Diagnosis and Treatment of Latent Tuberculosis Infection in Adults in South Korea

1. Exclusion of active tuberculosis

Current diagnostic tests for LTBI, including the TST and IGRA, can yield positive results in both active TB and LTBI [1]. Therefore, LTBI should be diagnosed only after excluding active TB in individuals with positive TST and IGRA. A baseline evaluation to rule out active TB should include a detailed medical history (including TB treatment history, contact with patients with pulmonary TB, and symptoms suggestive of TB [cough, sputum, hemoptysis, weight loss, and night sweats]) and physical examination, along with imaging, such as chest X-ray. If active TB is suspected, additional tests, such as microbiological examinations (acid-fast bacillus smear/culture tests and Xpert MTB/RIF test), and if necessary, computed tomography, should be performed to confirm active TB. If the diagnosis remains unclear, referral to a specialist should be considered. In situations where active TB cannot be definitively excluded, the diagnosis and treatment of LTBI must be deferred.

2. Diagnosis of LTBI in immunocompetent adults

Recent guidelines published by the World Health Organization (WHO) or Canada recommend that LTBI in immunocompetent adults can be diagnosed using either TST or IGRA [2-5]. We also recommend using either test for LTBI diagnosis in immunocompetent adults.

However, in South Korea, infants are recommended to receive the BCG vaccination at 4 weeks of age as part of the national standard immunization schedule. Given that TST can yield false-positive results due to BCG vaccination, TST results should be interpreted with caution. Due to possible false-positive TST results caused by BCG vaccination, guidelines commonly recommend using IGRA, rather than TST, for LTBI diagnosis in individuals with a history of BCG vaccination [2-5]. However, if BCG was administered only once during infancy (within the first 12 months of life), the likelihood of a false-positive TST result due to BCG decreases significantly by the time the individual reaches 10 to 15 years of age, and becomes negligible [6,7]. Therefore, in adults who received only one BCG vaccination during infancy, LTBI can be diagnosed using either TST or IGRA (Figure 1). For the TST, the criterion for a positive result is an induration size of ≥10 mm.

Fig. 1.

Diagnosis of latent tuberculosis infection (LTBI) in immunocompetent adults. For adults who received a single Bacillus Calmette-Guérin (BCG) vaccination during infancy or were not vaccinated, either tuberculin skin test (TST) or interferon gamma release assay (IGRA) can be used. For adults with a history of two or more BCG vaccinations or additional BCG vaccination after infancy, IGRA alone is recommended for diagnosing latent tuberculosis infection. If the BCG vaccination history is uncertain, IGRA alone is recommended for diagnosing latent tuberculosis infection.

On the other hand, adults who received BCG vaccination more than twice, or had additional vaccinations after infancy, have a higher likelihood of false-positive TST results. A recent study has reported that this false-positive effect can persist for up to 55 years after vaccination [7]. Therefore, in such cases, using IGRA rather than TST is preferable for LTBI diagnosis (Figure 1) [2]. In South Korea, the standard immunization policy was revised in 1985 [8]. Since then, individuals born after that year are recommended to receive BCG vaccination only once during infancy. If a TST test indicates a positive result in an adult who received the BCG vaccination more than twice, an additional IGRA test can be performed to rule out a false-positive TST result due to BCG. However, if the IGRA test is performed after a certain interval following the TST, caution is required, as the purified protein derivative used in the TST may cause a false-positive IGRA result (the boosting phenomenon) [9]. Since boosting can occur from 2 to 4 weeks after the TST, performing the IGRA test within approximately 1 week is advisable if it must be conducted following the TST [9]. If the additional IGRA test is negative, it can be considered a false-positive TST due to the BCG vaccine, thus ruling out LTBI. Conversely, if the additional IGRA test is also positive, a diagnosis of LTBI can be confirmed. Nevertheless, the TST induration ≥15 mm is more likely to indicate an actual M. tuberculosis infection, rather than a false-positive due to the BCG vaccine [10]. In such cases, LTBI can be diagnosed without an additional IGRA test.

However, obtaining accurate information regarding prior BCG vaccination history is often difficult. Additionally, since a scar does not always form after BCG vaccination, accurately determining prior BCG vaccination status or the number of vaccinations based solely on the presence of a scar is complicated. In cases where prior BCG vaccination history is uncertain, using IGRA alone for LTBI diagnosis is recommended.

