Abstract
Background
Individuals with inactive pulmonary tuberculosis (PTB) lesions were found to be high-risk populations for active PTB development. This retrospective study evaluated the association between different types of inactive PTB lesions and the development of active PTB aiming to provide epidemiological evidence for developing precise intervention strategies.
Methods
Based on a population-based PTB active-case-finding project conducted on the elderly, 154 028 subjects who had participated in the 2020 baseline survey were included in the current analysis.
Results
During the 2-year follow-up, 462 cases developed microbiologically confirmed active PTB with an overall incidence rate of 0.19 per 100 person-years. Among the study population, 15 037 (9.76%) showed chest radiography (CXR) abnormalities suggestive of inactive PTB, which was found to be independently associated with an increased risk of active PTB with an adjusted hazard ratio (aHR) of 6.00 (95% confidence interval (CI) 4.85–7.43) compared with normal CXR. Such a relationship was consistently observed for inactive lesions, including fibrosis, calcification, pleural thickening and nodule lesions with aHRs ranging from 2.94 to 6.55. Inactive PTB lesions alone or a history of anti-tuberculosis (TB) treatment alone were independently associated with the risk of active PTB with aHRs of 6.96 (95% CI 5.59–8.67) and 7.67 (95% CI 4.26–13.78), respectively. A combined effect between inactive PTB lesions and with history of anti-TB treatment was found with an aHR of 10.50 (95% CI 5.93–18.52).
Conclusion
Overall, individuals with inactive PTB lesions, regardless of lesion type and history of anti-TB treatment, are at increased risk of developing active PTB and deserve interventions for TB control.
Shareable abstract
Individuals with inactive PTB lesions, regardless of the lesion types and history of anti-TB treatment, are at increased risk of developing active PTB and deserve further interventions for TB control https://bit.ly/4h9ZL0T
Introduction
Pulmonary tuberculosis (PTB) is a major public health challenge globally [1]. Various host factors are associated with the risk of active PTB development, including impaired immune function, comorbidity or other clinical characteristics, such as having inactive PTB lesions [2, 3]. A meta-analysis revealed that the annual rate of progression to active PTB disease in participants with chest radiography (CXR) suggestive of inactive PTB was 1% (95% confidence interval (CI) 0.3–1.8), which was 10-fold of the incidence among participants with normal CXR [4]. Our previous population-based prospective study consistently observed that middle-aged and elderly residents with inactive PTB only accounted for 3% of the investigated rural population but contributed 30% of the active PTB incidences [5]. The inactive PTB lesions might be left behind either by prior active PTB with a history of treatments or by naturally recovered Mycobacterium tuberculosis (MTB) infections or early PTB without any treatment [6]. For cases with a history of treatments, strengthened active PTB surveillance is an important approach for early case-finding and control of recurrent diseases [7]. For individuals with inactive PTB lesions, but without a treatment history, the re-activation of PTB can be averted by tuberculosis (TB) preventive treatment (TPT), which has been reported with an efficacy ranging from 60% to 90% [8, 9]. Our previous randomised controlled trial (RCT) showed that the protective rates of the short regimen of 6 weeks of twice-weekly rifapentine plus isoniazid among this key population aged ≥60 years were 80.17% (95% CI 25.36–97.96%) [10]. This suggested that giving TPT to individuals with inactive PTB lesions in China may have significant community-level implications for reducing TB incidence. Guidelines for TPT in the United States and Canada suggested that individuals with fibrosis lesions should be targeted for latent TB infection testing and TPT [11, 12]. However, neither the World Health Organization (WHO) nor Chinese guidelines have yet to include them as the target TPT population due to the low quality of evidence [13, 14].
Furthermore, according to national standards and previous studies, the radiological classifications of inactive PTB mainly include 1) calcification lesions; 2) fibrosis lesions; 3) nodule lesions; and 4) pleural thickening [15–20]. Previous studies were mainly focused on the impact of fibrosis lesions [5, 10, 21]. There is a lack of relevant research on whether the risk of active PTB onset and the benefits of TPT are consistent among subgroups with different inactive lesions. This type of research is challenging because it requires a large sample size and prospective design; however, it is crucial for defining the targets for TPT among this important high-risk population. In 2020, Quzhou City in Zhejiang Province launched a 3-year PTB active-case-finding (ACF) project using CXR as a screening tool for rural elderly people aged ≥65 years. This project provides us with a rare opportunity to explore the correlation between different types of inactive PTB lesions and the risk of active PTB, and could provide reference data for developing precision intervention strategies.
Methods
PTB screening project conducted in Quzhou City
The retrospective study was based on a three-year population-based PTB ACF project conducted among elderly aged ≥65 years residing in six districts of Quzhou City, Zhejiang Province. County-level hospitals and primary medical facilities in Quzhou were responsible for organising and conducting screening examinations from March to October each year. The target population for the project included all individuals (both male and female) aged ≥65 years residing in Quzhou. Each participant completed a standard questionnaire and underwent physical examinations. A total of 96 township health centres were involved in the project (responsible for performing the CXR) along with 13 radiologists from county-level and city-level hospitals (responsible for reading the CXR). Each image was interpreted by two radiologists with extensive clinical experience from county-level hospitals using standardised criteria adapted from the National Health Industry Standard on Classification of Tuberculosis (WS196—2017) [16]. In case of any inconsistent CXR results, a third senior radiologist from a city-level hospital confirmed the findings. The final CXR results were categorised as: 1) normal; 2) abnormal, abnormalities suggestive of active PTB; 3) abnormal, abnormalities suggestive of either active or inactive TB; or 4) abnormal, non-TB pulmonary abnormalities [18, 22, 23].
