Assessment of Disease Activity and Complications in Patients of Pulmonary Tuberculosis by High Resolution Computed Tomography

Background: Tuberculosis (TB) is a global health problem and the second most common infectious cause of death. High-resolution computed tomography (HRCT) is far more superior to chest radiography as well as conventional CT for analyzing the pulmonary parenchyma. This study aimed to evaluate the role of HRCT in pulmonary tuberculosis (PTB) with respect to disease activity and complication after anti-tubercular therapy (ATT). Methods: This prospective observational study was conducted in the Department of Radiodiagnosis, Teerthanker Mahaveer Medical College & Research Centre (TMMC&RC) for a period of 1.5 years. A total of 50 cases of newly diagnosed TB were included in the study and a standard six-month ATT was given to the patients. Pulmonary involvement was evaluated by HRCT (128 slice multi-detector PHILIPS INGENUITY CT scanner), twice for each patient (first scan after diagnosis and second after treatment completion). The acquired HRCT images were reconstructed on a highresolution lung algorithm and parenchymal, bronchial, and extra parenchymal findings were recorded systematically. Results: Out of the 50 patients, 5 died within two months of the initiation of treatment and four were lost to follow-up. Thus, post treatment follow-up sample sizewas reduced to 41 patients. Ill-defined nodules (96%), tree-in-bud pattern (74%), consolidation (86%), cavitary lesions (98%), and ground glass opacities (58%) were the main imaging features of active cases of TB on HRCT. Resolution to thin-walled cavitary lesions (36.5%), bronchiectasis (41.5%), and fibrotic (parenchymal) bands (66%) were common complications or sequelae which were observed after completion of treatment. How to cite this article: Ashwani Jain, Ankur Malhotra, Deepti Arora, Mazher Maqusood, and Sunil Kumar (2021) “Assessment of Disease Activity and Complications in Patients of Pulmonary Tuberculosis by High Resolution Computed Tomography,” Sudan Journal of Medical Sciences, vol. 16, Issue no. 2, pages 159–177. DOI 10.18502/sjms.v16i2.9286 Page 159 Corresponding Author: Dr. Ankur Malhotra, Associate Professor; Department of Radiodiagnosis, Teerthanker Mahaveer Medical College & Research Centre (TMMC&RC), Moradabad, Uttar Pradesh244001, India. Contact no.: +918394097126 email: drankur.m7@gmail.com Received 12 March 2021 Accepted 18 June 2021 Published 30 June 2021 Production and Hosting by


Introduction
Tuberculosis (TB) is a global health problem and the second most common infectious cause of death [1]. It is a chronic granulomatous disease caused by mycobacterium tuberculosis (MTB) and characterized by caseating necrosis with remarkable susceptibility for calcification and fibrosis. Tuberculous involvement of lung may occur in variety of ways, for example, local spread, bronchogenic dissemination, or hematogenous route [2].
Timely diagnosis is of paramount importance in the control of TB and is the only efficient measure for stopping the transmission of disease [3]. A number of complication and sequelae can occur in cases of pulmonary tuberculosis (PTB) with or without treatment, and knowledge of radiological signs of these sequelae and complication is vital. Despite clinical recovery, the post-TB sequelae can continue to be identified on imaging which may raise concern regarding disease activity [4].
Chest radiography continues to be the front runner for early assessment of individuals with suspicion of PTB; although it may appear entirely normal or inconclusive even in active cases of PTB. Compared to the chest radiography, computed tomography (CT) is way better of diagnosing and analyzing vivid imaging features of PTB [6].
The overall diagnostic accuracy of chest radiography in PTB is only 49% (34% for primary and 59% for post-primary cases). Conventional CT, on other hand, gives correct diagnosis in nearly 91% PTB cases and can accurately differentiate between active-type (80%) and inactive-type (89%) diseases. The reported sensitivity and specificity of HRCT in smear positive cases of PTB approaches 90.9% and 96.4%, respectively [7,8].
With this background, the present study was undertaken to evaluate the role of HRCT in PTB with respect to disease activity and complication after anti-tubercular therapy (ATT).

