This study was presented in part at the 48th Annual Meeting of the Infectious Diseases Society of America, 21–24 October 2010 (Poster session, abstract no. 2647).
A Prospective Longitudinal Study Evaluating the Usefulness of a T-Cell-Based Assay for Latent Tuberculosis Infection in Kidney Transplant Recipients
Article first published online: 12 JUL 2011
© 2011 The Authors Journal compilation © 2011 The American Society of Transplantation and the American Society of Transplant Surgeons
American Journal of Transplantation
Volume 11, Issue 9, pages 1927–1935, September 2011
How to Cite
Kim, S.-H., Lee, S.-O., Park, J. B., Park, I.-A., Park, S. J., Yun, S.-C., Jung, J. H., Kim, Y. H., Kim, S. C., Choi, S.-H., Jeong, J.-Y., Kim, Y. S., Woo, J. H., Park, S.-K., Park, J. S. and Han, D. J. (2011), A Prospective Longitudinal Study Evaluating the Usefulness of a T-Cell-Based Assay for Latent Tuberculosis Infection in Kidney Transplant Recipients. American Journal of Transplantation, 11: 1927–1935. doi: 10.1111/j.1600-6143.2011.03625.x
Clinical Trial Registration Number NCT00682045
- Issue published online: 29 AUG 2011
- Article first published online: 12 JUL 2011
- Received 20 January 2011, revised 15 March 2011 and accepted for publication 11 April 2011
Vol. 11, Issue 11, 2541, Article first published online: 26 OCT 2011
- ELISPOT assay;
- kidney transplantation;
We evaluated whether ELISPOT assay can predict tuberculosis (TB) development in kidney-transplantation (KT) recipients with a negative tuberculin skin test (TST). All adult patients admitted to a KT institute between June 2008 and December 2009 were enrolled; TB development after KT was observed between June 2008 and December 2010. Isoniazid (INH) was given to those patients with positive TST or clinical risk factors for latent TB infection (LTBI). ELISPOT assay was performed on all patients, and TB development after KT was observed by a researcher blinded to the results of ELISPOT. A total of 312 KT recipients including 242 (78%) living-donor KT were enrolled. Of the 312 patients, 40 (13%) had positive TST or clinical risk factors for LTBI and received INH; none developed TB after KT. Of the remaining 272 patients, 4 (6%) of 71 with positive ELISPOT assay developed TB after KT, whereas none of the 201 patients with negative (n = 171) or indeterminate ELISPOTs (n = 30) developed TB after KT (rate difference between positive and negative/indeterminate ELISPOT, 3.3 per 100 person-years [95% CI 1.4–5.1, p<0.001]). Positive ELISPOT results predict subsequent development of TB in KT recipients in whom LTBI cannot be detected by TST or who lack clinical risk factors for LTBI.
The development of a reliable test for latent tuberculosis infection (LTBI) is a key issue in efforts to eliminate tuberculosis (TB). Traditionally, the tuberculin skin test (TST) has been the method of choice for the diagnosis of LTBI (1). However, the TST is not specific, as the antigens present in PPD cross-react with Bacilli Calmette–Guérin (BCG) and environmental mycobacteria, nor is it sensitive, due to anergy in immunocompromised patients (2). Interferon (IFN)-γ releasing assays (IGRAs), a new generation of diagnostic TB assays, have recently shown promising results in diagnosing LTBI (3) based on data extrapolated from contact-tracing studies (4,5) and studies of active TB (6–10). However, objective comparison of the diagnostic performance of the IGRAs and the TST is hindered by the lack of a gold standard reference test for LTBI, which makes it impossible to directly measure sensitivity and specificity (2). Therefore, the value of a given test for LTBI in predicting subsequent active TB can only be measured in a longitudinal study, such as the numerous longitudinal studies correlating the size of the TST reaction with future risk of active TB (11). These have shown that longitudinal studies require a large number of individuals from the general population, and long follow-up periods (11). Thus, the best candidates for such longitudinal studies of IGRAs are patients with LTBI at high-risk of developing subsequent active TB. A few studies have provided longitudinal data on IGRAs in the context of persons in recent contact with pulmonary TB (12–14) or HIV patients (15). However, no longitudinal studies have focused on immunocompromised patients, such as transplant recipients undergoing screening for remote LTBI and in whom TST is of limited clinical value.
