Introduction

Tuberculosis (TB) remains a significant cause of death as an infectious disease. In 2021, approximately 10.6 million people worldwide developed TB, resulting in an estimate of 1.4 million deaths1. Around 167.1 thousand individuals were confirmed to have multidrug-resistant tuberculosis (MDR-TB) globally, while 25.1 thousand patients were confirmed to have pre-extensively-drug resistant/extensively-drug resistant tuberculosis (pre-XDR/XDR-TB). The treatment success rate for multidrug-resistant/rifampicin-resistant tuberculosis (MDR/RR-TB) cases was only 60%1. Specifically, China carried 7.4% of the global TB burden and 16.8 thousand patients were confirmed MDR/RR-TB with a treatment success rate of 53%1. MDR/RR-TB continues to be a significant public health concern in China.

Tracheobronchial tuberculosis (TBTB) is a form of tuberculosis that specifically affects the tracheobronchial tree. TBTB may lead to chronic airway diseases, including lumen stenosis, obstruction, pulmonary atelectasis, and even respiratory failure. Based on previous researches, around 6–50% of patients diagnosed with pulmonary tuberculosis (PTB) also had TBTB2,3,4,5,6. TBTB is the leading cause of benign airway stenosis in China. However, the clinical manifestations of TBTB lack specificity, making it prone to misdiagnosis and missed diagnosis. In severe cases, such as multidrug-resistant tracheobronchial tuberculosis (MDR-TBTB), the difficulty of treatment is further increased. Currently, there is a lack of studies investigating the use of bedaquiline or delamanid in the treatment of MDR-TBTB. China introduced the clinical use of bedaquiline in 2018 under the project of New Drug Introduction & Protection, and according to the guidelines of the World Health Organization (WHO)7 and the consensus of China8 in 2018 and 2019 respectively, bedaquiline was recommended as the initial drug for MDR-TB patients in group A. Delamanid was introduced in China in March 2018 and was subsequently included in the recommended drugs for group C in 20198. During the period of 2017–2021, we utilized bedaquiline and delamanid and aimed to assess their impact, particularly in patients with MDR-TBTB, as part of this study. This study aims to provide a comprehensive description of the entire phase of MDR-TBTB and investigate the effects of anti-tuberculosis treatment regimens containing bedaquiline or delamanid on the clinical outcomes of this disease.

Methods

Study design and participants

This study was a retrospective cohort study conducted with patients diagnosed with multidrug-resistant/rifampicin-resistant/pre-extensively-drug resistant tracheobronchial tuberculosis (MDR/RR/pre-XDR-TBTB). The participants were recruited from Public Health Clinical Center of Chengdu (PHCC) (Chengdu, China) between November 2017 and November 2021. All patients were bacteriologically or clinically diagnosed as TBTB according to the guideline of China9, and the criteria was shown in the case definition section. This study included adults (≥ 18 years) with MDR/RR/pre-XDR-TBTB. Patients with drug-susceptible TBTB and viral infection were excluded. Patients enrolled were classified into the bedaquiline group, the delamanid group, and the control group.

This study was approved by the PHCC Human Ethics Committee (YJ-K2022-20-01). Informed consent was obtained from patients. All research procedures adhered to the guidelines outlined in the Declaration of Helsinki.

Case definition of TBTB

There is no globally accepted standard for diagnosing TBTB currently. In this study, qualified physicians followed the guideline for TBTB in China9. The diagnosis was based on various criteria, including clinical symptoms such as irritable cough, expectoration, hemoptysis, or dyspnea. Additionally, positive results in tests such as Acid-fast bacilli (AFB) smear, Mycobacterium tuberculosis (MTB) PCR, GeneXpert MTB/rifampin (RIF) (Xpert), and MTB culture were considered. Bronchoscopy and biopsy were also utilized for diagnosis9. In a previous study by Chung in 200010, TBTB was classified into seven subtypes. However, the classification used in this study differed from Chung's and included 6 subtypes such as inflammatory infiltration, ulceration necrosis, granulation hyperplasia, cicatrices stricture, tracheobronchial malacia, and lymph fistula9. For more detailed information about the bronchoscopic appearances of each subtype, please refer to the published study11.

Definition of drug resistance

According to WHO12, MDR/RR-TBTB referred to either multidrug-resistant (resistant to rifampicin and isoniazid) TBTB or rifampicin-resistant TBTB. Pre-XDR-TBTB referred to TBTB resistant to rifampicin (and might also be resistant to isoniazid), and that was also resistant to at least one fluoroquinolone (FQ) (either levofloxacin or moxifloxacin).

