Introduction

Tuberculosis (TB) and chronic hepatitis B virus infection (HBV) are preventable infectious diseases that impact public health and clinical medicine. Worldwide, there are an estimated 10 million people with active TB disease [1], 1.7 billion people with latent TB infection (LTBI) [2, 3], and 257 million people with chronic HBV [4]. During 2019, the USA reported 8920 TB cases, which represented a TB incidence rate of 2.7 cases per 100,000 population [5]. TB disproportionately impacts non-US-born persons, and especially non-Hispanic Asians; rates of TB are 15-fold and 35-fold higher among non-US-born and non-Hispanic Asians compared to US-born and Whites, respectively [5]. Although the USA has one of the lowest TB rates in the world, elimination of TB, defined as a TB rate of < 1 per 1,000,000 population, has been difficult to achieve in the USA, in part, because of challenges in identifying, screening, and treating persons with LTBI [6, 7]. An estimated 80% of TB cases in the USA result from LTBI, and 13 million persons in the USA are estimated to have LTBI [5, 8]. Newer and shorter course LTBI regimens such as rifampin and isoniazid and rifapentine have a lower risk of hepatotoxicity, higher treatment completion rates, and similar efficacy as isoniazid for prevention of active TB disease and are now the preferred LTBI treatment regimens [9]. Despite these advances in LTBI treatment regimens, there remain substantial gaps in the LTBI cascade of care as a recent meta-analysis estimated that only 19% of patients complete an adequate course of LTBI treatment [10].

Chronic HBV also has significant gaps in the cascade of care and disproportionately impacts non-US-born and Asians [11]. During 2019, an estimated 1.59 million US persons were infected with chronic HBV [12]. Non-US-born persons and non-Hispanic Asians have a disproportionately higher prevalence of chronic HBV infection compared to US-born (1.28% vs. 0.15%) and non-Hispanic Whites (3.85% vs. 0.08%), respectively [11]. Among chronic HBV-infected patients in the USA, 32% of patients were aware of their HBV diagnosis, and among those who were aware of their HBV diagnosis, 28% were started on HBV antiviral treatment [13]. Although an estimated 36% of insured chronic HBV patients are eligible for antiviral treatment, only 18% were prescribed treatment [14]. Reductions in chronic HBV burden will require increased screening and linkage to care and treatment among non-US-born persons.

Although LTBI and chronic HBV are common preventable infectious diseases in the USA that disproportionately affect non-US-born persons, there are few published data regarding LTBI and chronic HBV co-infection [15, 16]. In this paper, we conducted a systematic review and meta-analysis to estimate the prevalence of LTBI among chronic HBV-infected adults.

Methods

Search Strategy and Study Selection

This study was conducted and presented according to Meta-analysis Of Observational Studies in Epidemiology (MOOSE) [17] and Preferred Reporting Items for Systematic Review and Meta-Analyses (PRISMA) guidelines [18]. We conducted structured searches of MEDLINE/PubMed from inception to September 1, 2019, to identify English language peer-reviewed original research articles that included adults (age > 18 years) by using different combinations of keywords for hepatitis B virus and tuberculosis. The search strategy included all of the following terms:

  1. 1.

    (Tuberculosis[mesh] OR Antitubercular agents[mesh] OR Interferon-gamma release tests[mesh] OR Mycobacterium tuberculosis[mesh] OR Tuberculin[mesh] OR Tuberculin test[mesh] OR tuberculosis OR antitubercul* OR “interferon gamma release test” OR “interferon gamma release tests” OR “mycobacterium tuberculosis” OR tuberculin OR mantoux OR “TB test” OR “TB tests” OR “anti-TB” OR Quantifero* OR qft-git OR t-spot.tb OR TB[ti]) AND

  2. 2.

    (Hepatitis B[mesh] OR Hepatitis B virus[mesh] OR Hepatitis B antigens[mesh] OR Hepatitis B antibodies[mesh] OR “hepatitis b” OR HBV[tiab] OR HBcore OR “HB core” OR “homologous serum hepatitis virus” OR “dane particle” OR HBAg OR HBeAG OR “hepatitis Be antigen” OR “hepatitis Be antigens” OR “e antigen” OR “e antigens” OR HBeAg OR “HBNE Ag-1” OR “HBe Ag-2” OR “Australia antigen” OR “Australia antigens” OR “Anti-Australia antigen” OR “Anti-Australia antigens” OR “anti-HBAg”) AND

  3. 3.

    (English[la])

In addition, we conducted a detailed manual review of the bibliographies of included articles and relevant systemic reviews to identify additional articles. We included both prospective and retrospective observational studies, including cohort and cross-sectional study designs. When two or more studies reported data on the same cohort over the same study period, we utilized the study that included the most comprehensive cohort and study periods. All literature searches, reviews, and selections were conducted independently by two authors (RJW and AH). The outcome of each stage of literature search was reviewed independently by a third author (ASC), who also resolved any discrepancies.