Since TB is one of the important opportunistic infections in transplant recipients (16,17) it has been recommended that all transplant recipients undergo a TST before transplantation (18). However, because of anergy, only 20–25% of all cases of active TB after transplantation occur in patients who give positive TST reactions before transplantation (19), indicating that the ability of TST to diagnose LTBI in transplant candidates is suboptimal. We therefore performed a prospective longitudinal study to assess whether an enzyme-linked immunosorbent spot (ELISPOT) assay is capable of predicting active TB development in kidney transplant (KT) recipients with negative TST results, in Korea, a country with an intermediate TB burden. The incidence of TB has been reported to be 90 cases per 100 000 Korean persons (20) and 4.6% in Korean KT recipients (14 new TB cases out of 304 KT recipients between 1984 and 1994) (21).
All patients admitted for transplantation in a renal transplant unit between June 2008 and December 2009 at a 2700-bed, tertiary-care hospital in Seoul, South Korea, were prospectively screened. The resulting cohort was observed until December 2010, with a focus on TB development after KT. Each patient was interviewed for clinical risk factors of LTBI (i.e. close contact with an active TB patient, a history of untreated or inadequately treated TB, a recent history of TST) by a trained nurse. Tests for LTBI (i.e. chest X-ray, TST) were performed by a trained nurse 2–5 days immediately prior to scheduled transplant surgery for living-donor transplantation, or 1 or 2 days immediately after emergency transplant surgery for deceased-donor transplantation. If chest radiographic findings were abnormal, a sputum AFB smear and a CT scan were performed to rule out active pulmonary TB. Exclusion criteria were refusal of informed consent, presence of active TB, presence of skin disease that precluded TST, pediatric renal transplant candidates (<16 years old) and presence of any contraindication for KT (e.g. malignancy). Pancreas transplantation alone was also excluded. All individuals were informed of the nature of the study, and all participants provided written informed consent. This investigation was approved by the hospital Institutional Review Board. Preliminary cross-sectional data from this longitudinal cohort study have been published elsewhere (22).
ELISPOT assay and TST
A peripheral venous blood sample was collected from each patient for an ELISPOT assay for the IFN-γ-producing T-cell response (i.e. T-SPOT.TB, Oxford Immunotec, Abingdon, UK). All blood samples were collected prior to TST to avoid a possible boosting effect of TST on the ELISPOT assay. Detailed laboratory procedures and the criteria used for positive, negative and indeterminate outcomes have been described (6–10,22).
The TST was performed by the Mantoux technique, injecting a 2-TU (tuberculin unit) dose of purified protein derivative RT23 (Statens Serum Institut, Copenhagen, Denmark) intradermally into the forearm. The positive criterion for TST was 10 mm or greater size of induration 48–72 h after injection, as described previously (6–10,22) and in accordance with Korea Centers for Diseases Control and Prevention guidelines (23).
Study design and isoniazid treatment
The goal of this study was to evaluate positive ELISPOT tests for predicting active TB development in KT recipients who had neither a positive TST nor clinical risk factors for LTBI. Accordingly, we administered a 9-month course of isoniazid (INH) preventive therapy immediately after renal transplantation to those patients with any of the following criteria (22) recommended in previous guidelines (16,18): (i) positive TST (≥ 10 mm) before transplantation; (ii) (a) a history of close contact with pulmonary tuberculosis within the last year, (b) abnormal chest radiograph and no prior prophylaxis or (c) a history of untreated or inadequately treated TB; or (iii) newly infected persons (recent conversion of TST to positive). In addition, we administered isoniazid therapy to renal recipients if the kidney donor met above criteria (ii) or (iii). We did not perform the TST or ELISPOT assay in transplant donors. The study cohort consisted of all patients who received KT during the study period and did not meet the exclusion criteria; the final cohort consisted of patients who had neither positive TST nor clinical risk factors for LTBI.
