|Year : 2021 | Volume
| Issue : 1 | Page : 1-6
Utilizing flexible bronchoscopy for the diagnosis of endobronchial tuberculosis with negative sputum acid-fast bacillus
Nai-Chien Huan1, Nurul Izzati Azmin Mohd Yusof1, Hema Yamini Ramarmuty1, Teng Shin Khoo2, Yean Chen Lai2, Shan Min Lo1, Kunji Kannan Sivaraman Kannan1
1 Department of Respiratory Medicine, Queen Elizabeth Hospital, Kota Kinabalu, Sabah, Malaysia
2 Department of Medicine, Labuan Hospital, Federal Territory of Labuan, Malaysia
|Date of Submission||23-Apr-2020|
|Date of Decision||31-Aug-2020|
|Date of Acceptance||16-Oct-2020|
|Date of Web Publication||15-Feb-2021|
Department of Respiratory Medicine, Queen Elizabeth Hospital, Kota Kinabalu, Sabah, Malaysia
Source of Support: None, Conflict of Interest: None
Background: Endobronchial tuberculosis (EBTB) should be viewed as a distinct subset of tuberculosis due to various pitfalls: (a) often diagnosed late, (b) mimicking other conditions, for example, lung carcinoma, and (c) risks of bronchostenosis if treated late. Flexible bronchoscopy (FB) offers a reliable and safe modality to increase the diagnostic yield of EBTB when sputum acid-fast bacillus (AFB) is negative. In this study, we aim to determine the clinical presentation, bronchoscopic characteristics, and safety of FB for the diagnosis of EBTB at our centers. Methods: From September 2018 to December 2019, 25 patients with EBTB from Queen Elizabeth Hospital and Labuan Hospital, Malaysia were enrolled in this study. Histopathology and/or microbiology were diagnostic in all patients. All patients underwent FB only after sputum smears for AFB were negative. Results: The most common presenting complaint was cough (68%), followed by loss of weight (52%), loss of appetite (36%), fever (32%), hemoptysis (28%), and dyspnea (24%). Lung mass/nodule was seen on imaging in 56% of patients, followed by consolidation (36%), cavity (36%), tree-in-bud appearance (32%), and collapse (24%). The most common Chung’s classification of EBTB appearance was edematous-hyperemic (60%); other appearances in order of descending frequencies were: fibro-stenotic (24%), caseating (24%), tumorous (16%), and granular (16%). Bronchial lavage cultures and/or histopathology cultures were positive for tuberculosis in 72% of patients. No procedure-related mortality or major complications were reported. Conclusion: Even in tuberculosis endemic settings, empirical pharmacological treatment of all suspected EBTB cases will inadvertently lead to delay in attaining competing differential diagnoses in some. FB in selected patients is safe and effective as it enhances treatment confidence by providing histological and/or microbiological evidence thereby ruling out other differentials when initial sputum results were inconclusive.
Keywords: Bronchoscopy, endobronchial tuberculosis, tuberculosis
|How to cite this article:|
Huan NC, Mohd Yusof NA, Ramarmuty HY, Khoo TS, Lai YC, Lo SM, Sivaraman Kannan KK. Utilizing flexible bronchoscopy for the diagnosis of endobronchial tuberculosis with negative sputum acid-fast bacillus. J Assoc Chest Physicians 2021;9:1-6
|How to cite this URL:|
Huan NC, Mohd Yusof NA, Ramarmuty HY, Khoo TS, Lai YC, Lo SM, Sivaraman Kannan KK. Utilizing flexible bronchoscopy for the diagnosis of endobronchial tuberculosis with negative sputum acid-fast bacillus. J Assoc Chest Physicians [serial online] 2021 [cited 2021 Apr 10];9:1-6. Available from: https://www.jacpjournal.org/text.asp?2021/9/1/1/309470
Key Messages: Subjecting all suspected cases with EBTB for FB is risky for healthcare workers, whereas empirical pharmacological treatment of all cases will lead to delay in attaining competing differential diagnoses in some. In the absence of formal guidelines, we recommend a “middle ground” practice of obtaining at least three sputum samples for AFB before FB for patients with suspected EBTB.
| Introduction|| |
With more than 10 million new cases and 1.6 million deaths in the year of 2017 alone, tuberculosis is the deadliest infection among mankind. Endobronchial tuberculosis (EBTB) can be defined as a tuberculosis infection of the tracheobronchial tree with or without the involvement of adjacent lung parenchyma. The true incidence of EBTB remains unknown and is likely underestimated as bronchoscopy is not routinely performed. EBTB should be viewed as a distinct subset of tuberculosis due to various pitfalls: (a) often late diagnosis, (b) mimics other conditions, for example, lung carcinoma, and (c) risks of bronchostenosis if treated late. Although the treatment of smear-positive tuberculosis is straightforward and standardized, there is a lack of uniformity in management strategies among clinicians regarding smear-negative cases. Flexible bronchoscopy (FB) offers a safe and reliable modality to increase the diagnostic yield of EBTB when other noninvasive methods, for example, sputum smears for acid-fact bacilli (AFB) and/or imaging, were inconclusive. In this study, we aim to determine the clinical presentation, bronchoscopic characteristics, and safety of FB for the diagnosis of EBTB at our centers.