Data Extraction

Three authors independently (AH, RJW, ASC) extracted data from the included studies using a standardized data extraction form. Data extracted included year of study, country/region, study design, study period, aim of the study, study population, and diagnostic criteria for HBV and LTBI. We also extracted demographic data on patients including age, sex, and country of origin; the number of patients with chronic HBV; and the number of patients with LTBI. If available, data regarding the proportion of patients with HIV, HCV, or concurrent diabetes were also collected. For studies with missing data or incomplete data that precluded calculation of LTBI prevalence in patients with concurrent chronic HBV, corresponding authors were contacted to obtain necessary data. Any discrepancies in data extraction were resolved in consultation with a separate reviewer (RS).

Statistical Analysis

Prevalence of LTBI among patients with chronic HBV was calculated by dividing the number of patients with LTBI by the total number of patients with chronic HBV. Aggregate data on LTBI prevalence among chronic HBV-infected patients were evaluated using meta-analysis, and pooled prevalence rates were generated using random-effects models. All analyses were conducted using Review Manager (RevMan, Version 5.3. Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration).

Results

Of the 1205 studies identified by our search strategy, 1145 (95%) studies were excluded based on title and abstract review, and 60 (5%) studies were reviewed to determine whether they included data regarding chronic HBV and LTBI (Fig. 1). Of the 60 studies, 54 (90%) were excluded based on review of the full papers. A total of 6 studies were included in the systematic review and meta-analysis of LTBI among adults with chronic HBV [15, 16, 19,20,21,22]. These 6 studies included complete data required to calculate LTBI prevalence among patients with concurrent chronic HBV, and as a result the corresponding authors of these studies were not contacted.

Fig. 1
figure 1

PRISMA flow diagram of studies included in the meta-analysis of latent tuberculosis infection prevalence among adults with chronic hepatitis B virus infection

Full size image

A summary of the demographic and clinical characteristics of the persons included in the 6 studies is provided in Table 1. Three of the studies were cross-sectional, and three were cohort studies. Three of the 6 studies were published between 1986 and 1991, and three of the 6 studies were published during 2002–2019. Five of the 6 studies were conducted in North America: two of these studies only included Southeast Asian immigrants, predominantly from Vietnam; two studies included Alaskan and Northwest Canadian Natives; and one study included homeless US veterans. One study was conducted in Zhongmu County, Central China. The proportion of males in the studies varied from 45 to 53%. Most studies included persons aged 50 years and younger; however, one study only included patients aged 50 years and older. Two studies included limited data on comorbidities. In these studies, the prevalence of hepatitis C virus infection (based on positive hepatitis C virus antibody) among patients with LTBI was 52.03% and 0.49%; HIV infection was 2.44%; and diabetes mellitus type 2 was 7.84%.

Table 1 Characteristics of included studies evaluating the prevalence of latent tuberculosis infection among adults with chronic hepatitis B virus infection
Full size table

All studies defined chronic HBV as persons who had a positive hepatitis B surface antigen (HBsAg) test. Five of the 6 studies defined LTBI as persons who had a tuberculin skin test that was read as having 10 or more millimeters of induration; one study defined LTBI as having a positive QuantiFERON®-TB Gold In-Tube test. The number of persons with chronic HBV in the studies ranged from 7 to 950, and the number of persons with chronic HBV who had LTBI ranged from 1 to 259. The prevalence of LTBI among persons with chronic HBV ranged from 14.29 to 63.40%. In the meta-analysis of all 6 studies, the LTBI prevalence among persons with chronic HBV was 34.25% (95% Confidence Interval: 17.88–50.62%) (Fig. 2).

Fig. 2
figure 2

Forest plot of meta-analysis evaluating the prevalence of latent tuberculosis infection among adults with chronic hepatitis B infection

Full size image

Discussion

In this systematic review and meta-analysis, we detected a pooled prevalence of LTBI among adults with chronic HBV of 34.25%. This prevalence is substantially higher than published estimates of LTBI reported in the National Health and Nutrition Examination Survey (NHANES). During 2011–20,112, NHANES utilized a QuantiFERON®-TB Gold-In-Tube test to detect LTBI among participants, and noted an overall LTBI prevalence of 5.0%, a prevalence of 2.8% among US-born persons, and a prevalence of 15.8% among non-US-born persons [23]. To the best of our knowledge, our study is one of the first to estimate the prevalence of LTBI among persons with chronic HBV, and the findings from these analyses are relevant to clinical medicine and public health for several reasons.