Assessment of outcomes
The primary outcome of the study was the development of TB; secondary outcomes included mortality and rejection. Subjects were evaluated monthly during the first 6 months after transplantation and every 3 months thereafter. If active TB was suspected at any scheduled or unscheduled visit on the basis of symptoms and signs, patients were actively screened for TB in consultation with infectious diseases specialists (SH Kim and SO Lee). The development of TB after KT was observed by attending surgeons, nephrologists and infectious diseases specialists blind to the results of ELISPOT assays, to avoid a verification bias.
Patients with suspected active TB were categorized as confirmed, probable and possible TB as described previously (10). Mortality and rejection episodes were assessed at all scheduled and unscheduled visits.
The primary goal of the present study was to test the hypothesis that post-KT patients who were ELISPOT-positive developed active TB more frequently than ELISPOT-negative or indeterminate patients (comparator). Assuming a 20% positive ELISPOT result in KT patients with a negative TST, and extrapolating from studies published elsewhere (16,24), we calculated statistical power based on an estimated 10% TB development rate in the ELISPOT-positive patients and a 1% TB development in the comparator. We also assumed that the rate of negative TST in KT patients was 76% based on a previous study (16), and that the overall incidence of TB development after KT was 3%, also based on previous studies (16,21). We concluded that, with a type I error of 0.05 and an allowance of 10% loss to follow-up, a sample size ≥ 240 was needed to have 80% statistical power to detect a difference between the ELISPOT-positive group and the comparator. Since our hospital had performed 168 operations for KT in 2007, our plan was to enroll patients for 18 months to obtain ≥ 240 patients (predicted, 252) and observe the development of TB for an additional 12 months.
The incidence of TB was estimated using the person-year method and an exact confidence interval (CI) was calculated based on the Poisson distribution. Rate differences were calculated using test-based CIs, comparing TB incidence in the ELISPOT-positive patients with negative TST with that in the ELISPOT-negative or indeterminate patients with negative TST. All tests of significance were two-tailed; a p value of less than 0.05 was considered to indicate statistical significance. Calculations were performed using the SPSS for Windows software package, version 14.0K (SPSS Inc, Chicago, IL, USA) and MedCalc software (Mariakerke, Belgium).
Figure 1 outlines the study flow for all patients admitted for KT in our hospital between June 2008 and December 2009, and presents the results of observations on the development of TB after KT between June 2008 and December 2010. A total of 324 patients were initially enrolled in the study. Of these, 12 patients were ultimately excluded (Figure 1). The resulting study cohort of 312 patients was followed up for 556.3 person-years; thus, the median duration of follow-up was 1.8 years (interquartile range [IQR] 1.4–2.2). Of the 312 patients in the study cohort, 242 (78%) received living-donor KT, 54 (17%) received deceased-donor KT and 16 (5%) received simultaneous pancreas-kidney transplants. None of 312 patients reported a previous treatment history for LTBI. A total of 16 patients in the study cohort had clinical risk factors for LTBI and received INH preventive therapy (Figure 1). Of these 16 patients, three reported recent contact (≤ 1 year) with patients who had active pulmonary TB. However, none of the transplant recipients enrolled in this study reported TB exposure after KT over the study period. In addition, 24 (8%) of the 312 patients had positive TST results and received INH preventive therapy (Figure 1). Finally, the 272 patients with negative TST and without any clinical risk factors for LTBI were included in the final cohort analysis (Figure 1). Demographic data for the study groups are shown in Table 1. None of the 312 recipients and corresponding living- or deceased-donors was positive for HIV infection. A total of 256 (82%) of the patients had BCG vaccination histories or scars. BCG vaccination status was not associated with positive TST results (p = 0.95) or positive ELISPOT results (p = 0.36).