All participants suspected of having active PTB because of radiographical abnormalities consistent with active TB or clinical symptoms (coughing, expectoration for more than 2 weeks, and haemoptysis or bloody sputum) at baseline and during follow-up from all study sites, were referred to a local designated hospital for PTB diagnosis. A sputum smear with acid-fast bacillus microscopy was conducted first and an additional GeneXpert MTB/RIF assay and/or smear culture was conducted to confirm the results if the smear was positive.
Study design and participants in current study
Based on the above ACF project, a retrospective cohort study was designed using the 2020 screening as the baseline survey and the subsequent screenings in 2021 or 2022 as follow-up surveys. The inclusion criteria were: 1) subjects who participated in the ACF project in 2020; and 2) subjects with complete demographics and radiological results. The exclusion criteria were: 1) subjects identified as having suspected active PTB during the baseline survey; 2) subjects diagnosed with active PTB before or within 90 days after the baseline survey.
The study outcomes were incident microbiologically confirmed PTB occurring among eligible participants during the 2-year follow-up. Apart from patients who were diagnosed during the follow-up survey, anyone who was registered in the Tuberculosis Information Management System (TBIMS), a national electronic system that systematically records the information of incident PTB cases, including diagnosis, treatment regimen and outcomes, was exported and matched with our study participants during the study period. Microbiologically confirmed PTB cases were defined as individuals with positive results for any of the three kinds of bacteriological evidence (sputum smear with acid-fast bacillus microscopy, smear culture or GeneXpert MTB/RIF assay) according to the National Health Industry Standard on Diagnosis for Pulmonary Tuberculosis (WS288–2017) [24].
The study protocol was approved by the ethics committee of the Institute of Pathogen Biology and the Chinese Academy of Medical Sciences, and written informed consent was obtained from all participants or their legal guardians if necessary (IPB-2024-12).
Results
Classification of CXR
As some participants had more than one type of lesion, the standard classification for pulmonary abnormalities in the analysis was as follows: participants with abnormal CXR consistent with active PTB were classified as “abnormalities suggestive of active PTB”, regardless of the presence of inactive PTB lesions or “abnormal inconsistent with PTB”; after ruling out suspected PTB cases, participants consistent with inactive PTB lesions and without active PTB progression observed within 90 days of screening were classified as “abnormalities suggestive of inactive PTB”; finally, those “abnormal inconsistent with PTB” were classified as “nontubercular pulmonary abnormalities” [18, 22, 23].
For participants showing radiographical abnormalities suggestive of inactive PTB, calcification lesions, fibrosis lesions, nodule lesions or pleural thickening lesions were further classified according to the National Health Industry Standard on Classification of Tuberculosis (WS196—2017) [16]. Detailed information on the four lesion types is described in supplementary table S1 [16, 18, 23, 24].
Statistical analysis
The data were analysed using R software (v.4.3.1). The baseline demographic characteristics and physical examinations are described with categorical data represented by frequency (percentage). Body mass index (BMI) was calculated for each participant and then categorised as underweight (<18.5 kg·m−2), normal weight (≥18.5 kg·m−2 to <24.0 kg·m−2), overweight (≥24.0 kg·m−2 to 28.0 kg·m−2) or obese (≥28.0 kg·m−2). Pearson chi-squared tests were used for categorical variables, with p<0.05 indicating statistical significance. Follow-up time was defined as the period from the baseline survey to the diagnosis of incident PTB, date of death or the end of the study period of 2022 screening for those who completed the 2-year follow-up, or the date of the latest survey for those who were lost to follow-up. The proportional hazards assumption was first accessed with the Grambsch–Therneau global test. Then, Cox proportion hazards regression models or an adjusted non-proportional Cox model were fitted to identify associated factors related to microbiologically confirmed active PTB development. Backward stepwise selection regression was utilised to assess the independence and the strength of the associations estimated by adjusted hazard ratio (aHR) and 95% CI for various potential risk factors. Sensitivity analysis was conducted by including both clinically diagnosed active PTB and microbiologically confirmed active PTB together as outcomes. Evidence for interaction of abnormalities suggestive of inactive PTB and with history of anti-TB treatment was assessed using the likelihood ratio test by comparing multivariable Cox regression with and without an interaction term.
Participant enrolment and characteristics
A total of 156 049 subjects participated in the 2020 baseline survey, with a response rate of 37.14% (156 049 of 420 138) of the entire resident population aged ≥65 years in Quzhou. After the exclusion of 1325 participants diagnosed with active PTB before or within 90 days after the baseline survey, and 696 participants with radiological results consistent with suspected PTB, 154 028 eligible participants were included in the final analysis, as shown in figure 1.
FIGURE 1.
Flowchart of inclusion, exclusion and follow-ups for study participants. CXR: chest radiography; PTB: pulmonary tuberculosis. #: microbiologically confirmed PTB.
The baseline characteristics of the study population classified by CXR results are outlined in table 1. Overall, 52.75% (81 249 of 154 028) were female and 70.59% (108 741 of 154 028) were under 75 years old. Individuals with a BMI <18.5 or ≥28 kg·m−2 accounted for 9.26% (13 276 of 154 028) and 6.12% (8773 of 154 028) of included subjects, respectively. According to the TBIMS, 1491 (0.97%, 1491 of 154 028) had an anti-TB treatment history; 7 (0.47%, 7 of 1491) cases were identified as drug resistance and 1174 (78.74%, 1174 of 1491) cases completed their anti-TB treatment course. After further classification by CXR results, CXR abnormalities were observed most in males, those aged ≥75 years old, individuals with low BMI and individuals with history of anti-TB treatment (p<0.001), as shown in table 1.
