The prevalence and pattern of post tuberculosis lung disease including pulmonary hypertension from an Australian TB service; a single-centre, retrospective cohort study

Abstract

Introduction

Post Tuberculosis Lung Disease (PTLD) is increasingly recognised as a significant cause of morbidity internationally, but has not been described in an Australian setting. We aimed to determine the prevalence of PTLD among adult TB survivors from an Australian TB service and describe the pattens of lung function abnormalities and pulmonary disease, including pulmonary hypertension.

Methods

We conducted a single-centre retrospective cohort study in Sydney, Australia, including all adults who successfully completed TB treatment between January 2013 and December 2022. Baseline characteristics, post treatment pulmonary function, and thoracic computed tomography (CT) data were analysed to determine the prevalence and patterns of PTLD, defined as any lung function and/or radiological abnormality attributable to TB.

Results

Among 119 confirmed TB patients (mean age 46 ± 21 years, 61% males) PTLD was identified in 81/119 (68%). Pulmonary function testing was available for 51/119 (43%), of whom 38/51(75%) exhibited abnormalities. Obstructive deficits were found in 25/51 (49%), restrictive deficits in 11/51 (22%), and impaired gas transfer capacity in 26/51 (51%). Chest CT scans were completed in 76/119 (64%), with 70/76 (92%) demonstrating significant abnormalities, including pulmonary fibrosis 43/76 (57%), bronchiectasis 22/76 (29%), and emphysema 11/76 (15%). Pulmonary hypertension was suspected in 52/76 (68%) patients based on radiological findings.

Conclusion

Despite successful treatment, PTLD was frequently observed among our cohort of Australian TB survivors. Further research into optimal screening practices to diagnose chronic pulmonary diseases and pulmonary hypertension could provide opportunities for earlier intervention and management.

Summary at a glance.

Post-tuberculosis lung disease (PTLD) is an under recognised sequelae of tuberculosis, with unknown prevalence in Australia. We identified PTLD in 68% of patients who successfully completed TB treatment at an Australian clinic. The high prevalence of PTLD supports the need for routine screening after treatment.

Introduction

Tuberculosis (TB) remains a significant global health problem, with 10 million people developing the disease in 2023 and 1.5 million of these dying [1]. Antimicrobial treatment has dramatically reduced mortality from acute TB disease, yet chronic sequelae and its associated morbidity affecting TB survivors remains [2, 3]. Historically, the primary aim of TB treatment according to accepted international treatment outcome guidelines, has been a “successful treatment” outcome defined by microbiological cure [4]. Emerging evidence however, suggests that chronic respiratory impairment persists following “successful” TB treatment, challenging this paradigm [5, 6]. Studies demonstrate that (successfully treated) TB survivors experience significantly elevated all-cause mortality compared to the general population [7], persistent pulmonary function abnormalities, and significant symptom burden after the initial TB disease episode [8, 9].

The accumulating evidence of persistent long-term respiratory impairment following TB disease has led to the formal acknowledgement of Post Tuberculosis Lung Disease (PTLD). This chronic condition was defined at the first international symposium of Post Tuberculosis Lung Health (Stellenbosch, South Africa 2019) by Delphi consensus as “Evidence of chronic respiratory abnormality, with or without symptoms, attributable at least in part to previous tuberculosis” [10]. Reported rates of respiratory impairment following TB disease (largely based on spirometry) varies widely (between 18% and 87%) [6], reflecting the heterogeneity of study populations and the different criteria used for defining respiratory impairment. This highlights the challenges of how to define and characterise PTLD [11].

The most common PTLD pattern or phenotype, reported in the literature is airflow obstruction via spirometry. Previous studies have demonstrated a significant association between TB and the development of chronic obstructive pulmonary disease (COPD), even after adjusting for smoking history [9, 12]. Other PTLD phenotypes including pleural, parenchymal, and pulmonary vascular disease (including pulmonary hypertension) have been reported, although fewer studies have utilised plethysmography or other complex lung function assessments to better characterise these patterns of disease [2].

The prevalence and pattern of PTLD in Australia is not known. Some may think that early diagnosis of TB disease in Australia may result in less PTLD as our healthcare system conducts routine TB screening of migrants from TB endemic countries (with most cases of active TB disease in Australia occurring in migrants). This provides free access to diagnostic testing, and treatment of TB disease in Australia. However, as with most countries, the Australian model of care for TB remains focused on microbiological cure and there is no routine, programatic screening for persistent imaging or lung function abnormalities after TB treatment, although this has recently been recommended [13]. More information on the potential burden of PTLD disease in Australia is urgently needed and could also inform other high income countries with low TB burden and similar health systems.