|All (n = 312)||Isoniazid treatment (n = 40)||No isoniazid treatment (n = 272)|
|Clinical risk factors (n = 16)||TST ≥ 10 mm (n = 24)||ELISPOT (+) (n = 71)||ELISPOT (-) or indeterminate (n = 201)|
|Mean age, years (±SD)||42.5 (± 10.2)||45.6 (± 10.6)||47.4 (± 11.3)||46.0 (± 9.3)1||40.4 (± 9.9)1|
|Male gender||176 (56)||14 (88)||16 (67)||40 (56)||106 (53)|
|BCG vaccination history or scars||256 (82)||13 (81)||18 (75)||53 (75)||162 (81)|
|Primary reason for transplant|
|Glomerulonephritis||84 (27)||4 (25)||8 (33)||19 (27)||53 (26)|
|Hypertension||77 (25)||5 (31)||7 (29)||15 (21)||50 (25)|
|Diabetes mellitus||60 (19)||5 (31)||7 (29)||11 (16)||37 (18)|
|Unknown||66 (21)||1 (6)||1 (4)||17 (24)||41 (23)|
|Polycystic kidney disease||14 (5)||1 (6)||1 (4)||5 (7)||7 (4)|
|Others||11 (3)||0||0||4 (6)||7 (4)|
|Living donor kidney||242 (78)||11 (69)||24 (100)||66 (93)*||141 (70)a|
|Deceased donor kidney||54 (17)||3 (19)||0||4 (6)*||47 (23)a|
|Pancreas and kidney||16 (5)||2 (12)||0||1 (1)||13 (7)|
|Retransplantation||20 (6)||1 (6)||1 (4)||3 (4)||15 (8)|
|ABO-mismatch transplantation||16 (5)||4 (25)||1 (4)||3 (4)||8 (5)|
|Primary transplant induction therapy at transplant|
|Anti-IL-2 receptor antibodies||266 (85)||10 (63)||17 (71)||59 (83)||179 (89)|
|Antithymocyte antibodies||31 (10)||5 (31)||5 (21)||7 (10)||14 (7)|
|Rituximab||16 (5)||4 (25)||1 (4)||3 (4)||8 (4)|
|Median follow-up, months (IQR)||14.5 (9.9–19.6)||12.9 (10.5–19.9)||16.1 (13.9–18.6)||14.1 (8.4–19.7)||14.4 (8.5–18.9)|
|Mortality||2 (0.64)||0||0||0||2 (1)|
|Rejections||22 (7)||2 (13)||2 (8)||5 (7)||13 (6)|
ELISPOT results and development of TB
Of the 312 patients, 40 (13%) received INH preventive therapy, none of whom developed TB after KT (Figure 1). Of these 40 patients, three (8%) stopped INH treatment due to suspicious INH-associated hepatotoxicity; one patient gave an abnormal result in a liver function test 1 week after INH treatment, another 6 weeks after INH treatment and the third 8 weeks after INH treatment. A further two patients received INH treatment for only 4–6 months, respectively, due to compliance problems. The remaining 35 (88%) completed 9-month treatments. Rates of TB in relation to induration size in the TST are shown in Table 2.
|Variable||TB incidence rates|
|Number of patients||Number of TB cases||Number of person-years||TB rate per 100 person-years||95% CI|
|Total patients (n = 312)||312||4||556.26||0.72||0.19–1.84|
|——Tuberculin skin test ≥ 10 mm or clinical risk factors1 for LTBI with INH treatment||40||0||73.00||0||0–5.05|
|———Tuberculin skin test ≥ 10 mm with INH treatment||24||0||44.50||0||0–8.29|
|————Clinical risk factors for LTBI with INH treatment||16||0||28.50||0||0–12.94|
|——Tuberculin skin test|
|Tuberculin skin test < 5 mm||262||3||457.57||0.66||0.14–1.92|
|————Tuberculin skin test ≥ 5 mm (with INH treatment if tuberculin skin test ≥ 10 mm)||50||1||98.69||1.01||0.03–5.65|
|————Tuberculin skin test ≥ 10 mm with INH treatment||24||0||44.50||0||0–8.29|
|————Tuberculin skin test ≥ 15 mm with INH treatment||8||0||14.72||0||0–25.06|
|——Tuberculin skin test < 10 mm without INH treatment||272||4||483.25||0.83||0.23–2.12|
|————Positive ELISPOT results||71||4||122.10||3.282,3||0.89–8.39|
|————Negative or indeterminate ELISPOT results||201||0||361.16||02||0–1.02|
|Negative ELISPOT results||171||0||307.83||03||0–1.20|
|Indeterminate ELISPOT results||30||0||53.33||0||0–6.92|
|Living donor kidney transplantation (n = 242)||242||4||427.29||0.94||0.26–2.39|
|——Tuberculin skin test ≥ 10 mm or clinical risk factors for LTBI with INH treatment||35||0||62.49||0||0–5.90|
|————Tuberculin skin test ≥ 10 mm with INH treatment||24||0||44.50||0||0–8.29|
|————Clinical risk factors for LTBI with INH treatment||11||0||17.99||0||0–20.51|
|——Tuberculin skin test|
|——Tuberculin skin test < 5 mm||197||3||339.89||0.88||0.18–2.58|
|————Tuberculin skin test ≥ 5 mm (with INH treatment if tuberculin skin test ≥ 10 mm)||45||1||87.39||1.14||0.03–6.38|
|————Tuberculin skin test ≥ 10 mm with INH treatment||24||0||44.50||0||0–8.29|