TABLE 1.
Baseline characteristics of the study participants classified by chest radiography (CXR) results
| Total (%)# n=154 028 | CXR | p-value | ||
|---|---|---|---|---|
| Normal n=126 267 | Abnormal n=27 761 | |||
| Sex | ||||
| Female | 81 249 (52.75) | 69 182 (54.79) | 12 067 (43.47) | <0.001 |
| Male | 72 779 (47.25) | 57 085 (45.21) | 15 694 (56.53) | |
| Age (years) | ||||
| ≥65–<70 | 60 985 (39.59) | 52 719 (41.75) | 8266 (29.78) | <0.001 |
| ≥70–<75 | 47 756 (31.01) | 39 508 (31.29) | 8248 (29.71) | |
| ≥75 | 45 287 (29.40) | 34 040 (26.91) | 11 247 (40.51) | |
| Illiteracy | ||||
| Yes | 50 357 (35.41) | 41 596 (35.69) | 8761 (34.14) | <0.001 |
| No | 91 867 (64.59) | 74 968 (64.31) | 16 899 (65.86) | |
| BMI, kg·m−2 | ||||
| <18.5 | 13 276 (9.26) | 8646 (7.35) | 4630 (17.91) | <0.001 |
| ≥18.5–<24 | 83 774 (58.44) | 68 625 (58.38) | 15 149 (58.72) | |
| ≥24–<28 | 37 532 (26.18) | 32 636 (27.76) | 4896 (18.98) | |
| ≥28 | 8773 (6.12) | 7649 (6.51) | 1124 (4.36) | |
| With history of anti-TB treatment | ||||
| Yes | 1491 (0.97) | 615 (0.49) | 876 (3.16) | |
| Complete anti-TB treatment course¶ | 1174 (0.76) | 474 (0.37) | 700 (2.53) | <0.001 |
| Not completed anti-TB treatment course | 38 (0.02) | 8 (0.01) | 30 (0.11) | |
| No | 152 537 (99.03) | 125 652 (99.51) | 26 885 (96.84) | |
| Residential district | ||||
| Changshan | 23 807 (15.44) | 20 449 (16.21) | 3358 (12.10) | <0.001 |
| Jiangshan | 55 639 (36.12) | 43 123 (34.15) | 12 516 (45.08) | |
| Kaihua | 27 798 (18.05) | 23 413 (18.54) | 4385 (15.80) | |
| Kecheng | 17 015 (11.05) | 14 221 (11.26) | 2794 (10.06) | |
| Longyou | 17 445 (11.33) | 14 415 (11.41) | 3030 (10.92) | |
| Qujiang | 12 324 (8.00) | 10 646 (8.43) | 1678 (6.04) | |
Data are presented as n (%). BMI: body mass index; TB: tuberculosis. #: data might not sum to the total because of missing data. ¶: among them, seven received drug-resistant treatment, including two with normal CXR and five with abnormal CXR, showing no statistically significant difference (p=0.71). p-values were obtained by chi-squared test.
Radiological results classified by history of anti-TB treatment
The radiological results showed 81.98% (126 267 of 154 028) of the study population had normal CXR results and 18.03% (27 761 of 154 028) exhibited abnormal results, including 9.76% (15 037 of 154 028) with abnormalities suggestive of inactive PTB, and 8.26% (12 724of 154 028) with nontubercular pulmonary abnormalities. Of the 15 037 individuals with inactive PTB lesions, 13 084 had detailed radiological descriptions and were further classified into five types: with only fibrosis lesions (31.08%, 4066 of 13 084), with only calcification lesions (10.77%, 1409 of 13 084), with only pleural thickening lesions (19.57%, 2561 of 13 084), with only nodule lesions (30.16%, 3946 of 13 084) and with at least two types of lesion (8.42%, 1102 of 13 084) (table 2).
TABLE 2.
The radiological results of the study participants classified by history of anti-tuberculosis (TB) treatment
| Classification | Total n=154 028 | With history of anti-TB treatment n=1491 | Without history of anti-TB treatment n=152 537 | p-value |
|---|---|---|---|---|
| CXR classification | ||||
| Normal CXR | 126 267 (81.98) | 615 (41.25) | 125 652 (82.37) | <0.001 |
| Abnormalities suggestive of inactive PTB | 15 037 (9.76) | 612 (41.05) | 14 425 (9.46) | |
| Nontubercular pulmonary abnormalities# | 12 724 (8.26) | 264 (17.71) | 12 460 (8.17) | |
| The types of inactive PTB lesions | ||||
| Only with fibrosis lesions | 4066 (31.08) | 105 (25.43) | 3961 (31.26) | <0.001 |
| Only with calcification lesions | 1409 (10.77) | 16 (3.87) | 1393 (10.99) | |
| Only with pleural thickening lesions | 2561 (19.57) | 108 (26.15) | 2453 (19.36) | |
| Only with nodule lesions | 3946 (30.16) | 117 (28.33) | 3829 (30.22) | |
| With two or more lesions | 1102 (8.42) | 67 (16.22) | 1035 (8.17) | |
Data are presented as n (%). CXR: chest radiography; PTB: pulmonary tuberculosis. #: the majority of non-TB lesions were lung hyperinflation (2493, 19.57%), nodule (2290, 17.99%) and airway widening (638, 5.01%). p-values were obtained by chi-squared test.