Efforts to reduce morbidity and mortality associated with TB and chronic HBV will require healthcare systems to screen and treat persons at risk of both infections. Several studies have shown gaps in the cascade of care for both LTBI and chronic HBV [10, 14, 24, 25]. In one meta-analysis of the LTBI diagnosis and treatment cascade of care, only 72% of persons intended for TB screening received a TB screening test and 19% of persons diagnosed with LTBI completed an adequate course of LTBI treatment [10]. Primary care providers also reported suboptimal screening rates for HBV infection, and less than 20% of insured patients with chronic HBV were prescribed treatment [14, 24, 25]. Although improvements in the quality of care for LTBI and chronic HBV are needed, targeted efforts to identify individuals at risk of both infections will also be important in reducing morbidity and mortality of both infections. During 2019, there were a reported 9.2% of the US population who were uninsured, and many of these individuals are non-US-born, which is an important demographic risk factor for both TB and chronic HBV [26]. These uninsured individuals may have limited access to medical care, and targeted screening efforts may need to be conducted in community settings. TB and chronic HBV community-based screening efforts have been shown to increase diagnosis and treatment of both infections [27, 28].

Persons with LTBI and chronic HBV have a common demographic risk factor for place of nativity; however, our study also noted that they may have similar medical comorbidities. Xin et al. included data on several comorbidities [22] and detected a high prevalence of diabetes among persons with LTBI. The incidence of TB disease is higher among persons with diabetes, especially among non-US-born persons [29]. Similarly, for those with chronic HBV, an increasing proportion of patients have been noted to have diabetes [30, 31]. These findings suggest that targeted efforts for screening and treatment of LTBI and chronic HBV may be warranted among persons with diabetes who are at risk of both infections. As a result, healthcare systems may consider implementing quality improvement efforts for LTBI and chronic HBV screening and treatment among persons with diabetes. Because there are several Centers for Medicare & Medicaid Services Quality Measures regarding diabetes [32], there also may be financial incentives for healthcare systems to demonstrate that they can improve the diagnosis and treatment of other comorbidities among diabetic persons such as LTBI and chronic HBV.

Co-infection with chronic HBV may pose certain challenges for clinicians with regard to treatment of LTBI. Chronic HBV infection has been noted to increase the risk of hepatotoxicity during LTBI treatment [16, 33,34,35,36]. For example, Patel et al. evaluated 743 Vietnamese immigrants to the USA between 1991 and 1999 who underwent screening for HBV and LTBI [16]. Among this cohort, the prevalence of LTBI among patients with confirmed chronic HBV was 53%. The prevalence of drug-induced liver injury (DILI) from isoniazid prophylaxis treatment was significantly higher in HBV-TB co-infected patients than latent TB-infected patients without HBV (30.0% vs. 3.2%, p < 0.001). In addition, the authors observed that HBV e-antigen positivity was a strong predictor of developing DILI (RR 11.38, 955 CI 5.49–23.59, p < 0.001) [16]. These findings highlight the importance of screening for HBV among LTBI patients, and monitoring liver function tests during LTBI treatment among persons with chronic HBV. Other challenges that may be encountered during LTBI treatment among chronic HBV patients are drug-drug interactions between HBV antiviral therapy and LTBI medications. For example, rifamycin medications, which are commonly used in the treatment of LTBI, may impact the clearance of certain HBV antiviral therapies such as tenofovir alafenamide, which may require changes in HBV or LTBI treatment [37,38,39].

This systematic review and meta-analysis are subject to certain limitations. First, several studies included in the meta-analysis were published more than 20 years ago. These studies may have a higher LTBI and HBV prevalence compared to the current prevalence of both infections because of the implementation of public health measures, such as HBV vaccination. Based on the age groups of patients included in these studies and time periods these studies were conducted, it is likely that these study populations may not have been routinely vaccinated for HBV. Second, while all the studies utilized objective laboratory criteria to identify LTBI and HBV (e.g., tuberculin skin test and QuantiFERON®-TB Gold In-Tube test for LTBI and HBsAg for HBV), for the diagnosis of HBV infection in particular, a single positive HBsAg test to define chronic HBV may have contributed to some degree of misclassification bias (e.g., patients may have had acute HBV infection). Third, the literature search was limited to PubMed® and the English language. Nevertheless, we applied comprehensive and broad search criteria, including review of the references of included studies to ensure capture of all eligible studies for inclusion. Fourth, the studies included in this meta-analysis were primarily limited to persons from the Western Pacific Region of the world. This may limit the ability to generalize our findings to other regions of the world.

In conclusion, LTBI prevalence among adults with chronic HBV was two times higher than the prevalence of LTBI among all non-US-born persons. The high LTBI prevalence and increased risk of hepatotoxicity with TB medications among persons with HBV may warrant consideration of routine screening for HBV among persons who are tested for LTBI. Reducing morbidity and mortality associated with TB and HBV may require healthcare systems and public health to ensure that persons at risk of both infections are screened and treated for LTBI and HBV.