| Materials and methods|| |
From September 2018 to December 2019, 25 patients with EBTB from Queen Elizabeth Hospital and Labuan Hospital, Malaysia were included in this study. Queen Elizabeth Hospital is a tertiary referral center with full pulmonology services in the state of Sabah, Malaysia, whereas Labuan Hospital is a district hospital with medical specialists and bronchoscopy services.
All patients underwent FB only after noninvasive methods, for example, three consecutive morning sputum smears for AFB were negative. FB was performed in cases of suspected EBTB or to rule out competing differentials such as lung malignancy. Informed consent was taken from our patients following a thorough explanation of the purpose, nature, and possible complications of the procedure. The study was conducted in line with the latest amended Declaration of Helsinki (seventh revision 2013).
Flexible bronchoscopy procedure
All FB were conducted as in-patients at operating theatres or scope rooms by pulmonologists, pulmonology trainee/fellow, or by general medicine physicians under direct supervision. Endoscopists were strongly advised to wear N95 surgical masks during scopes as part of hospital airborne precaution infection control measures for tuberculosis.
All procedures were performed using FB (Olympus or Pantex). Patients were placed in a supine position and received oxygen at 2 to 3 L via a modified nasogastric tube into one of the nostrils. Conscious sedation in the form of fentanyl (25–50 mcg) and midazolam (1–2 mg) was given before the bronchoscope was introduced via the nasal route. The pharynx and vocal cords were anesthetized with lidocaine spray and lidocaine flush via FB, respectively. Bronchoalveolar lavage (BAL) was performed after a full inspection of the airways. BAL involved a gentle introduction of the bronchoscope until impacted or wedged on a subsegmental or segmental bronchus. The standard site of lavage were areas of greatest endobronchial abnormality and/or areas of greatest radiographic abnormality. If no obvious abnormalities were present, the right middle lobe or the left lingular lobe was chosen. Sterile isotonic saline in multiple aliquots of 30 to 50 mL (total volume of 100–250 mL) was instilled with syringes into the target airway through the biopsy channel of the bronchoscope and subsequently retrieved back with repetitive gentle suctioning into a sterile collection bottle. Endobronchial biopsies were performed by using biopsy forceps or snares depending on the characteristics of the endobronchial lesion.
Upon completion of the procedure, patients were monitored in observation bays before discharged or transferred back to the respective wards. Images and clinical findings were stored in a computer database system, available for reference later when needed.
Study procedure and diagnostic criteria
All patient data including demographics, presenting symptoms, imaging findings, bronchoscopy, microbiology, and histopathology results were gathered and evaluated, retrospectively, from our computer database system and patients’ case notes. Bronchoscopic findings were categorized according to Chung’s classifications into the following seven subtypes: actively caseating, fibrostenotic, edematous–hyperemic, tumorous, ulcerative, granular, and nonspecific bronchitic. Patients with bronchorrhea, another common feature of EBTB, were recorded as well.
Microbiological samples in the form of endobronchial biopsy, and/or BAL were obtained from all patients via bronchoscopy. The diagnosis of EBTB was established by (a) demonstrating caseating granulomas or epitheloid cell granulomas on microbiological samples, (b) the presence of Mycobacterium tuberculosis on the staining of microbiological samples, (c) positive cultures for Mycobacterium tuberculosis from microbiological samples, and (d) radiological and clinical improvement upon commencement of antituberculosis medications as per clinician judgment.
All the data are presented as a number with a percentage or mean. All statistical analyses were performed by using Statistical Product and Service Solutions (SPSS) version 2015.
| Results|| |
During our study period from September 2018 to December 2019, a total of 152 flexible bronchoscopes were performed at our centers, out of which 25 cases of EBTB were diagnosed. Out of the total of 25 cases, 22 cases (88%) were from Queen Elizabeth Hospital with the remaining three cases (12%) from Labuan Hospital. Eighteen (72%) of our patients were male and seven patients (28%) were female. The mean age of our patients was at 50.32 years (range: 25–76 years old).