In immunocompetent adults with spontaneously healed pulmonary TB lesions detected on a chest X-ray, if there is no history of prior TB treatment, or if the prior treatment was inadequate, LTBI can be diagnosed based on positive TST or IGRA.

3. Diagnosis of LTBI in immunocompromised adults

An immunocompromised adult refers to individuals in specific circumstances, such as those living with HIV, patients scheduled for treatment with biologics or small molecule inhibitors, individuals taking or expected to take immunosuppressive therapy following organ transplantation, patients with autoimmune diseases requiring long-term steroid use, individuals with end-stage renal disease undergoing dialysis, and patients with hematologic malignancies. For these populations, diagnostic methods different from those recommended for immunocompetent adults are recommended.

A recent meta-analysis has reported that in cases when the TST result is negative (<10 mm), the incidence of TB is 5.1 times higher if the additional IGRA test is positive, compared to the negative IGRA result [11]. The same study also reported that when the IGRA result is negative, the incidence of TB is 3.6 times higher if the additional TST is positive, compared to the negative TST result [11]. Reflecting this study result, recent guidelines recommend that a negative TST or IGRA alone should not be considered sufficient to rule out LTBI in immunocompromised adults [2,4,5]. Instead, performing an additional test to confirm the presence or absence of LTBI is recommended (i.e., if the TST is negative, perform an IGRA; if the IGRA is negative, perform a TST) [2,4,5]. This is because in immunocompromised adults at a higher risk of progressing to active TB, increasing the sensitivity of the LTBI diagnosis is crucial.

However, there is a higher likelihood of false-negative results in LTBI tests in immunocompromised adults, compared to immunocompetent adults. This likelihood of false-negative results is significantly higher with TST than with IGRA [12-14]. Therefore, due to the high likelihood of false-negative results, we do not recommend initially using TST alone for LTBI diagnosis in immunocompromised adults. In this guideline, it is instead recommended to perform the IGRA first. If the IGRA result is positive, a diagnosis of LTBI can be made. If the IGRA result is negative, a TST should be performed, determining the presence or absence of LTBI (Figure 2). Alternatively, both tests can be conducted concurrently from the beginning, and the diagnosis of LTBI can be made following the same diagnostic flow. For the TST, the criterion for a positive result is an induration size of ≥10 mm, regardless of BCG vaccination status (with a 5-mm threshold for people living with HIV).

Fig. 2.

Diagnosis of latent tuberculosis infection (LTBI) in immunocompromised adults. Initially, a patient should be tested with interferon gamma release assay (IGRA) alone; if the result is negative, the diagnosis of LTBI is determined based on the tuberculin skin test (TST) result. Alternatively, performing both IGRA and TST from the beginning is possible, interpreting the results according to the same algorithm.

In immunocompromised adults with spontaneously healed pulmonary TB lesions detected on a chest X-ray, if there is no history of prior TB treatment, or if the prior treatment was inadequate, a diagnosis of LTBI can be made, even if both TST and IGRA results are negative.

In immunocompromised adults, the IGRA result may be reported as indeterminate. In such cases, either repeating the IGRA, or performing a TST, is recommended to diagnose LTBI [2].

4. Diagnosis of LTBI in individuals with a history of positive LTBI test or tuberculosis treatment

If an individual previously tested positive for LTBI, repeating the test is unnecessary, as it is likely to yield a positive result again. Additionally, individuals treated for active TB or LTBI in the past can still test positive on LTBI tests, regardless of whether the treatment was adequate. This is because current LTBI tests cannot distinguish whether M. tuberculosis has been fully eradicated or partially cleared, or if there is a new infection following treatment for LTBI or active TB.

Therefore, in cases where an individual with a history of LTBI or active TB treatment requires an assessment and treatment for LTBI, the decision should be based on a comprehensive evaluation that includes the previous treatment regimen, treatment duration, the individual’s risk of developing TB, and occupational factors.