|————Tuberculin skin test ≥ 15 mm with INH treatment||8||0||14.72||0||0–25.01|
|——Tuberculin skin test < 10 mm without INH treatment||207||4||244.02||1.64||0.45–4.19|
|——Positive ELISPOT results||66||4||112.67||3.554,5.||0.97–9.09|
|——Negative or indeterminate ELISPOT results||141||0||252.13||04||0–1.46|
|——Negative ELISPOT results||126||0||226.71||05||0–1.63|
|——Indeterminate ELISPOT results||15||0||25.42||0||0–14.51|
Of the final cohort comprising 272 patients (i.e. those without positive TST or clinical risk factors for LTBI), 71 (26%) gave positive results, 171 (63%) negative results and 30 (11%) indeterminate results, in ELISPOT assays before KT (Figure 1). Four (6%) of the 71 patients with positive ELISPOTs developed TB after KT, whereas none of the 201 patients with negative (n = 171) or indeterminate ELISPOTs (n = 30) developed TB after KT (rate difference between positive and negative/indeterminate ELISPOT, 3.3 per 100 person-years [95% CI 1.4–5.1, p < 0.001]). If patients with indeterminate ELISPOT results were excluded from the final analysis, the rate difference between positive and negative ELISPOT was 3.3 per 100 person-years [95% CI 1.3–5.3, p = 0.002]). Of the 4 patients who developed TB after transplantation, all had BCG vaccination histories or scars. Therefore, 4 (0.4%) of 256 patients who had BCG vaccination histories or scars developed TB after transplantation while none of 56 patients without BCG vaccination developed TB after transplantation (p > 0.99). In a subgroup analysis that included patients who received living-donor KT (n = 242), 4 (6%) of 66 patients with a positive ELISPOT developed TB, whereas none of the 141 patients with negative (n = 126) or indeterminate ELISPOTs (n = 15) developed TB after KT (rate difference between positive and negative/indeterminate ELISPOT, 3.6 per 100 person-years [95% CI 1.2–5.9, p = 0.003]). The rates of TB in relation to the results of ELISPOT assays are shown in Table 2. The diagnostic performances of the TST and ELISPOT assays for the development of TB in KT recipients are also shown in Table 3.
|Variable||Number of patients||Number of TB cases||Sensitivity||Specificity||Accuracy||PPV||NPV|
|Tuberculin skin test ≥ 10 mm or clinical risk factors for LTBI with INH treatment||40||0||0 [0/4] (0–60)||87 [268/308] (82–91)||86 [268/312] (82–89)||0 [0/40] (0–9)||99 [268/272] (96–100)|
|Tuberculin skin test ≥ 5 mm (with INH treatment if tuberculin skin test ≥ 10 mm)||50||1||25 [1/4] (0–80)||84 [259/308]1 (79–88)||83 [260/312] (79–87)||2 [1/50] (0–11)||99 [259/262] (97–100)|
|Tuberculin skin test ≥ 10 mm with INH treatment||24||0||0 [0/4] (0–60)||92 [284/308]2 (89–95)||91 [284/312] (87–94)||0 [1/24] (0–14)||99 [284/288] (96–100)|
|Tuberculin skin test ≥ 15 mm with INH treatment||8||0||0 [0/4] (0–60)||97 [300/308]3 (95–99)||96 [300/312] (93–98)||0 [0/8] (0–37)||99 [300/304] (97–100)|
|Positive ELISPOT results||71||4||100 [4/4] (40–100)||78 [241/308]123 (73–83)||79 [245/312] (74–83)||6 [4/71] (2–14)||100 [241/241] (98–100)|
Incident TB cases
Table 4 presents the demographic characteristics and test results of the four patients who subsequently developed TB after KT. In three cases (patients no. 160, 168 and 298 in Table 4), cultures of clinical specimens tested positive for pan-drug-susceptible M. tuberculosis; hence each was classified as confirmed TB. The remaining patient (patient no. 242 in Table 4) presented with severe headache with cerebrospinal fluid findings of lymphocytic pleocytosis (white blood cell count, 300/mm3; 60% lymphocytes, raised protein levels (119 mg/dL) and sterile cultures, and responded successfully to antituberculosis therapy. Therefore this case was classified as probable TB. Initial donor screening revealed no chest X-ray abnormities or clinical risk factors for LTBI in any of the four donors, and none reported any symptoms or signs suggestive of active TB in telephone interviews conducted at the end of the study period. There were no rejection episodes before and after the onset of TB development in these four patients.