Data collection

Epidemiological, demographic, clinical data, baseline laboratory tests, imaging, and bronchoscopy reports, were extracted from medical records. Prognoses were followed up either in person or by telephone. Two physicians (QC and TH) reviewed all the data, and a third researcher (WH) assessed any discrepancies in interpretation between the two primary reviewers.

Laboratory examination

All patients underwent testing for human immunodeficency virus (HIV) with HIV antibody. Respiratory samples containing sputum and bronchoalveolar lavage fluid (BALF) were collected and examined for Xpert and using smear microscopy for AFB. The samples were then pretreated for purification and homogenization13, and the BACTEC MGIT960 system (Becton Dickinson & Co., Franklin Lakes, NJ, USA) was utilized for MTB culture. As this study was retrospective, not all patients had tested completely. Therefore, the lacking test results were not presented. Drug susceptibility testing was performed on all patients using either molecular or genotypic techniques to identify resistance-conferring mutations or phenotypic methods to evaluate susceptibility to anti-tuberculosis drugs.

Bronchoscopy examination and interventional treatment

All patients underwent bronchoscopy upon admission. The TBTB subtype and lesion location were carefully documented, and the degrees of stenosis were categorized as mild (< 50%), moderate (50–75%), or severe (> 75%) based on previous studies14,15. Ten patients underwent interventional treatment, of whom three received cryotherapy and topical administration, two received cryotherapy, two received topical administration, two received balloon dilatation, and one received cryotherapy along with balloon dilatation and argon plasma coagulation. The median number of interventional therapy was 1.00 (1.00, 3.00).

Anti-tuberculosis treatment

These regimens were long-term and managed by qualified physicians, following the guidelines of WHO16,17,18,19 and China3. The bedaquiline-containing regimen received either 400 mg of bedaquiline once daily for 2 weeks, followed by 200 mg three times a week for 22 weeks, and the delamanid-containing regimen received 100 mg of delamanid twice daily for 24 weeks. New anti-tuberculosis drugs were used in combination with a preferred background regimen.

Definitions of outcomes

The definitions of treatment outcomes were classified into 7 categories: cured, treatment completed, treatment failed, died, lost to follow-up, not evaluated, and treatment success20. Treatment success included cured and treatment completed. Culture conversion was defined as two or more consecutive cultures taken at least 30 days apart, that tested negative in a patient who initially tested positive at baseline.

Two experienced radiologists analyzed the chest CT images together. In this study, completely pulmonary reexpansion was defined as the recovery of lung volume exceeding 80% of the estimated original volume from chest CT images21. Partially reexpansion was defined as the recovery of lung volume less than 80% of the estimated original volume. In Table 4, the term ‘constant’ indicated that the chest CT imaging features did not show significant changes, while the term ‘deterioration’ referred to further progression. Additionally, the term ‘receiving surgery’ referred to receiving pulmonary lobectomy. The term ‘improved’ was defined as a reduction in the severity of airway stenosis compared to the baseline.

Statistical analysis

Continuous variables that followed a normal distribution were expressed as mean ± standard deviation, while others were expressed as medians (Q1, Q3). Categorical variables were represented as n (%). One-way ANOVA was used to compare continuous variables among multiple groups. If variances were not uniform, the Kruskal–Wallis H test was used. The Chi-square test was used to compare categorical variables between groups. If the categorical variables were ordinal rank variables, a single ordinal link table analysis (Kruskal–Wallis H test) was performed. The Kaplan–Meier curve was used to represent the time of culture conversion, and the log-rank test was used to analyze differences between groups. Predictors of culture conversion during early treatment were selected based on the baseline clinical characteristics, examinations, and treatments. Univariable logistic regression was initially performed. Multivariable logistic regression was performed by including variables that were significant at P < 0.2 in univariable logistic regression, along with all treatment-related variables. Collinearity analysis was conducted to estimate the correlation between predictors. Two-tailed p-values less than 0.05 were considered statistically significant. All statistical analyses were performed using SPSS version 24.0 (SPSS Inc, Chicago, IL, USA).