5. Interpretation of the IFN-γ level in IGRA

Among the currently used IGRAs, the QuantiFERON-TB Gold Plus test (Qiagen, Hilden, Germany) involves collecting the patient’s peripheral blood and adding it to tubes containing TB antigens tube (containing early secretory antigenic target-6 [ESAT-6], culture filtrate protein 10 [CFP-10]), a negative control tube (nil tube), and a positive control tube (mitogen tube). The blood is then incubated for 16 to 24 hours, and the concentration of IFN-γ is measured using ELISA. The measured difference of IFN-γ concentration between the TB antigen tube and nil tube is reported as positive if ≥0.35 IU/mL, and negative if <0.35 IU/mL. Previously, this value itself was considered to have little clinical significance. However, several recent studies have reported that the interpretation can vary, based on the IFN-γ result value itself.

First, if the result is positive with a relatively low IFN-γ level (e.g., <0.7 IU/mL), repeating the test could yield a negative result (i.e., <0.35 IU/mL) [9,15]. Similarly, if the IGRA result is negative with the IFN-γ level slightly below the cut-off (e.g., around 0.2 IU/mL), repeating the test could provide a positive result (i.e., ≥0.35 IU/mL). Thus upon retesting, relatively low result values near 0.35 IU/mL can change from positive to negative (or vice versa). These values are referred to as the borderline zone, and are thought to be due to immunological responses or technical issues, rather than the LTBI itself [1,15]. The borderline zone is most problematic when IGRA tests are repeated, particularly in healthcare workers who undergo annual testing. In this group, nonspecific conversions and reversions due to the borderline zone are not uncommon [16]. If the IGRA result converts from negative to positive within the last 2 years, it is generally considered to be a case requiring LTBI treatment [17]. However, values within the borderline zone (≤0.7 IU/mL) may revert to negative upon retesting. Therefore, in special cases when IGRA tests are regularly repeated, such as among healthcare workers, the borderline zone can be considered to decide whether to retest or initiate treatment [18]. Further research is needed on this matter. In contrast, in groups where IGRA is performed only once without repetition (e.g., contact investigation or high-risk group screening, such as organ transplant recipients), interpretation of the results without applying the borderline zone is recommended.

Moreover, recent studies have revealed that the risk of developing active TB can vary, based on IFN-γ level. That is, a higher IFN-γ level may indicate a greater risk of developing active TB, compared to a lower level [19]. A meta-analysis of 34 studies reported that an IFN-γ level of 0.35 IU/mL had a relative risk of 1.64 for developing active TB compared to an IGRA-negative result, whereas a concentration of 4 and 7 IU/mL had a relative risk of 9.32 and 15.07, respectively, indicating a significant increase in the risk with higher IFN-γ values [20]. Therefore, when deciding on LTBI treatment, the IFN-γ level from the IGRA can be an important factor to consider.

1. Daily rifampin monotherapy for 4 months

In a randomized controlled trial (RCT) conducted in TST-positive and HIV-negative adults with silicosis, shorter regimens with rifampin (3 months of daily rifampin [3R], 3HR) and 6 months of INH (6H) showed higher TB prevention rates of 63%, 48%, and 41% in 3R, 3HR, and 6H, respectively, compared to the placebo-controlled group [21]. However, the differences among the treatment groups were not statistically significant [21]. Several recent studies have reported that the 4R regimen was superior to the daily INH monotherapy for 9 months (9H) regimen in terms of treatment completion rate and adverse events (AEs), including hepatotoxicity [22,23]. In a large RCT involving more than 6,800 patients (including Korean patients), 4R was found to be non-inferior to 9H in TB prevention (<0.01/100 person-years in both groups), and was superior in treatment completion (78% in 4R vs. 63% in 9H) with fewer AEs [24]. Furthermore, in high-income countries, the healthcare cost of 4R was lower than that of 9H, with a ratio of 0.76 (95% confidence interval [CI], 0.70 to 0.82) [25]. Accordingly, 4R is recommended as first-line therapy for LTBI in many countries, including the United States (US), Canada, and Taiwan [26-28]. However, a potential disadvantage of rifampin-based regimens is the drug-drug interaction with a wide range of drugs. Therefore, before initiating LTBI treatment with a rifampin-based regimen, it is essential to assess current medications and potential drug interactions.