|Patient number||Age/sex||KT type||Time to diagnosis, months||Type of TB||Results of culture||TST, mm||Initial ELISPOT (ESAT-6/CFP-10)1||Follow-up ELISPOT (ESAT-6/CFP-10)1|
|160||55/M||Living-donor KT||4.4||Miliary TB||(+) from sputum [S]||0||4/464||288/1192|
|168||43/M||Living-donor KT||1.8||Pulmonary TB||(+) from PCNBx and BAL [S]||0||344/20||332/1052|
|242||57/M||Living-donor KT||4.4||Miliary TB||(+) from sputum and BAL [S]||0||300/24||64/44|
|298||44/M||Living-donor KT||10.8||TB meningitis||(−) from CSF||7||16/518||0/16|
The frequency of IFN-γ-secreting T cells increased after KT in two of the four patients but decreased after KT in one of them (Table 4). The remaining patient exhibited discordant responses, giving an increased response to one M. tuberculosis-specific peptide (CFP-10) but a decreased response to another (ESAT-6), between pre-transplantation and the development of active TB (Table 3).
During follow-up, two of the 312 study cohort patients died of causes unrelated to TB; both were ELISPOT-negative subjects with negative TSTs. One of the patients died of a bowel infarction 3 weeks after KT and the other of CMV colitis and Chryseobacterium indologenes bacteremia 4 weeks after KT. Rejection episodes occurred in 22 (7%) of the 312 patients but the frequency of these episodes was not different between groups (Table 1).
In this prospective study we found that ELISPOT-positive KT patients with negative TST results had a higher risk of progression to active TB than ELISPOT-negative or -indeterminate KT patients with negative TSTs. It is noteworthy that none of the four patients who developed active TB had any clinical risk factors for LTBI and all exhibited induration sizes < 10 mm in the TST before KT, indicating that IGRA provides greater sensitivity for detecting those likely to develop active TB.
Data relating to the predictive value of IGRAs in patients with recent contacts are available from four longitudinal studies (12–14, 25). A study from Germany found that a significantly higher proportion of untreated household contacts with a positive IGRA (ELISA-based assay) progressed to TB than contacts with a positive TST (12). However, a study in Turkey found that TB incidence among ELISPOT-positive contacts was similar to that of TST-positive contacts (13). In addition, a study from Gambia showed that neither TST nor ELISPOT was predictive of subsequent progression to TB (14). Recently, a study of 339 immigrants to the Netherlands also reported that TST and IGRA based on ELISA assays performed similarly in predicting subsequent active TB (25). However, intrinsic problems prevent these longitudinal studies on recent contacts from being generalized to immunocompromised patients or the general populations. Thus, because INH treatment is the standard of care for recent contacts with positive TST, INH prophylaxis allocation bias should not be controlled for, as it was in these studies in which only patients who refused INH treatment for the development of TB after recent contact were followed up. In addition, the results should not be generalized to persons who are being screened for remote latent TB infection because it is possible that the positive IGRAs merely indicate past encounters of the immune system with TB, rather than genuine “latent” infections. Indeed, short-incubation (e.g. overnight) IGRAs are less sensitive for detecting past latent TB than long-incubation (e.g. 5 days) IGRAs or TSTs (26). In this context, our findings are important in that they indicate that the ELISPOT assay has predictive value for the subsequent development of TB in immunocompromised patients who are being screened for remote LTBI and for whom TST is of limited clinical value. However, remote versus recent LTBI is a somewhat arbitrary distinction because no one knows the exact time of exposure to TB. Therefore further studies are needed to evaluate the usefulness of different diagnostic tests for LTBI in a number of clinical settings and regions.