At St Vincent’s hospital Sydney, all patients with a history of TB are routinely offered follow-up with the TB service after they complete TB treatment. Whilst not protocolled, this is often at 3 months, 6 months, 12 months and then annually for as long as deemed clinically appropriate. Investigations are individualised depending on patient symptoms and physician preference but typically include chest X-rays and lung function testing. Our primary aim of this study was to determine the prevalence of PTLD in patients with successful TB treatment outcomes (defined microbiologically per standard criteria) for Mycobacterium tuberculosis in an Australian setting. Our secondary objectives were to describe (i) the patterns of radiological abnormalities, (ii) pulmonary function abnormalities, and (iii) the prevalence of pulmonary hypertension in this cohort of treated TB patients.

Methods

This was a single-centre cohort study of all patients with a successful TB treatment outcome (treatment complete or cure) at St Vincent’s Hospital, Sydney, Australia between 1st January 2013 and 31st December 2022. Ethics approval was obtained from the St Vincent’s Hospital Human Research Ethics committee (2020/ETH02889). As with other TB services in New South Wales, St Vincent’s has a successful TB treatment completion rate of over 90% and a re-infection/relapse rate of under 1%, thus almost all included patients had a single episode of TB disease. St Vincent’s hospital only treats adult patients, thus all patients were aged over 18 years at the time of their initial TB diagnosis. We retrospectively reviewed the electronic medical records of all eligible patients. Patient characteristic data including age, gender, and co-morbidities were collected. Details of patients’s TB episode including time of diagnosis, sputum acid fast bacilli smear microscopy, mycobacterial culture positivity, extra-pulmonary involvement, and drug resistance were recorded. Data on the TB treatment course including total treatment duration and time to microbiological cure (if available/relevant) were also collected.

Consistent with the consensus definition from the first International symposium of Post Tuberculosis Lung Health, PTLD was defined as any pulmonary function or radiological abnormality in patients (independent of symptoms) that could be attributable to prior TB disease [10]. We reviewed complex pulmonary function tests (PFTs) conducted following TB treatment that were requested by the treating physician and completed in the lung function laboratory at St Vincent’s Hospital. All PFTs were performed in accordance with the standards of the American Thoracic Society/European Respiratory Society (ATS/ERS) to ensure quality control and reproducibility of results and post-bronchodilator spirometry (400mcg inhaled salbutamol) was also performed [14]. Obstructive lung function defects were defined as an FEV1/FVC ratio below the predicted lower limit of normal for the reference population. Restrictive defects were defined as total lung capacity (TLC) below the predicted lower limits of normal. A gas transfer capacity abnormality was defined as the measured DLCO below the predicted lower limit of normal, adjusted for haemoglobin. The reference used to generate predicted values have been published previously [15].

Radiological outcomes were evaluated and included radiologist reported respiratory disease or significant pulmonary abnormality, and markers of computed tomography (CT) defined suspected pulmonary hypertension. Thoracic CT scans performed after the completion of TB treatment were reviewed by the reporting radiologist/s (RR) and a study investigator (XH). We specifically noted radiological features of chronic structural lung damage including emphysema, bronchiectasis, and pulmonary fibrosis.

We assessed pulmonary hypertension directly from right heart catheterisation (RHC) if available, with pulmonary hypertension defined as mean pulmonary artery pressure (mPAP) > 20 mmHg [16]. Indirect evidence of pulmonary hypertension was obtained from trans-thoracic echocardiogram (TTE) and CT imaging. CT images were reviewed for pulmonary artery diameter (PAD) and pulmonary artery: aorta (PA: A) ratio measurements. These features of pulmonary hypertension have previously demonstrated good sensitivity and specificity for detecting pulmonary hypertension indirectly [13]. We used PAD > 29 mm and PA: A ratio > 1 as thresholds for suspected pulmonary hypertension.

Statistical analysis

All continuous data were presented as mean ± standard deviation (SD) or median with interquartile range (IQR). Categorical data were reported as number and/or percentage. We compared baseline and TB disease characteristics between patients with and without PTLD. Continuous variables were compared using the Student’s T-test for normally distributed data or Mann-Whitney U-test. Categorical variables were compared using Fisher’s Exact or Chi-squared test, as appropriate. A p-value of < 0.05 was considered statistically significant. All statistical analyses were performed using IBM SPSS^© Version 29.