2. Daily rifampin and isoniazid combination therapy for 3 months

The 3HR regimen is recommended in the guidelines of the US, WHO [27,29], and the United Kingdom (UK), based on findings that long-term (≥10 years) use of 3−4HR in HIV-negative children is both effective and safe [30]. A meta-analysis of RCTs found that 3−4HR and 6H were equivalent in treatment effectiveness [31]. In a study of HIV-negative Greek children over an 11 year period, patient adherence to 3−4HR was superior to that to 9H, and new radiographic findings suggestive of active disease were lower in 3−4HR than in 9H (24% vs. 11%−13.6%) [32]. The incidence of active TB did not differ between 3HR and 4HR regimens. For the patient’s preference, 78.7% of children and adults aged <35 years chose 3HR over 6H when given the option, and the completion rate was higher in those who had a choice and in those using 3HR in a UK study [33]. However, unlike other regimens, 3HR has not been investigated in HIV-negative adults in a large RCT. Additionally, although hepatotoxicity was less likely to occur with 3HR compared to 6H or 9H, the rate of discontinuation due to other AEs was reportedly higher in those with 3HR [34].

3. Once weekly rifapentine and isoniazid combination therapy for 3 months

In late 2011, the US Centers for Disease Control and Prevention (CDC) recommended a 3-month intermittent therapy with INH and rifapentine as a new regimen for LTBI [35]. Rifapentine, a derivative of rifamycin, is ideal for intermittent treatment, due to its longer half-life, and equivalent anti-TB effectiveness. This recommendation is based on three prospective studies showing that rifapentine and INH combination therapy for 3 months (3HP) did not differ in treatment effectiveness from 9H, but had a better completion rate [36,37]. Compared to INH monotherapy, regimens including rifapentine have less hepatotoxicity, but more systemic AEs, such as flu-like syndrome [38,39]. In South Korea, rifapentine was introduced in 2016, and investigated for the treatment of LTBI among healthcare workers. There were more cases of flu-like syndrome, and four instances of anaphylaxis in individuals treated with 3HP [40]. Therefore, further research is necessary before broadly introducing rifapentine into South Korea.

4. Daily rifapentine and isoniazid combination therapy for 1 month

A 1-month regimen of daily INH/rifapentine therapy (1HP) was compared to the standard 9H regimen in HIV-positive patients [41]. The 1HP regimen was non-inferior to 9H in preventing TB, and had fewer AEs and a higher completion rate [41]. Although research has not yet been conducted with HIV-negative patients, the WHO recommends the 1HP regimen for the treatment of LTBI, regardless of HIV status [29].

5. Daily isoniazid monotherapy for 9 or 6 months

The efficacy of INH treatment for LTBI was established in the 1960s, and mono-INH therapy remained the standard treatment for over 50 years. Numerous RCTs have shown its preventive effect on TB. Nevertheless, INH preventive therapy has significant limitations, including a long duration that leads to high discontinuation rates and hepatotoxicity.

Summarizing the findings of studies conducted in the US, Comstock [42] concluded that compared to 6H, 12H is more effective in preventing TB, and the optimal duration of INH therapy is 9 to 10 months, as no additional benefit was observed beyond 12 months [42,43]. Therefore, 9H is recommended as the first-line treatment for LTBI.

However, 6H has also been shown to be effective and cost-efficient. A meta-analysis demonstrated that 6H was non-inferior to 9H in terms of effectiveness and hepatotoxicity, and had better patient adherence [44,45]. Despite this data, no clinical trials have directly compared 6H and 9H, and the preventive effect of 6H on TB was found to be approximately 64% that of 9H [43]. Moreover, a recent meta-analysis concluded that 9H was more effective, and had fewer AEs than 6H [46]. Therefore, while 6H can be used as an alternative regimen, it is recommended only when 9H cannot be used.

6. The choice of treatment regimen in South Korea

The results of the drug susceptibility test for the index case should guide the choice of drugs for the treatment of LTBI. For the contacts of patients with drug-susceptible TB, a shorter rifampin-based regimen, either 4R or 3HR, is recommended as the preferred regimen for the treatment of LTBI in South Korea, based on recent clinical trials and real-world data. When rifampin-based regimens cannot be used due to intolerance, feasibility issues, or contraindications, 9H can be an alternative option. Additionally, 6H may be considered for patients who are unlikely to complete the 9H treatment due to hepatotoxicity or other AEs. Table 2 summarizes the recommended treatment regimens for LTBI.