To our knowledge, only three longitudinal studies on IGRAs in populations other than recent contacts are available: on HIV patients (15), HIV-infected pregnant women (27) and silicosis patients (28). The HIV study demonstrated that three (0.4%) of 772 HIV patients developed active TB, all of whom had positive IGRAs (ELISA-based assay) at baseline. However, in that study, only IGRA-positive patients received TST, and INH preventive therapy was not given to the HIV-positive patients with positive TSTs. Furthermore, patients with a positive IGRA result were less likely to have received antiretroviral therapy than were patients with negative IGRA results. Accordingly, an antiretroviral therapy allocation bias could not be ruled out. The other study on HIV-infected pregnant women, which showed that positive ELISPOT assays were associated with postpartum active TB and mortality among mothers and their infants, lacked maternal TST results, so that we cannot compare the IGRA results with TST results (27). The study of Hong Kong silicosis patients demonstrated that a positive ELISPOT assay result was significantly more effective in predicting the subsequent development of active TB than a positive TST (28). However, since only one third of the TST-positive patients (≥ 10 mm) accepted INH preventive therapy, the uneven allocation of INH treatment may have had a confounding effect on the development of TB. In our study, we provided INH treatment to KT recipients based on strict criteria established by recent INH treatment guidelines (18). We thus observed subsequent TB development in patients who did not need to receive INH treatment, so allowing us to control for the confounding effect of INH treatment. The other strength of the current study is that attending physicians were blind to the results of the ELISPOT assays, thus preventing closer monitoring of ELISPOT-positive patients and eliminating verification bias.
We defined positivity on TST as an induration size ≥10 mm, whereas the American Thoracic Society uses a criterion of ≥5 mm for transplant recipients (29). However, previous studies from intermediate TB-burden countries used a cut-off of 10 mm because of high false-positive rates when lower cut-offs were employed (6–10,22,28). Had we used a TST-reactivity cut-off of ≥ 5 mm, an additional 10% of KT recipients would have received INH prophylaxis, but the three cases in which active TB developed would not have been prevented (Table 2).
The local epidemiology of TB and the prevalence of HIV play important roles in the development of TB in transplant recipients. Therefore, it is worth mentioning the local epidemiology of TB and the prevalence of HIV in our study population. South Korea is an intermediate TB-burden country (90 cases per 100 000 Korean persons), but the burden of TB in Korea is higher than in other developed countries (i.e. 30 per 100 000 populations in Spain, 22 per 100 000 populations in Japan, 15 per 100 000 populations in UK and 4 per 100 000 populations in the USA) (20), while the prevalence of HIV infection is low (< 0.1%). The incidence of TB was found to be 4.6% in Korean kidney transplant recipients (21), which is higher than the 0.7% in Spanish kidney transplant recipients (30). We therefore suggest that further studies are needed to confirm our findings in regions with different TB epidemiologies of TB and HIV prevalences.
A previous study reported that positive TST results were associated with BCG vaccination status in hemodialysis patients (31,32). However, in agreement with our previous observations (22), we found no association between BCG vaccination status and positive TST results in the KT recipients. The reason for this discrepancy is not clear, but may relate to regional differences in epidemiologic settings (i.e. a high BCG coverage rate, intermediate TB-burden country and high proportion of immunocompromised patients in our study). Interestingly, we found that BCG vaccination status was not associated with the development of TB after transplantation. Indeed, the true efficacy of the BCG vaccine has been debated for decades (33). Furthermore, few studies have addressed the effect of BCG vaccination on the development of TB in transplant recipients.