Results

A total, of 119 patients were included in this study, with a mean age of 45 ± 21 years and 72 (61%) male patients. We identified 13 (11%) patients with pre-existing respiratory disease and 21 (18%) were ever-smokers. No patients had a known history of pre-existing pulmonary hypertension. The patient demographics and co-morbidities at time of TB infection diagnosis are presented in Table 1.

Table 1 Baseline characteristics of patients with tuberculosis (TB) who completed treatment at an Australian TB clinic between January 2013 and December 2022

The TB diagnosis was confirmed by positive mycobacterial culture and/or TB polymerase chain reaction (PCR) in 75 (63%) patients, while a clinical diagnosis was made in 44 patients (37%) based on histopathological and/or radiological evidence in the absence of culture or PCR positivity and with an expected improvement with TB treatment. There were 36 (30%) patients with extra-pulmonary involvement, with affected sites including lymph node (n = 19,16%), pleura (n = 9, 8%), musculoskeletal (n = 3, 3%), urinary tract (n = 2, 2%), peritoneum (n = 1,1%), central nervous system (n = 1, 1%) and ocular (n = 1, 1%). Drug resistance to at least one first-line medication (Rifampicin, Isoniazid, Pyrazinamide and Ethambutol) was found in 10 (8%) of patients.

All patients achieved a successful treatment outcome in accordance with international definitions of treatment completion [4]. The mean (SD) treatment duration was 9 (3) months and time to microbiological cure, where relevant, was 47 (39) days. Other TB disease characteristics and treatment course are summarised in Table 2.

Table 2 Tuberculosis disease characteristics and treatment course

A summary of post-treatment investigations is presented in Table 3. The most commonly performed investigation was chest X-ray (CXR) in 100 patients (84%), followed by chest CT scans in 76 patients (64%), and complex pulmonary function testing including spirometry, plethysmography, and gas transfer capacity in 51 patients (43%). Directed investigation for pulmonary hypertension was not possible as no patient underwent a RHC and only 13 (11%) completed a TTE.

Table 3 Summary of investigations completed post TB treatment

Laboratory lung function testing was performed in 51 (43%) of patients following TB treatment: 38 (75%) had abnormal lung function testing results: 25 (49%) an obstructive defect, 11 (22%) a restrictive defect, and 26 (51%) had impaired gas transfer capacity (DLCO).

Of the 20 (17%) patients with available lung function testing performed before and after TB diagnosis. The mean(SD) time interval between tests was 2.3 (1.6) years with the mean (SD) reduction in FEV1, FVC and TLC 140 (40) mls, 80 (77) mls and 0.26 (1 L) respectively. The corresponding annual rate of decline in FEV1, FVC and TLC was of 60 mls/yr, 35 ml/yr and 113 mls/yr respectively (Table 4).

Table 4 Pulmonary function abnormalities after successful TB treatment

Chest CT imaging was performed in 76 (64%) patients following TB treatment. Radiological abnormalities were observed in 70 (92%) patients: nodular changes (n = 49, 65%), fibrosis (n = 43, 57%), bronchiectasis (n = 22, 29%), lymphadenopathy (n = 21, 28%) and radiological emphysema (n = 11, 15%) (Table 5).

Table 5 Radiological abnormalities following successful TB treatment

No direct evidence of pulmonary hypertension was found following TB treatment in our cohort as no patient underwent RHC. Only 13 (11%) patients had TTE following TB treatment, of which 2/13 (15%) patients had echocardiogram defined measurements consistent with pulmonary hypertension. Pulmonary hypertension was suspected in 52 patients based on chest CT findings, representing 44% of the entire cohort and 68% of patients who had a chest CT scan post TB treatment. Of patients who had a chest CT scan post-TB treatment (n = 76), 43% had a raised PAD > 29 mm and 33 (42%) had a PA: A ratio > 1 (Table 5).