Among patients with EBTB, the commonest presenting complaint was cough (68%), followed by loss of weight (52%), loss of appetite (36%), fever (32%), hemoptysis (28%), and dyspnea (24%). Lung mass/nodule was seen on imaging in 56% of patients, followed by consolidation (36%), cavity (36%), tree-in-bud appearance (32%), and collapse (24%). The commonest Chung’s bronchoscopic subtype for EBTB was edematous-hyperemic (60%); other appearances in order of descending frequencies were: caseating (24%), fibro-stenotic (24%), tumorous (16%), and granular (16%). Profound bronchorrhea was observed in 24% of patients. Patient demographics, clinical features, and radiological features are summarized in [Table 1] while a few photos depicting various appearances of EBTB are shown in [Figure 1].
|Table 1 Patient demographics, clinical features, and radiological features.|
Click here to view
|Figure 1 The many faces of endobrachial tuberculosis. Top left: Edematous appearance of left upper lobe bronchus. Top right: Stenotic right middle and lower lobe bronchus. Middle left: Excessive bronchorrhea of the main trachea. Middle right: Erythematous bronchial wall with the presence of caseous material. Lower left: Tumorous lesion at the right upper lobe. Lower right: Caseous material on a background of edematous bronchia wall.|
Click here to view
Although all patients had negative sputum AFB, 72% patients had positive smears for AFB in BAL. BAL and/or histopathology cultures for Mycobacterium tuberculosis were positive in 72% of patients. Ten patients (40%) had a combination of more than one bronchoscopic subtype finding in different parts of the airway, for example, bronchostenosis at the right middle and right lower segmental bronchus with granular lesions at bronchus intermedius. The highest rate of smear positivity for AFB was found among those with fibrostenotic subtype, in which all six patients (100%) were tested positive. Other bronchoscopic subtypes recorded positivity for AFB in order of descending frequencies were: caseating (83%), edematous-hyperemic (67%), granular (50%), and tumorous (25%). All patients with features of bronchorrhea were tested positive for AFB in their BAL fluids. Endobronchial biopsies taken from all six patients with tumorous and granular subtypes demonstrated caseating granulomatous lesions. The results of different bronchoscopic subtypes together with their smears for AFB and culture results are shown in [Table 2].
Anatomically, most patients (56%) had abnormalities in upper lobes: left upper lobe at 32% and right upper lobe at 24%. Generalized nonspecific erythematous changes were seen in 28% of cases, whereas 20% cases had abnormalities at the right middle lobe. The main trachea, left main bronchus, right main bronchus, left lingular lobe, left lower lobe, and right lower lobe each had one patient (4%) with bronchoscopic abnormalities. The findings are further summarized in [Figure 2].
|Figure 2 Distribution of bronchoscopic abnormalities among patients with EBTB. EBTB = endobronchial tuberculosis.|
Click here to view
There were no serious complications such as deaths directly related to the procedure, injuries to major thoracic organs, and massive bleeding leading to hemodynamic instability related to bronchoscopy in our study.
| Discussion|| |
The exact pathogenesis of EBTB is not fully elucidated yet. Possible explanations include (a) direct implantation of Mycobacterium tuberculosis into the airways from adjacent mediastinal lymph nodes, (b) hematogenous spread, (c) erosion of lymph node or tuberculoma into the airway, or (d) direct spreading from adjacent diseased lung parenchyma., Chung and colleagues in their studies have classified EBTB into seven subtypes based on bronchoscopic appearances, namely: nonspecific bronchitic, granular, edematous-hyperemic, caseating, ulcerative, tumorous, and fibrostenotic. This classification can be regarded as a continuum of EBTB infection: from an early stage (seen as nonspecific inflammatory and erythematous changes) to an intermediate stage in which necrosis takes place (demonstrated by the presence of airway ulcers or caseous materials) in a late fibrotic healing state (airway fibrostenosis). More than one subtype can coexist within the same patient in different parts of the airway, as seen in 40% of patients in our study. Various earlier studies have demonstrated low sputum yield among patients with EBTB as compared to tuberculosis with lung parenchymal involvement., Possible explanation for this is due to sputum and mucous entrapment by proximal granulation tissue. All our patients had negative sputum samples for AFB. Negative sputum results for AFB, therefore, does not preclude the diagnosis of EBTB.
In our study, we performed further sub-analysis of microbiological yield for different bronchoscopic subtypes. Our results demonstrated high AFB yields among those with edematous-hyperemic (66.7% positive yield) as well as a caseating subtype (83.3% positive yield). We believe that the reason behind this finding could be due to the fact that edematous-hyperemic and caseating appearances were early infective lesions in which microbiological replication was still active. To our surprise, all six patients with fibrostenotic subtype were tested positive for AFB in their BAL fluids, which is inconsistent with the understanding that firbostenosis represent a late state of EBTB in which bacterial replication is no longer active. A possible explanation for this result is that most of our patients with endobronchial fibrostenosis (four of six) had other concurrent bronchoscopic features in different parts of the airway (three patients had concurrent caseating features, one had airway hyperemia).