Table 2.

Summary of recommended treatment regimens for latent tuberculosis infection

1. Pretreatment evaluation

Before treating LTBI, it is critical to exclude active TB, to avoid undertreatment and the subsequent development of drug resistance. The initial assessment should include drug allergies, comorbidities, hepatitis status, alcohol consumption, and contact with multidrug-resistant TB (MDR-TB). To evaluate the risk of drug-drug interaction with rifampin, current medication should be reviewed. Rifampin induces the production of cytochrome P450 enzymes, which decreases the serum levels of other drugs by enhancing their metabolism. Important categories of interacting medications include antihypertensives, anticoagulants, antifungal agents, methadone, immunosuppressive agents, hormonal contraceptives, and antiretrovirals. These medications may need adjustment, cessation, or substitution during LTBI treatment.

For baseline tests, complete blood count (CBC), aspartate transaminase (AST)/alanine transaminase (ALT), bilirubin, and hepatitis B and C tests should be performed. HIV testing is recommended for patients at high risk of HIV infection. Patients should also be informed about potential AEs, and instructed to immediately discontinue antitubercular agents if severe AEs occur.

2. Evaluation during treatment

Evaluation at the end of the first month of treatment is essential to assess medication tolerability and encourage adherence. Liver function and bilirubin tests must be performed, regardless of the regimen used. If a rifampin is used, CBC should also be performed. If baseline liver function tests are abnormal, or if risk factors for hepatotoxicity are present, liver function tests should be performed monthly.

3. Management of adverse events

For mild-to-moderate AEs, frequent monitoring while continuing treatment is recommended. If severe AEs occur, limiting daily activities or life-threatening, treatment should be discontinued, either until recovery from the AEs, or permanently. Regarding liver function, treatment should be stopped if AST or ALT levels exceed five times the upper limit of normal without symptoms of liver dysfunction, or three times the upper limit of normal with symptoms. Once the patient recovers, an alternative regimen should be considered.

4. Treatment completion

LTBI treatment completion is defined as taking 80% of the prescribed dosage within a set period, depending on the regimen: 12 months for 9H, 6 months for 4R, and 4 months for 3HR. For 3HP, completion is defined as taking the drugs at least 11 times in 16 weeks [37]. Currently, there are no tests to confirm the outcome of LTBI treatment after completion. Consequently, neither TST nor IGRA should be used to assess treatment effectiveness.

1. Pregnant and breastfeeding patients

Data regarding the use of most LTBI treatment regimens during pregnancy are limited. Although no firm guideline for LTBI treatment during pregnancy exists, women with a high risk of progression to active TB should not delay LTBI treatment due to pregnancy [47]. Treatment should begin in the first trimester for women who are HIV-positive, recent contacts of patients with active TB, or immunocompromised. LTBI treatment should start after the first trimester for those who have had a TST conversion within the past 2 years. Most studies on LTBI treatment during pregnancy have investigated the 9H regimen. An observational study reported an increased risk of hepatotoxicity in women treated with INH during pregnancy, or within 3 months postpartum [48,49]. However, INH therapy did not increase the risks of grade 3 or 4 AEs, hepatotoxicity, and death in HIV-negative pregnant women [50]. A RCT that included HIV-positive pregnant women found poorer pregnancy outcomes, such as stillbirth, spontaneous abortion, or congenital anomalies, in the group that received INH for 28 weeks (odds ratio, 1.51; 95% CI, 1.09 to 2.10) [49]. However, another observational study reported that the INH group had fewer adverse pregnancy outcomes, and a lower (by 30%) risk of active TB [51]. No data have been published regarding the use of rifampin for LTBI treatment during pregnancy. However, rifampin is considered safe for treating active TB during pregnancy, suggesting that it may also be safe for LTBI treatment [52]. Pregnant women taking INH should visit the hospital at least once a month, and be thoroughly examined for symptoms and signs of hepatitis. They should take pyridoxine daily to prevent peripheral neuropathy. Both INH and rifampin are excreted in breast milk in small quantities, well below the usual therapeutic neonatal dose. Pyridoxine is recommended for breastfed infants whose mothers are taking INH, but who are not on INH themselves.