Materials and Methods
This prospective observational and descriptive study was conducted in the Department of Radiodiagnosis of a tertiary care hospital in northern India for a period of 1.5 years ( January 2019-June 2020) after obtaining prior clearance from the Institutional Ethics Committee (letter no. TMMC&RC/IEC/18-20/079 dated December 27, 2018). Written informed consent was obtained from all patients or their attendants before enrolment.
A total of 50 cases of newly diagnosed smear/Cartridge-based Nucleic Acid Amplification Testing (CBNAAT)-positive PTB were included in the study. Patients with known malignancy, pregnancy, or prior history of ATT intake were excluded from the study. A detailed clinical history, complete physical examination, and routine laboratory investigations were done in all patients and findings were recorded.
All patients received a six-month standard ATT consisting of a two-month initial phase (rifampicin, isoniazid, pyrazinamide, and ethambutol) and a four-month maintenance phase (rifampicin and isoniazid). Pulmonary involvement was evaluated by HRCT which was done twice for each patient. The first scan was obtained after the diagnosis of TB and within 28 days of the initiation of treatment. The second scan was obtained within 30 days after the treatment completion when patient was considered cured as per the clinical criteria. Patients were imaged on 128 slice multi-detector CT scanner (PHILIPS INGENUITY) set to 0.6-mm collimation and a pitch of 1.5. The images were reconstructed with a 1-mm slice thickness in the axial plane using a high spatial frequency bone algorithm. Sagittal and coronal reconstructions were also obtained.
After the acquisition of images, they were systematically analyzed by an experienced radiologist (with eight years of experience in reporting HRCT lung) who was blinded to the clinical details. The various pulmonary, bronchial, and extrapulmonary findings were interpreted as follows: nodule was characterized as "A focal, round opacity, at least moderately well marginated and no greater than 3 cm in maximum diameter" [6]; "Tree-in-bud pattern was recognized by small centrilobular nodules of soft-tissue attenuation connected to multiple branching linear structures of similar caliber originating from a single stalk" [6]; "Homogeneously increased lung opacity with obscuration of underlying vessels" was interpreted as consolidation [6]; "A lucent area within a zone of pulmonary consolidation, or a nodule that may or may not contain a fluid level and that was surrounded by a wall, usually of varied thickness" was labelled as cavitation [9]; "Hazy opacity without obscuration of underlying broncho-vascular structures " was recorded as ground glass opacity [10]; "A linear opacity, 1-3 mm thick and up to 5 cm long which extends to the visceral pleura was considered as parenchymal band" [11]; "Bronchial dilatation (relative to accompanying pulmonary artery) with lack of bronchial tapering was interpreted as bronchiectasis" [11]. "Reduced volume accompanied by increased attenuation in the affected part of lung was considered collapse/atelectasis" [11]; "Homogenous opacification of costophrenic angle and hemidiaphragm was considered as pleural effusion" [12]; "Separation of visceral pleural line from the chest wall by a trans radiant zone devoid of vessels was recorded as pneumothorax" [13]; "Empyema was characterized as "non-dependent elliptical or lenticular shaped opacity having obtuse angle with chest wall with splitting of visceral and parietal pleural surfaces by fluid giving split pleural sign" [14].

Statistical analysis
Data were entered into the Microsoft excel and the statistical analysis was performed by SPSS (version 21.0). The quantitative variables (numerical variables) were presented in the form of mean and SD and the qualitative variables (categorical variables) were presented in the form of frequency and percentage. The Chi-square test was applied for comparing the categorical variables such as gender and adverse events between the two groups. The sensitivity, specificity, positive, and negative predictive values and accuracy were calculated. The p-value < 0.05 was considered to be significant.

