likely to get active TB than the general population. Part of the increased risk of TB is caused by progression from LTB to active disease due to impaired immunity associated with CKD.[6] However, routine tuberculin skin testing (TST), also called Mantoux test (MxT), is mostly not performed in these patients, partly because of high rates of anergy among MHD patients.[27] Further, in conditions which alter the cutaneous response to TST, like CKD, significance of TST in the diagnosis of LTB is controversial. There is limited information on TST and LTB in MHD patients from India as well as from the other parts of the world. As there is no gold standard with which TST can be compared for the diagnosis of LTB, the best approach is to follow these patients for the development of TB at a later stage and take disease development as gold standard. This study was conducted to find out the significance of TST for the diagnosis of LTB in patients on renal replacement therapy (RRT).

Materials and Methods

Results

Between May 2000 to May 2006, 554 patients got MHD followed by RT at our hospital. Pediatric patients (23), patients having active TB (57), patients getting TB prophylaxis in clinical trials done at our institute (212) and patients in whom TST could not be done before RT (62), were excluded from the study. Remaining 200 patients were included for the analysis for determining TST for LTB, TST was positive in 21 and negative in 179. Details of the patients are shown in Table 1. In TST negative group 20 (11.1%) and in TST positive group 5 (23.8%) patients developed TB following RT. None of the patients developed TB before RT while he/she was on maintenance dialysis. This is also because of the fact that majority of patients remained on maintenance dialysis for just few weeks before RT was done. Patients who had TB before TST was done were analyzed as having history of TB. Of the 20 cases in TST negative group, who developed post-RT tuberculosis the site of lesion were 1 disseminated TB (bone marrow acid fast bacilli AFB positive), 3 exudative pleural effusion, 5 pulmonary TB (4 bronchoalveolar levage positive), 2 tubercular meningitis, 3 miliary TB, 5 pyrexia of unknown origin (PUO) and one skin TB (AFB positive) were found. In the TST positive group, two had miliary TB, two had pulmonary TB (one bronchoalveolar levage AFB and one sputum AFB positive) and one case had PUO. There was no death attributed directly to TB. However, in one patient it was associated with cytomegalovirus (CMV) infection. In univariate analysis, the follow-up was longer in TST negative group. The number of acute rejection (AR) and new onset diabetes mellitus after transplantation (NODAT) before development of TB were significantly more in TST negative group. On the contrary, CMV disease before development of TB was more in TST positive group. On multivariate Cox regression analysis, while adjusting follow-up period, MMF, tacrolimus, induction therapy, AR, NODAT and CMV before development of TB, hazard of positive TST for development of post-RT TB was 2.7 [P = 0.11, confidence interval (CI) 0.78-9.7]. TB free survival estimates by TST positivity were not significant [Figure 1](log rank test, P = 0.08).

Close

Figure 1

Tuberculosis free survival by TST

Export to PPT

Of the 179 patients who were negative for TST, 156 (78%) did not show any reaction to TST. TB free survival estimates by TST results also had no significant difference in survival between TST negative vs. anergy on TST test (log rank test, P = 0.18). On multivariate Cox regression analysis, while adjusting follow-up, AR and NODAT before development of TB, there was no significant risk of developing TB in patients who were positive for TST as compared to TST negative patients.

Discussion

TB prevention and control includes three strategies: a) identifying and treating persons who have active TB, b) screening persons who have had contact with TB patients to determine whether they are infected with Mycobacterium tuberculosis or have active TB, and providing appropriate treatment and c) screening and treatment of populations at high risk for TB infection to prevent progression to active TB.[10] The practice of screening low-risk groups is not recommended. Patients on MHD are a high-risk group and deserve screening.[11] The American thoracic Society and Centre for Disease Control[11] it is clearly documented that patients of CKD on MHD need tuberculin testing for diagnosis of LTB. It is also made clear that in countries like India where bacilli Calmette-Guerin BCG is used for vaccination for tuberculosis, there is no test that differentiates between BCG effect vs. LTB. Thus, till a new test is developed, TST is the only test that can be used for diagnosis of LTB in this clinical setting. We at our own institution had previously published that isoniazid (INH) prophylaxis in these patients can prevent active TB.[11–13] However, in an endemic area, giving prophylaxis to every patient going for MHD followed by RT may not be justified. Thus, we need to detect patients at significant risk of developing TB. In this regard, TST till now should be the accepted test for the diagnosis and management of LTB.

There are a few reports on TST in relation to patients on RRT, in the literature.[14–26] Major features of these studies have been shown in Table 2. The prevalence of TST positivity ranged from 9% to 55%. However, no definite factors could be attributed to suggest the high prevalence in some studies.[1617192425] In some studies, it has been shown that a repeat booster testing after few days of first skin test had improved the positivity of TST test.[16172024] But its relevance in terms of treatment or future development of active disease has not been evaluated. Some of the factors attributed in different studies increasing TST positivity are male sex,[2025] elderly patients,[23] past history of TB[17] and better nutrition.[19] In this study nearly 80% patients were males and only 5-8% patients had history of past TB.

The main aim of this study was to see the effect of TST on future development of active TB. There is only one study from Indian subcontinent,[19] which can be compared with the present study. It included 108 patients and of them only 79 patients were followed-up after RT. The maximum follow-up period was 30 months, while in our study we had longest follow-up up to 8 years. Though post-RT TB is more common in first year, we have shown that cumulative incidence of post-RT TB goes on increasing even after the first year (unpublished data). However, even with a large number of patients and longer follow-up, we did not find any significant difference in the risk of development of TB in TST positive as compared to TST negative patients. As per our results, sensitivity of TST for predicting post-RT TB is 20% and specificity is 9% only.

T-cell dysfunction in CKD has been suggested as one of the reasons for TST-hypo and unresponsiveness. However, Basoglu et al,[22] could not demonstrate T-cell dysfunction and its association with TST response. Further, Passalent et al,[26] while comparing T-spot test with TST in patients on MHD, demonstrated that T-spot gave 35% positivity as compared to TST, which showed positive response in 12.8% patients.

Discovery of the role of T lymphocytes and interferon-γ assay (IFN-γ) in the immune process has led to the development of an in vitro assay for cell-mediated immune reactions to M. tuberculosis. [27] The IFN-γ assay is a whole-blood test for diagnosing LTB. If a patient is infected with M. tuberculosis, the lymphocytes recognize the tuberculin and release IFN-γ in response. The test is found to be moderately concordant with TST.[28] TST and IFN-γ assay do not measure the same component of the immunologic response and are not interchangeable. Streeton et al,[29] in their study using IFN-γ assay test showed a specificity and sensitivity of 98 and 90%, respectively. Converse et al,[30] studied the suitability of IFN-γ assay in high-risk adults with intravenous drug abuse. However, there are no adequate published data of this test in CKD patients on RRT.

Our study has some limitations. Firstly, we have not used other antigens to do skin test for demonstration of anergy in these patients. Secondly, in some of the patients the diagnosis of TB was made on the basis of response to ATT and one can question the diagnosis in these patients. However, once the same criteria have been followed in both TST positive and negative groups, this should not have any effect while comparing two groups.

To conclude, our study shows that TST in patients on maintenance hemodialysis is an insensitive and nonspecific test to make a diagnosis of LTB infection and this should not be taken as the criterion to prescribe prophylaxis of TB in these patients.