2. Contacts of drug-resistant TB cases

A meta-analysis of 21 studies, including five comparative studies, showed that among contacts with MDR-TB patients, TB occurred in 1.1% of the treatment group, and 14.3% of the non-treatment group [53]. Fluoroquinolone and ethionamide combination was the most effective treatments, but it was significantly more expensive than any other regimen [53]. An observational study compared 104 MDR-TB contacts treated with fluoroquinolone for 12 months to 15 MDR-TB contacts who refused treatment in the Federated States of Micronesia; after 36 months, TB occurred in three individuals in the non-treatment group, and none in the treatment group [54]. Another study in Pakistan with family contacts of MDR-TB patients treated with fluoroquinolone for 6 months found that TB occurred in two individuals over 2 years [55]. However, RCTs have not yet been completed on this topic. Future recommendations will be based on findings from large-scale RCTs currently being conducted using 6-month levofloxacin (The V-QUIN MDR trial and TB-CHAMP) and delamanid (PHOENIx MDR-TB) [56-58]. Contacts of MDR-TB cases should be evaluated for active TB and LTBI, and followed up for at least 2 years. Contacts of INH mono-resistant cases can be treated with 4R, while contacts of rifampin mono-resistant cases may be treated with 9H.

3. Patients with at risk of hepatotoxicity

Risk factors for hepatotoxicity include liver transplantation, history of severe liver disease, history of drug-induced liver disease, hepatitis B or C infection, alcoholic hepatitis, fatty liver disease, and liver cirrhosis. The decision to treat patients with these risk factors should be based on a careful weighing of the potential risks and benefits of LTBI treatment.

4. Patients aged 65 years or older

The risk of toxicity, particularly hepatotoxicity from INH, increases with age. Previously, the 4th Korean TB guideline did not recommend LTBI treatment for patients aged 65 years or older. However, recent several studies have reported that LTBI treatment can be completed safely in this age group.

A study conducted in a tertiary hospital in South Korea reported LTBI treatment outcomes in patients with immune-mediated inflammatory diseases by age group [59]. The completion rate in patients aged ≥60 years was acceptable, although lower compared to those aged <60 years (85.1% vs. 91.9%). While AEs occurred more frequently in patients aged ≥60 years compared to those aged <60 years (22.9% vs. 9.8%), most AEs were not severe, and little difference was observed regarding hepatotoxicity between the groups (2.0% vs. 2.2%) [59].

Another retrospective multicenter study on LTBI treatment in patients aged >65 years found completion rates of 88.4%, 77.3%, and 75% for 3HR, 4R, and 9H regimens, respectively [60]. Treatment was discontinued due to AEs in two patients treated with 4R, and one treated with 3HR. The three regimens did not differ significantly in hepatotoxicity rates (9.1%, 16.7%, and 11.6% for 4R, 9H, and 3HR, respectively) [60].

A prospective observational study analyzing 406 Taiwanese patients reported an 81.1% completion rate for 3HP, and a 68.6% completion rate for 9H in those aged >60 years (n=167), without significant differences in systemic AEs, including hepatotoxicity [61]. However, in patients aged ≥80 years, the treatment discontinuation rate was higher, due to systemic AEs [61]. In a post hoc study with large-scale RCTs, hepatotoxicity increased with age in the INH group, but was not significantly correlated with age in the rifampin group [62]. For the 3HP regimen, it is unclear whether AEs are increased in the older population, and studies have reported mixed findings. A prospective observational study in Taiwanese patients demonstrated that the oldest group with ages 65 years or older had the lowest completion rate (73.9%), but did not differ significantly in treatment discontinuation rate or hepatotoxicity, compared to the younger groups [63]. Systemic AEs were most common in the middle-aged group 35 to 64 years, while hypertension accompanied by flu-like syndromes was most common in the oldest group 11.2% [63]. However, a Chinese study comparing 3HP, 2HP (twice weekly), and placebo in 3,738 patients aged 50 to 69 years was terminated early, due to AEs in the HP groups [64].

Consequently, LTBI treatment can be considered for patients aged ≥65 years, with careful weighing of individual risks and benefits. For this age group, the 4R regimen is preferred, due to its lower risk of hepatotoxicity.