IGRAs are more expensive than TSTs. However, health economics analyses have found that they are cost-effective as they reduce the number of individuals needing unnecessary chemoprophylaxis and monitoring whilst on drug therapy (2). Contact studies have also shown that the strategy of targeting chemoprophylaxis to recent IGRA-positive contacts may prevent a similar number of cases to the use of TST but requires treatment of fewer contacts (12). However, since the positive rate for ELISPOT assays in KT recipients is threefold higher than for TST because of anergy to the TST in transplant recipients, the cost-effectiveness of INH treatment based on IGRA results is not known. In this context, the high risk of TB observed in the ELISPOT-positive KT patients with negative TST highlights the critical need for health economics analysis and our ongoing prospective randomized trial of INH treatment based on IGRA results in this population (clinical trial No. NCT01087190), because this study was not designed to evaluate the health economics of IGRA or the effectiveness and toxicity of INH prophylaxis.
This study has a few limitations. First, although we did screen the living KT donors for chest-X abnormalities and interviewed living related donors to identify any clinical risk factors for LTBI, we did not perform TST or ELISPOT assays on the transplant donors. In addition, some workers have proposed that a relatively short interval between transplantation and the development of TB points to donor-derived TB or active TB missed at enrollment; the latter, however, is not likely considering the strict exclusion criteria we used. Thus, donor-derived TB after KT could not be completely ruled out. However, only 4% of active TB after organ transplantation is of donor origin (16). In intermediate TB-burden countries, the need for donor screening for LTBI is controversial because INH toxicity to recipients would outweigh a small potential risk of donor-derived TB transmission due to the high prevalence of LTBI in healthy transplant donors. In addition, an indication to prescribe INH prophylaxis 6–9 months before kidney donation following donor screening for LTBI would only delay the KT with the negative consequences that that could involve for recipients (34). Indeed, one study from Mexico demonstrated that the use of INH prophylaxis in TST-positive donors made no difference to the risk of transmission of TB to the recipient or the development of TB after KT (35). Second, because of the long wait associated with shortages of donor organs in Korea (several years after initial registration and evaluation for transplantation), ELISPOT assays could not be performed before transplantation in deceased-donor KT or pancreas-kidney transplant recipients. In those patients, ELISPOT assays were performed 1 or 2 days after emergency transplant surgery, in contrast to living-donor KT, where ELISPOT assays were performed 2–5 days before the scheduled transplant surgery. Because of this difference, induction immunosuppressive therapy might have affected the results of ELISPOT assays in the deceased-donor KT or pancreas-kidney transplant recipients. However, when we analyzed a subgroup that included only living-donor KT, this subgroup exceeded the expected sample size and also yielded similar results (Table 2), indicating that these factors did not substantially affect our study results. Other potential limitations of the study include its single-center nature, the small number of active TB cases included, and the fact that it was conducted in an intermediate-burden TB country.
In conclusion, our longitudinal study demonstrates that positive ELISPOT results predict the subsequent development of TB in KT recipients who lack clinical risk factors for LTBI or in whom LTBI cannot be detected by TST.
Funding sources: Basic Science Research Program through National Research Foudation (NRF) funded by the Ministry of Education, Science and Technology (MEST) (grant 2008-E00136).
The authors of this manuscript have no conflicts of interest to disclose as described by the American Journal of Transplantation.
- 18Guidelines for the prevention and management of infectious complications of solid organ transplantation. Am J Transplant 2004; 4(Suppl 10): 164–166.
- 20World Health Organization. Available at: http://extranet.who.int/sree/Reports?op=Replet&name=%2FWHO_HQ%2FSTB_TME%2FPublic%2FTBCountryProfile&ISO2=KR&outtype=html [accessed 28 February 2011]
- 21Tuberculosis complicating kidney transplantation. J Korean Soc Transplant 1995; 9: 95–102 [Korean., , , et al.
- 23Korea Centers for Diseases Control and Prevention. TB control guideline 2010. (Accessed July 14, 2010, at http://tbnet.cdc.go.kr/cm/information/bbs/Bbs_TB_Indicate_DT.dm.)
- 29Targeted tuberculin testing and treatment of latent tuberculosis infection. Am J Respir Crit Care Med 2000; 161(Pt 2): S221–S247.
- 33Tuberculosis vaccines. In: Plotkin S, Orenstein W, Offit P eds. Vaccines. Philadelphia , PA : Saunders/Elsevier 2007: 857–885., , .