The distribution of radiological classifications among individuals with a history of anti-TB treatment varied significantly. Approximately 40% (612 of 1491) presented with abnormalities suggestive of inactive PTB. Compared with individuals without a history of anti-TB treatment, those who have undergone anti-TB treatment were less likely to present with fibrosis lesions, calcification lesions and nodule lesions but more likely to exhibit pleural thickening lesions and multiple types of lesions (p<0.001).
Association analysis with respect to microbiologically confirmed active PTB development during 2-year follow-up
The follow-up surveys were completed by 103 854 participants in 2021 and 108 431 participants in 2022 covering 85.22% (131 270 of 154 028) of baseline participants. Females and individuals aged < 75 years old were more likely to participate in the follow-up (supplementary table S2). During the 2-year follow-up, a total of 462 cases developed microbiologically confirmed active PTB with an overall incidence rate of 0.19 per 100 person-years. Details of the incidence rate of microbiologically confirmed PTB among different subgroups are provided in table 3. In the univariate analysis, sex, age, illiteracy conditions, BMI, history of anti-TB treatment, residential districts and radiological classifications were found to be significantly related to microbiologically confirmed PTB incidences (p<0.05). In the multivariable analysis, sex, BMI, history of anti-TB treatment, residential districts and abnormal radiological results were independently associated with the risk of active PTB. Radiological abnormalities suggestive of inactive PTB were observed to be a potential independent risk factor for microbiologically confirmed PTB development with an aHR of 6.00 (95% CI 4.85–7.43) compared with those with normal CXR. The incidence rate of active PTB among participants with inactive PTB lesions was 0.73 per 100 person-years, whereas that among individuals with a normal CXR was 0.11 per 100 person-years.
TABLE 3.
Association analysis with respect to microbiologically confirmed active pulmonary tuberculosis (PTB) development
| Variables | Incidence rate per 100 person-years | p-value | Adjusted HR (95% CI)# |
|---|---|---|---|
| Total | 462/254 337 (0.19) | ||
| Sex | |||
| Female | 145/135 792 (0.11) | <0.001 | Ref. |
| Male | 317/118 546 (0.27) | 1.94 (1.59–2.37) | |
| Age, years | |||
| ≥65–<70 | 158/102 691 (0.16) | 0.006 | |
| ≥70–<75 | 138/81 974 (0.17) | ||
| ≥75 | 166/69 672 (0.24) | ||
| Illiteracy | |||
| Yes | 131/84 421 (0.16) | 0.01 | |
| No | 307/154 255 (0.20) | ||
| BMI, kg·m−2 | |||
| <18.5 | 106/21 570 (0.50) | <0.001 | 1.98 (1.57–2.49) |
| ≥18.5–<24 | 259/140 407 (0.19) | Ref. | |
| ≥24–<28 | 67/63 595 (0.11) | 0.68 (0.52–0.89) | |
| ≥28 | 12/14 767 (0.08) | 0.57 (0.32–1.02) | |
| With history of anti-TB treatment | |||
| Yes | 27/2515 (1.10) | <0.001 | 2.28 (1.53–3.38) |
| No | 435/140 407 (0.18) | Ref. | |
| Multidrug resistance | |||
| Yes | 0 | ||
| No | 27/2515 (2.45) | 1.00 | |
| Complete anti-TB treatment course | |||
| Yes | 21/2957 (1.07) | 0.05 | |
| No | 3/246 112 (4.35) | ||
| Residential district | |||
| Changshan | 113/42 104 (0.28) | <0.001 | 2.03 (1.32–3.12) |
| Jiangshan | 165/102 793 (0.16) | 0.81 (0.54–1.24) | |
| Kaihua | 72/45 472 (0.16) | 1.25 (0.80–1.97) | |
| Kecheng | 58/24 446 (0.24) | 2.18 (1.36–3.47) | |
| Longyou | 26/20 681 (0.13) | Ref. | |
| Qujiang | 28/18 841 (0.15) | 1.11 (0.65–1.90) | |
| CXR classification | |||
| Normal | 215/209 099 (0.11) | <0.001 | Ref. |
| Abnormalities suggestive of inactive PTB | 179/25 092 (0.73) | 6.00 (4.85–7.43) | |
| Nontubercular pulmonary abnormalities | 68/20 147 (0.35) | 2.45 (1.85–3.23) | |
BMI: body mass index; CXR: chest radiography; HR: hazard ratio; TB: tuberculosis. #: all variables were included in Cox proportion hazards regression models and selected using the backward stepwise selection method. ¶: data might not sum to the total because of missing data. p-values were obtained by chi-squared test.
The results were not changed when we included both clinically diagnosed active PTB and microbiologically confirmed active PTB in the analysis (supplementary table S3).
The relationship between different inactive PTB lesions and microbiologically confirmed active PTB development
As shown in figure 2, all types of inactive PTB lesions showed a significantly increased microbiologically confirmed PTB incidence (0.31 to 0.71 per 100 person-years) when compared with normal CXR (0.11 per 100 person-years) as a reference. In the multivariable analysis, fibrosis, calcification, pleural thickening and nodule lesions were significantly associated with an increased risk of microbiologically confirmed active PTB after adjusting for the risk factors of sex, age, illiteracy conditions, BMI, history of anti-TB treatment and residential districts. Compared with those with a normal CXR, the risk of developing microbiologically confirmed active PTB ranged from 2.94 (95% CI 1.44–5.98) for those with calcification lesions, to 6.55 (95% CI 4.66–9.21) for those only with nodule lesions.