In contrast to smear-positive tuberculosis, the diagnosis of EBTB with negative sputum AFB is more challenging where diagnostic delays are often observed. Currently, there are no standardized clinical practice guidelines or recommendations with regards to the management of patients with suspected EBTB. In both of our hospitals, patients were offered FB only after at least three of their sputum results were negative for AFB. Smear positive cases were empirically treated with antituberculosis medications without performing FB. We believe that such “middle ground” practice is suitable in our local setting in which tuberculosis is endemic. On one end, subjecting all patients with suspected EBTB for FB will pose unnecessary risks to healthcare providers as FB is an aerosol-generating procedure with high risk of disease transmission during scope. On the other hand, empirical pharmacological treatment of all suspected patients with EBTB will inadvertently lead to delay in attaining competing differential diagnosis especially lung carcinoma in some cases. With the aid of FB, different samples including bronchial washing, BAL, bronchial brushing, biopsy (forceps, or ultrasound-guided endobronchial biopsy) can be performed with a good safety profile. We advocate performing FB in selected patients to rule out EBTB only when initial noninvasive workup in the form of sputum tests for AFB and imaging were inconclusive.
At both of our centers, bronchial washing, BAL, bronchial brushing, and biopsy samples were routinely tested for AFB smears and sent for Mycobacterium tuberculosis cultures. Due to cost-related issues, nucleic acid amplification tests such as Gene Xpert MTB/RIF were only offered in selected cases with a high index of suspicion for multidrug-resistant tuberculosis (MDR-TB), for example, the previous history of MDR-TB, recurrent pulmonary tuberculosis, and history of close contact with patients with MDR-TB. Zhang et al. in a recent publication on the role of GeneXpert MTB/RIF for rapid diagnosis and rifampicin resistance detection of endobronchial tuberculosis demonstrated that diagnostic yields of Gene Xpert MTB/RIF in bronchial brushings were significantly higher than sputum and brushing smears for AFB. Furthermore, among EBTB patients who were AFB negative by both sputum smear and brushings smear, GeneXpert MTB/RIF analysis of bronchial brushings detected positivity in 38.5% of patients. Despite higher costs, GeneXpert MTB/RIF appears to be an attractive option as it enables a more sensitive and rapid diagnosis of EBTB together with information of rifampicin resistance, which is crucial for timely management of tuberculosis.There are some limitations to our study. First, some patient data might be lost as this is a retrospective study. Further sub-analysis of various endobronchial subtypes might not be representative of the actual population in view of two main reasons. First, a significant proportion of our patients (40%) had more than one endobronchial subtype. Besides, a relatively small sample size means that we were not able to employ more complex statistical methods for comparison and analysis of various endobronchial subtypes.
| Conclusion|| |
We hope that our study demonstrated that FB in selected patients is safe for both patients and healthcare providers and is an effective option for the diagnosis of EBTB. It enhances treatment confidence of tuberculosis by providing microbiological and/or histological evidence thereby ruling out other differentials when initial workup for tuberculosis was inconclusive. Finally, we believe that a “middle ground” practice of subjecting patients with suspected EBTB for FB only when initial sputum smears for AFB were negative is of the best interest to both healthcare workers and patients.
Financial support and sponsorship
Conflicts of interest
The authors reported no conflicts of interest.
| References|| |
Kashyap S, Solanki A. Challenges in endobronchial tuberculosis: from diagnosis to management. Pulm Med 2014;2014:594806.
Lee JH, Park SS, Lee DH, Shin DH, Yang SC, Yoo BM. Endobronchial tuberculosis: clinical and bronchoscopic features in 121 cases. Chest 1992;102:990-4.
Kim YH, Kim HT, Lee KS, Uh ST, Cung YT, Park CS. Serial fiberoptic bronchoscopic observations of endobronchial tuberculosis before and early after antituberculous chemotherapy. Chest 1993;103:673-7.
Chung HS, Lee JS. Bronchoscopic assessment of the evolution of endobronchial tuberculosis. Chest 2000;117:385-92.
Yu W, Rong Z. Clinical analysis of 90 cases with endobronchial tuberculosis. Zhonghua Jie He He Hu Xi Za Zhi 1999;22:396-8.
Aggarwal AN, Gupta D, Joshi K, Behera D, Jindal SK. Endobronchial involvement in tuberculosis: a report of 24 cases diagnosed by flexible bronchoscopy. J Bronchol 1999;6:247-50.
Ozkaya S, Bilgin S, Findik S, K̈ok HC, Yuksel C, Atici GA, Endobronchial tuberculosis: histopathological subsets and microbiological results. Multidiscip Respir Med 2012;7:34.
Zhang Q, Zhang Q, Sun B et al.
GeneXpert MTB/RIF for rapid diagnosis and rifampin resistance detection of endobronchial tuberculosis. Respirology 2018;23:950-5.
[Figure 1], [Figure 2]
[Table 1], [Table 2]