Results

Baseline demographic and clinical characteristics

This study recruited 101 patients with MDR/RR/pre-XDR-TBTB. Among them, 32 were assigned to the bedaquiline group, 25 to the delamanid group, and 44 to the control group. The median age was 33 years, and 40.6% of patients were male. The average BMI was 20.47 ± 2.69 kg/m2, and 26 patients (25.7%) were underweight. Eighteen patients (17.8%) had heavy alcohol consumption. The most common comorbidities observed were anemia (16.8%), diabetes mellitus (9.9%), and hypoproteinemia (6.9%). Only one patient in the control group was HIV infected. About 77.2% of patients had a history of TB treatment. Most of patients (68.3%) received treatment within 6 months from the onset of symptoms (data not shown). A positive AFB of sputum smears was detected in 33 patients (32.7%). The control group had a higher prevalence of chronic renal disease (13.6%), combined with extrapulmonary TB (50.0%) and MDR/RR-TBTB (81.8%) (all P < 0.05). Additionally, the bedaquiline group had a higher rate of positive culture of sputum (P = 0.027) (Table 1).

Table 1 Baseline characteristics of patients with MDR/RR/Pre-XDR-TBTB at treatment initiation.
Full size table

Imaging features

In the baseline chest CT images, it was lung damage was observed in 15 patients (14.9%), atelectasis in 43 patients (42.6%), airway stenosis in 70 patients (69.3%), roughness of bronchial wall in 65 patients (64.4%), and cavities in 63 patients (62.4%). TB lesions involving bilateral lungs and multiple lobes were the most common observation (72.3%). Additionally, the bedaquiline group showed a lower proportion of atelectasis among the three groups (P = 0.001) (Table 2).

Table 2 CT imaging features of patients with MDR/RR/pre-XDR-TBTB using bedaquiline or delamanid or none.
Full size table

Bronchoscopic examination

All patients underwent bronchoscoy upon admission, with the most common TBTB subtype being ulceration necrosis (32.7%). The majority of lesions observed under bronchoscopy were found in both upper lobe bronchi and the left main bronchus. Stenosis was observed in 74.3% of the patients, while 14.9% presented obstruction. (Table 3).

Table 3 Bronchoscopic examination characteristics of patients with MDR/RR/pre-XDR-TBTB using bedaquiline or delamanid or none.
Full size table

Treatment and outcome

The median number of anti-tuberculosis drugs was 6. The bedaquiline group used FQ less frequently than the other two groups, while the bedaquiline or delamanid group used linezolid more frequently than the control group (all P < 0.001). The majority of patients (98.2%, 56/57) continued to use bedaquiline or delamanid for 24 weeks. Only one patient discontinued the use of bedaquiline at 16 weeks due to QTcF > 500 ms.

Sixty-seven patients (66.3%) were cured, 25 (24.8%) completed the treatment, 3 (3.0%) could not be evaluated and 6 (5.9%) were lost to follow-up. Additionally, 92 patients (91.1%) achieved treatment success. The cure rates of the delamanid and bedaquiline group were higher than the control group (76.0% vs. 56.8% and 71.9% vs. 56.8% respectively), but the differences of cure and success rates among groups were not statistically significant (all P > 0.05). The sputum culture conversion (SCC) rate reached 97.6% (83/85) at the end of treatment. Two cases had unknown outcomes of culture conversions. Moreover, the duration of treatment in the delamanid group was significantly shorter than the other two groups (P = 0.039). The SCC rates of the bedaquiline or delamanid group were significantly higher than the control group after 2 or 6 months of treatment, and the mean times to SCC were significantly shorter (all P < 0.001).

Among the patients who underwent chest images at the end of treatment, 69.4% of patients (25/36) showed reexpansion of atelectasis, and half of patients (30/60) showed improvement in lumen stenosis with a cavity closure rate of 59.3% (32/54). Patients in the delamanid and control group showed better improvement in lumen stenosis compared to the bedaquiline group, respectively (P = 0.042). (Table 4).

Table 4 Treatment and outcomes of patients with MDR/RR/pre-XDR-TBTB using bedaquiline or delamanid or none.
Full size table

Sputum culture conversion

Eighty-three patients (97.6%, 83/85) achieved culture conversion at the end of treatment. The culture results were unknown for two patients. The sputum culture monitoring frequency for 3 patients exceeded two months after six months of treatment. As a result, the conversion times of sputum culture for these five patients were recorded as missing values. The median time for culture conversion was 4 weeks [95% confidence interval (CI) 2.249–5.751]. There was a significant difference in the time to culture conversion among the three groups (control group: 12 weeks, bedaquiline group: 2 weeks, delamanid group: 2 weeks, P < 0.001) (Fig. 1).

Figure 1
figure 1

Time to culture conversion according to the drug group.