FIGURE 2.
The relationship between different inactive pulmonary tuberculosis (PTB) lesions and the development of microbiologically confirmed active PTB. The risk of different types of inactive PTB lesions progressing to active PTB is shown. Adjusted hazard ratio (aHR) for PTB development in groups with various types of inactive PTB lesions, compared with those with normal chest radiography (CXR) are shown along with 95% confidence intervals. aHRs were analysed by the Cox proportion hazards regression models after adjusting for sex, age, illiteracy conditions, body mass index, history of anti-tuberculosis treatment and residential districts. #: incidence rate per 100 person-years.
Subgroup analysis on inactive PTB lesions and the risk of microbiologically confirmed active PTB
A stratified analysis of the association between inactive PTB lesions and the risk of microbiologically confirmed active PTB is shown in figure 3. As for the participants with inactive PTB lesions among different sex, age groups, illiteracy conditions, BMI categories and residential districts, the incidence rates of microbiologically confirmed active PTB ranged from 0.38 to 1.47 per 100 person-years, which was higher than that of individuals with a normal CXR with incidence rates ranging from 0.06 to 0.27 per 100 person-years.
FIGURE 3.
Stratified analysis on the association of inactive pulmonary tuberculosis (PTB) lesions with the risk of active microbiologically confirmed PTB. Subgroup analyses of the effect of inactive PTB lesions on PTB development are shown. Adjusted hazard ratio (aHR) for PTB development in groups with inactive PTB lesions compared with those with normal chest radiography (CXR) are shown along with 95% confidence intervals. aHRs were analysed by the Cox proportion hazards regression models after adjusting for sex, age, illiteracy conditions, body mass index, history of anti-tuberculosis (TB) treatment and residential districts. #: incidence rate per 100 person-years.
Multivariable analysis consistently demonstrated the significant effects of inactive PTB lesions on microbiologically confirmed active PTB development across various subgroups. However, for individuals with and without a history of anti-TB treatment, the influence of inactive PTB lesions on PTB recurrence was different with an aHR of 1.08 (95% CI 0.49–2.38) and 6.42 (95% CI 5.17–7.99), respectively. A further classified analysis was conducted to explore the potential interaction between the history of anti-TB treatment and inactive PTB lesions, as shown in table 4. In the multivariable analysis, inactive PTB lesions alone or a history of anti-TB treatment alone were independently associated with the risk of active PTB with aHRs of 6.96 (95% CI 5.59–8.67) and 7.67 (95% CI 4.26–13.78), respectively. Additionally, a combination effect of a history of anti-TB treatment and inactive PTB lesions on the potential factors on the risk of developing active PTB were found (p-value for interaction <0.001), with an elevated aHR of 10.50 (95% CI 5.93–18.52).
TABLE 4.
Combined effect of inactive pulmonary tuberculosis (PTB) lesions and history of anti-tuberculosis treatment on the development of active PTB
| Abnormalities suggestive of inactive PTB | With history of anti-TB treatment | n/N (%) | Incidence rate per 100 person-year | Adjusted HR (95% CI)# | p-value for interaction |
|---|---|---|---|---|---|
| No | No | 203/125 652 (0.16) | 0.10 | Ref. | <0.001 |
| Yes | No | 166/14 428 (1.15) | 0.71 | 6.96 (5.59–8.67) | |
| No | Yes | 12/615 (1.95) | 1.17 | 7.67 (4.26–13.78) | |
| Yes | Yes | 13/612 (2.12) | 1.29 | 10.50 (5.93–18.52) |
HR: hazard ratio. #: Cox proportion hazards regression models were used and adjusted by sex, age, illiteracy conditions, body mass index and residential district.
Discussion
Through retrospectively tracking the incidence of PTB among 154 028 elderly people based on an ACF project, we found the presence of inactive PTB lesions was independently associated with a six-fold increased risk of active PTB development. The risk remained constant regardless of the type of inactive PTB lesions. In addition, a history of anti-TB treatment and inactive PTB lesions showed a significant interaction; a combined impact on developing active PTB was observed with an aHR of 10.50 (95% CI 5.93–18.52). The results indicated that individuals with inactive PTB lesions require attention for TB control regardless of whether there is a history of anti-TB treatment or not.
Inactive PTB lesions were found to be independently associated with the risk of development of active PTB in our study. This finding was consistent with a recent meta-analysis that showed that during 2–4-year follow-ups, progression for participants with baseline radiographical results of inactive PTB to active PTB occurred at an annualised rate of 1% (95% CI 0.30–1.80) among 24 cohort studies, whereas progression for participants with baseline radiographical results of normal to positive disease occurred at an annualised rate of 0.1% (95% CI 0.10–0.20) among four cohorts [4]. Moreover, our findings revealed that inactive PTB lesions and a history of anti-TB treatment were independently associated with the development of active PTB. For individuals with inactive PTB lesions, the risk of developing active PTB is approximately seven-times higher, even without a past anti-TB treatment history, compared with those with a normal CXR and without a history of anti-TB treatment. Furthermore, the combination effect of a history of anti-TB treatment and inactive PTB lesions indicated the essential strengthening of active PTB monitoring among those high-risk subgroups.