There were 81 patients in the cohort who fulfilled the definition of PTLD based on lung function and/or radiological abnormalities, corresponding to 68% of the entire cohort and 95% of patients who underwent either pulmonary function testing or chest CT imaging post TB treatment. To identify predictors of PTLD, baseline patient and TB disease characteristics were compared between patients with and without PTLD. Although the PTLD cohort had a higher proportion of male patients (65% vs. 50%) and drug-resistant cases (11% vs. 3%), neither variable reached statistical significance (p.>0.05). Other factors including age (45 (20) vs. 47 (21), p = 0.44), extra-pulmonary involvement (30% vs. 32%, p = 0.83), culture positivity (57% vs. 47%, p = 0.34), time to microbiological cure (46 (41) vs. 50 (37) days, p = 0.48) and total treatment duration (9 (3) vs. 9 (2), p = 0.99) were not significantly different between patients with and without PTLD (Table 6).

Table 6 Baseline characteristics of patients with and without Post TB Lung Disease

Discussion

Post Tuberculosis Lung Disease (PTLD) was identified in 68% of all adults in this Australian cohort following successful TB treatment. We identified high rates of abnormal lung function by spirometry and gas transfer capacity as well as radiological abnormalities by CT imaging. Whilst only 6% of the cohort reported COPD as a co-morbid condition and 2% pulmonary fibrosis, we found evidence of obstructive lung function defects in 49% of the cohort and restrictive defects in 22%. Imaging demonstrated fibrotic changes in 57% of those that underwent CT scans, bronchiectasis in 29% and emphysema 15%. Indirect evidence of pulmonary hypertension with abnormal dilation of the pulmonary artery of greater than 29 mm or larger than the aortic diameter was found in 68% of TB survivors with CT scans performed.

The prevalence of PTLD in our cohort, although high, is likely to have been underestimated as not all TB survivors underwent all diagnostic testing. Just over two thirds (72%) of the cohort had either a chest CT scan or complex lung function testing after TB treatment, both of which provide an opportunity to diagnose PTLD. When considering those TB survivors who underwent either pulmonary function testing or chest CT scan, PTLD was detected in 95% of those that had successfully completed TB treatment. The clinical significance of these findings, and their correlation with symptoms or healthcare utilisation was not assessed in this study but are considered by the authors to be important for future research. Importantly, this study does not suggest that TB is the sole cause of all respiratory disease and associated impairments diagnosed in this cohort, only that this TB cohort has significant respiratory disease (consistent with a PTLD definition). Regardless of co-factors to TB implicated in chronic respiratory disease, past literature suggests that people who successfully complete TB treatment face higher long-term mortality rates despite successful treatment outcomes [7]. This suggests a need for patient follow-up beyond TB treatment completion and the consideration of respiratory diagnostic assessments to optimise lung health [17, 18].

Our TB cohort of adults with mostly pulmonary TB (only) demonstrated a high prevalence of lung function impairment post TB treatment. Whilst 30% of the cohort were extra-pulmonary TB, we know that some of those people will also have some degree of pulmonary involvement if you look hard enough (such as pulmonary nodules or other radiological findings on CT in the patient with lymph node or pleural TB). Nonetheless, it is possible that people with pulmonary TB were more likely to have PFT performed. We found that 75% of TB survivors who underwent pulmonary function testing post TB demonstrated abnormalities below the lower limit of normal. For example, over half (51%) had impaired gas transfer capacity, while obstructive and restrictive lung function defects were common (49% and 22% respectively). In other studies, spirometry is the most common assessment of lung function used post TB treatment [6]. A recent meta-analysis for example, found a similar prevalence of restrictive defects to our study (21%), but lower rates of obstructive defects (18%) after TB treatment [19]. Our study, which also included CT imaging, identified emphysema and bronchiectasis, which may explain the higher prevalence of obstructive pulmonary defects that we identified in our cohort, since radiological findings conducive in our cohort are frequently associated with obstructive lung disease.

The inclusion of gas transfer capacity in our study increases the sensitivity to detect PTLD as a reduction in DLCO in post TB survivors may also correspond to the presence of pulmonary hypertension. While direct measurements of the pulmonary pressure via right heart catheterisation (RHC) is required for a diagnosis of pulmonary hypertension, the prevalence of suspected pulmonary hypertension in our study (44%) was based on indirect CT findings (raised PAD diameters and/or an elevated PA). Unfortunately, no TB survivors in our cohort underwent a RHC and only 13 (11%) completed a TTE, of which 2/13 (15%) demonstrated evidence of pulmonary hypertension. The interpretation of these results are complicated by cardiorespiratory co-morbidities in these two TB survivors, however neither patient had known pulmonary hypertension at the time of TB diagnosis. Nonetheless, a prevalence of Post TB pulmonary hypertension of 15%, if replicated in larger studies, is concerning given the large number of TB survivors globally.