Full size image

Predictors of culture conversion at the early phase of treatment

Tables 5 and 6 presented the predictors of culture conversion after 2 and 6 months of treatment in 83 patients. The results of the multivariable logistic regression analysis showed that treatment with bedaquiline (aOR = 13.417, 95% CI 4.067–44.260) and delamanid (aOR = 9.333, 95% CI 2.498–34.878) were associated with culture conversion after 2 months of treatment (Table 5). Furthermore, treatment with bedaquiline (aOR = 13.333, 95% CI 3.379–52.610) and delamanid (aOR = 5.000, 95% CI 1.357–18.426) were also associated with culture conversion after 6 months of treatment (Table 6).

Table 5 Predictors of culture conversion at 2 months in the patients with MDR/RR/pre-XDR-TBTB.
Full size table
Table 6 Predictors of culture conversion at 6 months in the patients with MDR/RR/pre-XDR-TBTB.
Full size table

Discussion

In this study, we described the demographic and clinical characteristics, as well as treatments, and outcomes of 101 patients with MDR/RR/pre-XDR-TBTB. Additionally, we analyzed the culture outcomes in the early treatment phase for 83 patients. Our findings indicated that the bedaquiline or delamanid group achieved faster culture conversion compared to the control group during the early treatment phase for MDR-TBTB.

Young females were more susceptible to MDR-TBTB, consistent with previous findings on TBTB15,22,23. The control group had a higher prevalence of chronic renal disease and extrapulmonary TB. Possibly due to that during the early use of new anti-tuberculosis drugs, the population selected had less renal function damage and fewer extrapulmonary TB to minimize the interference on treatment effects and safety.

It was observed that the clinical manifestations of TBTB were non-specific, consistent with previous findings15,22,23. Pulmonary atelectasis was observed in 42.6% of the patients, which is higher than previous report (11.4%)15. This could be attributed to the higher ratio of drug resistance in this study compared to the previous (100% vs. 31.4% in adults)15. Notably, imaging and bronchoscopic features can occasionally be contradictory. Chest imaging only reflect indirect signs, while bronchoscopy is more intuitive and precise, but typically limited to reaching areas up to the segmental bronchus. Therefore, for an accurate diagnosis of TBTB, it is necessary to combine bronchoscopy with thin-slice chest CT.

According to the guideline of TBTB in China9, the subtypes of inflammatory infiltration, ulceration necrosis, granulation hyperplasia, and lymph fistula are classified as the clinical active stage, while cicatrices atricture and tracheobronchial malacia are classified as the clinical stable stage9. In this study, a higher percentage (76.2%) of the clinical active stage of TBTB was observed, and lesions involving the left main bronchus was frequently observed, consistent with previous research15, as well as the rates of stenosis (74.3%) and obstruction (14.9%) under bronchoscopy. These findings suggest that drug resistance has a limited impact on TBTB subtypes, lesion location, lumen stenosis, and obstruction.

In the treatment section, the bedaquiline group had a lower proportion of FQ containing, mainly because this group had a higher proportion of pre-XDR-TBTB compared to the other two groups (P < 0.05). Additionally, linezolid was more frequently used in the bedaquiline and delamanid group compared to the control (P < 0.05). Possibly, these two groups primarily followed the guidelines of WHO and China after 20193,18,19, which recommended linezolid as one of the initial choices in the treatment regimen. In contrast, the control group predominantly received the traditional long-term regimen. Furthermore, only 9.9% of the total patients received interventional therapy, and the limited number may reduce the impact of interventional therapy on clinical outcomes. The interventional therapy for MDR/XDR-TBTB still need further investigation.

The treatment outcomes of all three groups showed high rates of success (84.0–96.9%). The bedaquiline group exhibited a high treatment success rate of 96.9%, which aligns with previous studies showcasing the efficacy of bedaquiline in reducing all-cause mortality and either increased treatment success or showed no reduction24,25,26,27,28,29,30. The delamanid group had a treatment success rate of 84.0%, consistent with previous studies revealed that delamanid-containing treatment was not inferior to the placebo group26,31. A meta-analysis found that the success rates of delamanid-containing regimens in observational and experimental studies were 80.9% (95% CI 72.6–87.2) and 72.5% (95% CI 44.2–89.8) respectively32, which were similar to our study. Furthermore, no statistically significant differences were found in cure and treatment success rates among the three groups, consistent with a previous study showing that the use of new drugs did not compromise the treatment success rate33. These findings suggested that following the anti-tuberculosis regimen based on previous and existing guidelines after 20163,16,17,18,19, along with strict follow-up monitoring and management, can lead to a favorable outcomes for MDR/RR/pre-XDR-TBTB.