The large sample size of our study not only further strengthened the evidence of the positive relation between inactive PTB lesions and PTB development but also allowed us to specifically explore the impact of different types of inactive lesions in greater detail. Previous studies mainly focused on fibrosis and calcification lesions. The statement of the American Thoracic Society and the Centers for Disease Control and Prevention demonstrated a markedly elevated risk of active PTB development in individuals with fibrosis lesions and past TB history, with risks ranging from 2- to 14-times higher than those with a normal CXR [11]. Similarly, in Japan, males aged over 40 years with fibrosis lesions have a 16-fold increase in TB incidence compared with counterparts with a normal CXR, whereas females in the same age group exhibit a notable 24-fold heightened risk [25]. Canadian guidelines reported a 6–19-times increased risk of developing active PTB for those with fibrosis lesions compared with those with a normal CXR [12, 26]. Although individuals with calcification lesions had lower risks of developing active PTB compared with those with fibrosis lesions, the incidences of latent TB infection among these two groups were approximately similar when the area of the calcification lesion exceeded 1.5 cm2, leading to an increased risk of active PTB [20, 21]. The evolution of inactive PTB lesions, progressing from peripheral fibrosis to calcification stages, has been demonstrated in animal models, and these findings confirm the infectious potential of inactive PTB lesions, with active MTB detected in both fibrosis lesions and calcification lesions with 1×104 CFU·g−1 and 10 CFU·g−1, respectively, further emphasising the complexities of PTB occurrence risks associated with different lesion types [27, 28]. In our study, apart from fibrosis and calcification lesions, pleural thickening and nodule lesions also contributed to a significantly increased active PTB risk.
Although research on pleural thickening and nodular lesions as inactive PTB lesions contributing to the development of active PTB remains limited, it is noteworthy that both the Chinese national standard [17] and previous studies [18, 20] classify these lesions as inactive. Our study using a large sample size that included comprehensive data on these two types of lesions provided additional evidence for expanding the types of inactive PTB lesions, understanding associated risks and uncovering potential mechanisms of active PTB development. Notably, we consistently found that different inactive lesions were associated with three-fold to seven-fold risks of active PTB development. These increasing risks are comparable with that of TB contacts (aHR 8.0, 95% CI 4.8–13.4), illicit drug users (aHR 6.0, 95% CI 2.4–13.4), homeless (aHR 7.27, 95% CI 0.51–103.70) and health workers (aHR 2.4, 95% CI 1.2–3.6) who are targeted as populations for TPT in the WHO guidelines [3, 29]. The WHO consolidated guidelines on TB preventive treatment also cited our previous findings and emphasised that individuals, including those with fibrosis lesions, may be at increased risk of TB progression [5, 14]. This study further supported our previous findings, expanded the PTB risk lesion types and demonstrated the necessity of taking into consideration the inclusion of the key population with inactive PTB lesions as the target for intervention strategies. Although all types of inactive lesions were related to increased PTB incidence, this did not imply that all individuals with inactive PTB lesions who lacked a history of anti-TB treatment should be prioritised as a key population for TPT. The criteria for identifying the key population for TPT must consider not only the risk of developing active PTB but also the sensitivity to TPT. Currently, guidelines from the United States and Canada recommend TPT predominantly for individuals with fibrosis lesions [11, 12]. Our previous research also supported the finding that populations with fibrosis lesions showed the most favourable responses to TPT [10]. In the future, additional RCTs and mechanistic research are required to gather evidence regarding protective efficacy among cases with various inactive PTB lesions, including calcification lesions, pleural thickening lesions and nodule lesions.
A previous study conducted in China revealed that the overall cost of PTB screening by CXR among the those aged ≥65 years with an incidence rate of 0.50 cases per 100 person-years was US$1881 per PTB case and that the cost decreased with the increase in TB development risk [30]. In our study, the incidence rate among those aged ≥65 years with inactive PTB lesions was 0.73 cases per 100 person-years, and the incidence rate among those with both inactive PTB lesions and a history of anti-TB treatment was as high as 1.20 cases per 100 person-years. Therefore, the cost of PTB screening by CXR among those aged ≥65 years with inactive lesions might be cost savings from the previous report in China. In addition, our findings highlight the importance of developing a risk assessment framework based on the risk factors identified and refining the corresponding control measures. For individuals with suspected TB lesions, early diagnosis and treatment should be the priority. For individuals with inactive TB lesions and no history of anti-TB treatment, latent TB infection testing and TPT are recommended. For individuals with a history of anti-TB treatment, active PTB surveillance is an important approach for the control of recurrent diseases. By integrating TPT with active-PTB case-finding, more cost-effective intervention models may be established for this key population.