There is paucity of published data on pulmonary hypertension caused by PTLD. Prior small studies based on TTE data found pulmonary hypertension in 9–47% of people following TB treatment [20, 21]. The clinical significance of pulmonary hypertension in PTLD is also uncertain, but concerning particularly as it pertains to overall survival. Pulmonary hypertension is known to be an adverse prognostic factor in other respiratory conditionings including interstitial lung diseases [22, 23] and COPD [24, 25]. Thus, the diagnosis of Post TB pulmonary hypertension may have significant long-term survival implications, and at the same time, allow potential options for treatment given the availability of targeted therapies such as pulmonary vasodilators in selected patients [26]. Clearly more research is needed in this fascinating area of PTLD.

Nearly all TB survivors who underwent CT chest imaging after TB treatment showed abnormalities likely to be clinically significant and medically important. For example, pulmonary fibrosis, bronchiectasis and emphysema were identified in 57%, 29% and 15% of patients respectively which is consistent with a prior systematic review, which reported rates of pulmonary fibrosis (25–75%), bronchiectasis (35–86%) and emphysema (15–45%) after TB treatment [27]. The heterogeneity of structural and functional abnormalities suggests the presence of clinical phenotypes within PTLD which may be amendable to targeted therapies through a “treatable traits” approach to complex respiratory disease management that has been used for asthma and Long COVID management [28]. Further research, as advocated by the 2nd international Post TB symposium, is required to better characterise the clinical phenotypes of PTLD and to describe the natural history [11].

A unique aspect of our study was the availability of lung function data prior to TB diagnosis in a subset of patients. We were able to demonstrate an apparent accelerated lung function decline following TB disease despite successful treatment as some of our TB survivors (uncommonly) had lung function test results available before and after their TB diagnosis as they had known prior respiratory disease or clinical symptoms that required lung function testing. These TB survivors experienced a mean FEV1 decline in of 60 mls/yr. Previous studies in healthy adults and those with smoking-related COPD have reported the rate of FEV1 decline to be 22.4 ml/yr and 66.3 ml/yr respectively [29, 30]. Based on these estimates, our PTLD cohort experienced a decline in lung function that was 2.7 times faster than in healthy adults and comparable to that seen in COPD patients. Our analysis of lung function decline in our cohort is limited by the small proportion with known lung function prior to TB (n = 20, 17%) and the presence of concurrent respiratory co-morbidities in 45% of these patients. The variables or co-factors resulting in lung function stability or decline after successful TB treatment are unclear and warrant further investigation to consider the trajectory of lung function impairment Post TB [31].

The predictors of PTLD (or no PTLD) also remain elusive. We found no statistically significant association between baseline patient characteristics and the development of PTLD in our study. While the PTLD cohort had a higher proportion of males (65% vs. 50%, p = 0.11) and drug resistant TB (11% vs. 3%, p = 0.12), these associations did not reach statistical significance, possibly due to the small sample size. Other research has suggested that host and pathogen (TB) factors are implicated in PTLD development [12, 32]. Host factors with a significant association with PTLD include advanced age, tobacco smoking, and malnutrition [33] while pathogen factors include drug resistance, culture positivity, extent of pulmonary involvement, and prolonged treatment duration [12, 34]. Identifying PTLD predictors could provide insight into the pathophysiology of the disease, improve prognostication, and help to personalise TB care.

Our study had several limitations. The prevalence of PTLD was likely underestimated given not all participants had the required investigations for detection. This limitation was most evident for the assessment of pulmonary hypertension, where the prevalence could only be inferred from TTE and CT imaging data. Despite these limitations, the high prevalence of PTLD observed supports the need for comprehensive assessment of PTLD among TB survivors to better understand its relationship with symptoms, healthcare utilisation, and long-term survival.

Conclusion

Post-tuberculosis lung disease appears to be highly prevalent in our Australian setting despite clinical and microbiological cure. The spectrum of PTLD identified includes obstructive lung disease, bronchiectasis, pulmonary fibrosis, restrictive lung function and pulmonary hypertension. These chronic respiratory conditions are likely to contribute to the morbidity and mortality experienced by TB survivors. Further research is needed to understand how to best diagnose and manage these complex respiratory conditions particularly in resource limited countries where ready access to the investigations required to confirm the diagnosis may be restricted.