The SCC rates of the bedaquiline and delamanid group at 6 months were 96.8% and 100% respectively, which were higher than the control (72.2%) (P < 0.001), and also higher than previous studies for MDR-TB (bedaquiline-containing regimen: 78.8–85.3%25,34,35; delamanid-containing regimen: 45.4–87.6%26,27,36,37,38,39). Additionally, the median time to culture conversion for these two groups was significantly shorter than the control (P < 0.001). It appeared that the better benefit might be achieved using shorter regimens with bedaquiline or delamanid for MDR-TBTB, particularly in the first six months of treatment. Moreover, this approach could potentially be more effective for culture conversion treating TBTB compared to PTB.

At the end of treatment, it was observed that the rate of stenosis improvement for the delamanid group (64.3%) was higher compared to the bedaquiline (26.3%) and control group (59.3%). This suggested that delamanid may be more effective in treating lesions in the tracheobronchial tree. A previous study showed that delamanid was broadly distributed to various tissues in addition to the lungs in rats40. Additionally, another study simulated the plasma and tissue concentrations of delamanid in humans and found that the simulated concentrations of delamanid at steady state in the lung, brain, liver, and heart were higher than the effective concentration for MTB in vivo41. This indicated that delamanid may achieve similar concentrations in various organs to that of the lung41. Therefore, delamanid may have a steady state concentration in tracheobronchial tissues, and appeared to have the potential to effectively improve airway stenosis of TBTB. Moreover, a previous study found that delamanid can suppress the expression of CXCL10 via regulation of JAK2/STAT1 signaling and correlates with reduced inflammation in MDR-TB patients42. Therefore, delamanid may potentially contribute to the improvement of airway stenosis in MDR-TBTB. Additional research is required to validate these findings and to further investigate the specific mechanism of delamanid for MDR-TBTB.

In our study, we observed that the median time for culture conversion in the bedaquiline and delamanid group were significantly shorter than the control (2 weeks vs. 12 weeks), which were even shorter compared to previous studies (bedaquiline-containing regimen: 83 days43; delamanid-containing regimen: 74.5 days34) in patients with MDR-TB respectively. The reason need to be further explored.

The results of multivariable logistic regression indicated that treatment with bedaquiline or delamanid was associated with culture conversion within 2 and 6 months of treatment. Furthermore, patients with MDR/XDR-TB treated with a regimen containing both bedaquiline and delamanid could achieve a SCC rate ranging from 74 to 100% after 6 months of treatment44,45,46,47,48,49. These findings suggested that a shorter anti-tuberculosis regimen containing bedaquiline and/or delamanid could be considered for MDR-TBTB to achieve faster culture conversion during the early treatment. Additionally, it is worth noting that the involvement of multiple bronchi (P = 0.077) and the presence of stenosis (P = 0.087) observed by bronchoscopy could be potential risk factors in achieving culture conversion after 2 months of treatment. However, treatment with linezolid (P = 0.082) may be beneficial. Moreover, inflammatory infiltration (P = 0.086) may be advantageous in achieving culture conversion after 6 months of treatment. To ensure comprehensive results, it is recommended to expand the sample size for future studies.

Moreover, the high cost of MDR-TBTB treatment should be considered, including expenses for bronchoscopy, interventional treatments, and long-term regimens, especially with new anti-tuberculosis drugs, which further exacerbates the financial challenges faced by low-income families in a developing country like China. In China, the per capita cost of patients with MDR-TB receiving regimen containing bedaquiline (45% patients receiving a short-course treatment lasting 9 months) was 138,000 yuan, roughly double that of the conventional treatment regimen (lasted for 16 months)50. However, there is limited research on the cost of regimens containing delamanid in China. Therefore, there is a potential opportunity to use shorter regimens (6–9 months) with new anti-tuberculosis drugs like bedaquiline and/or delamanid for MDR/XDR-TBTB patients in China. This approach has the potential to reduce treatment duration and lower costs.

This study still has some limitations. As a retrospective study, we did not conduct regular follow-up of bronchoscopy or lung function. Furthermore, the sample size of interventional treatment was limited. Additionally, our study did not include regimen containing both bedaquiline and delamanid. Therefore, it would be beneficial to consider prospective analyses and expanding the sample size for further investigation.

Conclusions

Anti-tuberculosis regimens containing bedaquiline or delamanid may accelerate the culture conversion and improve the SCC rates during the early treatment phase in MDR-TBTB, and delamanid appears to have the potential to effectively improve airway stenosis in MDR-TBTB.