When interpreting the results of the present study, several limitations should be kept in mind. First, the study population did not undergo MTB infection testing at baseline. Consequently, it was challenging to definitively establish whether these inactive PTB lesions were reasonably linked with MTB infection or represented early-stage active PTB [31]. However, previous research conducted by our team disclosed that individuals over 50 years old with inactive PTB lesions in China exhibited a substantial latent TB infection rate of up to 50% tested by the interferon-gamma release assay (IGRA) [5]. In high TB-burden countries, studies have shown high positivity rates for MTB infections in patients with inactive PTB lesions. In South Korea, positivity rates of latent TB infection among cases with inactive PTB lesions were reported at 54.6% and 77.7% for the tuberculin skin test and IGRA, respectively, and among individuals with positive IGRA results, inactive PTB lesions were discovered in 61.9% of cases [31]. Similarly, a study in India revealed that the positivity rate of latent TB infection exceeded 60% in a population with inactive PTB lesions [18]. The research indicated that individuals with inactive PTB lesions are highly likely to have been infected with MTB. Second, we did not measure the size of fibrosis and calcified lesions, and could not further evaluate the relationship between lesion size and the occurrence of PTB [21]. Third, selection bias could not be fully excluded. Although our study includes a large population aged ≥65 years, it only represents approximately 37% of the entire population in this age group, and cannot perfectly represent the local elderly population. Fourth, although CXR results were interpreted by a panel of radiologists, the inherent subjectivity of human readers may introduce potential misclassification bias. Furthermore, the prevalence of inactive PTB lesions, along with their associated risk factors, may differ across various populations. Therefore, these findings should not be directly extrapolated to other populations or settings. Fifth, the current study was a retrospective report based on real-world data; parallel CXR reporting by two radiologists was not implemented at the site and only one final report was recorded, which made it impossible to assess radiologists' agreement rates. Thus, it limited us to further evaluate the capacity of radiologists from primary health institutes. Sixth, the incidence rate may be underestimated in consideration of misdiagnosis or delayed diagnosis among individuals who should have been referred during the ACF project. In addition, while several demographic factors were controlled in the multivariable analysis, other potential risk factors influencing PTB development, such as comorbidities and socioeconomic status, were not included, which may introduce confounding bias.
In conclusion, individuals aged ≥65 years with inactive PTB lesions, regardless of the lesion types and history of anti-TB treatment, are at increased risk of developing active PTB and require further intervention for TB control.
Acknowledgements
We thank the staff in Quzhou Centers for Disease Prevention and Control (CDC) and sentinel hospitals, who collected demographic data and performed physical examinations. We also thank the participants for their cooperation.
Footnotes
Provenance: Submitted article, peer reviewed.
Author contributions: The study was conceived by L. Gao, F. Wang and P. Zhu. Data collection was led by Y. Zhang, X. Cao, J. Du, Y. Di, J. Huang, Y. He, B. Feng, Z. Li, J. Liang, W. Wang, Y. Peng, X. Hao, C. Fang, B. Xu, X. Wang, B. Chen and Z. Wang. Data cleaning and verification was performed by H. Xin, Y. Di and L. Shen, and data analysis and interpretation by L. Shen, H. Xin and L. Gao wrote the first draft of the manuscript. L. Gao and H. Xin revised the manuscript. All authors contributed to and approved the final manuscript.
Ethics statement: The ethics committees of the Institute of Pathogen Biology and the Chinese Academy of Medical Sciences approved the current study (IPB-2024-12).
Conflict of interest: None declared.
Support statement: This work was supported by the CAMS Innovation Fund for Medical Sciences (2021-I2M-1–0370) and the National Natural Science Foundation of China (82373647). The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication. Funding information for this article has been deposited with the Open Funder Registry.
Supplementary material
Please note: supplementary material is not edited by the Editorial Office, and is uploaded as it has been supplied by the author.
Supplementary material
01231-2024.SUPPLEMENT
Data availability
The anonymised datasets used in this study can be made available upon reasonable request to the corresponding author.
References
- 1.World Health Organization . Global Tuberculosis Report 2023. Geneva, World Health Organization, 2023. [Google Scholar]
- 2.Bagcchi S. WHO's Global Tuberculosis Report 2022. Lancet Microbe 2023; 4: e20. doi: 10.1016/S2666-5247(22)00359-7 [DOI] [PubMed] [Google Scholar]
- 3.World Health Organization . Appendix to the Guidelines on the Management of Latent Tuberculosis Infection. Geneva, World Health Organization; 2015. [PubMed] [Google Scholar]
- 4.Sossen B, Richards AS, Heinsohn T, et al. The natural history of untreated pulmonary tuberculosis in adults: a systematic review and meta-analysis. Lancet Respir Med 2023; 11: 367–379. doi: 10.1016/S2213-2600(23)00097-8 [DOI] [PubMed] [Google Scholar]
- 5.Gao L, Li X, Liu J, et al. Incidence of active tuberculosis in individuals with latent tuberculosis infection in rural China: follow-up results of a population-based, multicentre, prospective cohort study. Lancet Infect Dis 2017; 17: 1053–1061. doi: 10.1016/S1473-3099(17)30402-4 [DOI] [PubMed] [Google Scholar]
- 6.Richards AS, Sossen B, Emery JC, et al. Quantifying progression and regression across the spectrum of pulmonary tuberculosis: a data synthesis study. Lancet Glob Health 2023; 11: e684–e692. doi: 10.1016/S2214-109X(23)00082-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Liu K, Ai L, Pan J, et al. Survival analysis and associated factors for pulmonary tuberculosis death: evidence from the information system of tuberculosis disease and mortality surveillance in China. Risk Manag Healthc Policy 2022; 15: 1167–1178. doi: 10.2147/RMHP.S368593 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Stagg HR, Zenner D, Harris RJ, et al. Treatment of latent tuberculosis infection: a network meta-analysis. Ann Intern Med 2014; 161: 419–428. doi: 10.7326/M14-1019 [DOI] [PubMed] [Google Scholar]
- 9.Zenner D, Beer N, Harris RJ, et al. Treatment of latent tuberculosis infection: an updated network meta-analysis. Ann Intern Med 2017; 167: 248–255. doi: 10.7326/M17-0609 [DOI] [PubMed] [Google Scholar]
- 10.Zhang H, Xin H, Du Y, et al. Tuberculosis preventive treatment among individuals with inactive tuberculosis suggested by untreated radiographic abnormalities: a community-based randomized controlled trial. Emerg Microbes Infect 2023; 12: e2169195. doi: 10.1080/22221751.2023.2169195 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Targeted tuberculin testing and treatment of latent tuberculosis infection. This official statement of the American Thoracic Society was adopted by the ATS Board of Directors, July 1999. This is a Joint Statement of the American Thoracic Society (ATS) and the Centers for Disease Control and Prevention (CDC). This statement was endorsed by the Council of the Infectious Diseases Society of America. (IDSA), September 1999, and the sections of this statement. Am J Respir Crit Care Med 2000. 161: S221–S247. doi: 10.1164/ajrccm.161.supplement_3.ats600 [DOI] [PubMed] [Google Scholar]
- 12.Public Health Agency of Canada . Canadian Tuberculosis Standards 7th edition. Ottawa, Public Health Agency of Canada, 2014. [Google Scholar]
- 13.Xu C, Zhao Y. Chinese Guidelines for Preventive Treatment of Tuberculosis 2023. Beijing: People's Medical Publishing House. [Google Scholar]
- 14.World Health Organization . WHO Consolidated Guidelines on Tuberculosis: Tuberculosis Preventive Treatment 2020. Geneva, World Health Organization, 2020. [Google Scholar]
- 15.Hartman-Adams H, Clark K, Juckett G. Update on latent tuberculosis infection. Am Fam Physician 2014; 89: 889–896. [PubMed] [Google Scholar]
- 16.National Health and Family Planning Commission of the People's Republic of China . Classification of tuberculosis (WS196-2017). Date last accessed: Dec 2018. Date last updated: 2017.. https://www.nhc.gov.cn/ewebeditor/uploadfile/2017/11/20171128164208411.pdf
- 17.Chinese Antituberculosis Association . Expert consensus on diagnosis and prevention of inactive pulmonary tuberculosis. Tubercle Lung Dis 2021; 2: 197–201. [Google Scholar]
- 18.Joshi R, Patil S, Kalantri S, et al. Prevalence of abnormal radiological findings in health care workers with latent tuberculosis infection and correlations with T cell immune response. PLoS ONE 2007; 2: e805. doi: 10.1371/journal.pone.0000805 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.International Labor Office . Guidelines for the use of the ILO International Classification of Radiographs of Pneumoconiosis. Geneva: International Labor Office. Occupational Safety and Health Series; 2002. [Google Scholar]
- 20.Uzorka JW, Kroft L, Bakker JA, et al. Abnormalities suggestive of latent tuberculosis infection on chest radiography; how specific are they. J Clin Tuberc Other Mycobact Dis 2019; 15: 100089. doi: 10.1016/j.jctube.2019.01.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Uzorka JW, Wallinga J, Kroft L, et al. Radiological signs of latent tuberculosis on chest radiography: a systematic review and meta-analysis. Open Forum Infect Dis 2019; 6: ofz313. doi: 10.1093/ofid/ofz313 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.World Health Organization . Chest Radiography in Tuberculosis Detection: Summary of Current WHO Recommendations and Guidance on Programmatic Approaches. World Health Organization; 2016. [Google Scholar]
- 23.World Health Organization . Systematic Screening for Active Tuberculosis: Principles and Recommendations. Geneva, World Health Organization, 2013. [Google Scholar]
- 24.World Health Organization . The WHO Manual of Diagnostic Imaging Radiographic Anatomy and Interpretation of the Chest and the Pulmonary System. Geneva, World Health Organization, 2013. [Google Scholar]
- 25.Shimouchi A, Ozasa K. The incidence rate of active pulmonary tuberculosis among adult population with fibrotic lesions. Kekkaku 2003; 78: 5–13. [PubMed] [Google Scholar]
- 26.Gale-Rowe M, Menzies D, Sutherland J, et al. Highlights of the new 7th edition of the Canadian Tuberculosis Standards. Can Commun Dis Rep 2014; 40: 113–116. doi: 10.14745/ccdr.v40i06a04 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Lin PL, Rodgers M, Smith L, et al. Quantitative comparison of active and latent tuberculosis in the cynomolgus macaque model. Infect Immun 2009; 77: 4631–4642. doi: 10.1128/IAI.00592-09 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Lin PL, Dartois V, Johnston PJ, et al. Metronidazole prevents reactivation of latent Mycobacterium tuberculosis infection in macaques. Proc Natl Acad Sci USA 2012; 109: 14188–14193. doi: 10.1073/pnas.1121497109 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.World Health Organization . WHO Operational Handbook on Tuberculosis. Module 1: Prevention Tuberculosis Preventive Treatment. Geneva, World Health Organization, 2024. [Google Scholar]
- 30.Zhang C, Ruan Y, Cheng J, et al. Comparing yield and relative costs of WHO TB screening algorithms in selected risk groups among people aged 65 years and over in China, 2013. PLoS ONE 2017; 12: e0176581. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Jeong YJ, Yoon S, Koo HK, et al. Positive tuberculin skin test or interferon-gamma release assay in patients with radiographic lesion suggesting old healed tuberculosis. J Korean Med Sci 2012; 27: 761–766. doi: 10.3346/jkms.2012.27.7.761 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Please note: supplementary material is not edited by the Editorial Office, and is uploaded as it has been supplied by the author.
Supplementary material
01231-2024.SUPPLEMENT
Data Availability Statement
The anonymised datasets used in this study can be made available upon reasonable request to the corresponding author.