Results
This prospective observational study consisted of 50 patients who were enrolled on the basis of the inclusion and exclusion criteria. Out of the 50 patients, 5 died within two months of the initiation of treatment while four were lost to follow-up. Thus, posttreatment follow-up sample size was reduced to 41 patients. Of these, five patients presented showing a relapse in treatment but they were all asymptomatic.
Ill-defined Nodules were found in 48 (96.0%) cases prior to the initiation of treatment without any zonal predominance. While in 25 out of the 41 patients, the nodules completely disappeared after treatment, 16 patients showed persistent residual nodules, although with reduced nodule burden (Table 1). Tree-in-bud appearance was seen in 37 (74.0%) cases before treatment. It persisted in six patients (14.6%) post-treatment ( Table 2), out of which, five were defaulters and were not taking treatment properly ( Figure 1). Prior to the treatment, 43 (86%) cases showed consolidation which persisted in 8 patients (19.5%) after treatment (Table 3). In three out of eight cases, the size of consolidation patch decreased after treatment while the rest of the five patients were simply defaulters ( Figure 2). Chi-square value = 10.908, p-value < < < 0.001*   Chi-square value = 23.944, p-value < < < 0.001* Chi-square value = 13.955, p-value < < < 0.001*   (Figure 3). Ground glass opacity was found in 21 (42%) pretreatment cases and persisted in 3 patients after treatment (Table 5). Collapse/volume loss was seen in one patient before treatment and three patients posttreatment. Pleural effusion was found in nine (18%) cases before treatment. It resolved completely in all nine patients. However, in one patient who had no sign of pleural effusion at the time of the initiation of therapy, pleural effusion was detected posttreatment (Figure 4). Lung cyst was found in seven (14%) cases before treatment and persisted in only one of them after treatment. Nine (18%) cases showed emphysematous changes before treatment which increased to eleven posttreatment. Majority of the patients showed paraseptal emphysematous changes.
Pretreatment, fibrotic band was found in eight (16%) cases, which increased to 27 cases after treatment (Table 6)  3 cases, respectively (Table 7). There was no case of pneumothorax in posttreatment follow-up. HRCT is a useful technique for imaging evaluation of both lung parenchyma and airways. Patients who are clinically suspicious for PTB but are sputum smear negative can be best assessed by HRCT [13]. HRCT can easily detect complications after the completion of treatment and can also predict risk of complications [6]. The present study was undertaken to identify the HRCT findings of sputum-positive cases of PTB and assess the role of HRCT in disease activity and complications.

Discussion
Majority of our patients belonged to the age group of 51-60 years (22%) followed by 41-50 years (18%), which matches with the previous studies [17,18]. The mean age

HRCT findings in pre-and posttreatment cases
Ill-defined nodules are mostly found in the initial stages of TB and appear 5-6 mm in diameter with fuzzy margins with a tendency to coalesce. These nodules are usually not detectable on chest radiograph and the presence of nodules on HRCT is an early sign of active TB [6]. Im et al. [21] stated that TB progresses primarily in respiratory bronchioles followed by focal inflammation, and that these pathological processes are reflected in form of nodules seen on HRCT thorax. Pathologically, these micronodules represent solid caseous material within or around the terminal or respiratory bronchioles [9]. In the present study, ill-defined nodules were seen in 96% of the patients before treatment and persisted in 39% of them posttreatment, although the burden of nodules appeared to be reduced. These findings are in accordance with Lee et al. [6] who found ill-defined nodules in 88% pretreatment cases and 56% posttreatment cases. Capone et al. [19] also reported that 86.4% of their patients showed nodules in active cases of TB which persisted in 48.6% patients after treatment.
Tree-in-bud pattern is most commonly due to the presence of caseation necrosis and granulomatous inflammation within and surrounding the terminal and respiratory bronchioles and alveolar ducts, and reflects endobronchial spread of TB [22]. Rossi et al. [23] also stated that tree-in-bud appearance occurs generally in cases with endobronchial spread of MTB and is found it to be an indicator of active TB. However, the opinion of Im et al. [21] differs and according to them, "tree-in-bud" pattern is not pathognomonic for active PTB as it can also be recognized in many other entities like infections (bacterial, viral, parasitic, and fungal), congenital disorder, inhalation of foreign body, and vascular diseases. In the present study, tree-in-bud configuration was detected in 74% of patients before treatment which persisted in 14.6% posttreatment.
However, five patients were defaulters who were not taking treatment properly. Our results nearly match with Lee et al. [6] who reported tree-in-bud sign in 87% patients before treatment and in none of the cases posttreatment. Bombarda et al. [4] also observed tree-in-budsign in 60% of their patients before treatment and 5% patients posttreatment.
In the present study, cavitary lesions were observed in 98% cases before treatment.
Of them, 96% patients showed cavitation with smooth internal border and thick external border, 12% showed thin-walled cavitary lesion, and 6% showed cavitation with air fluid level. These cavities completely disappeared in 43.9% patients posttreatment. About 36.5% cases showed persistent thin-walled cavitary lesion, while 26.8% (11 out of 41) patients had thick-walled cavities. Out of these 11 cases with persistent thick-walled cavities, 4 were defaulters and in 7 patients mural thickness was reduced. Our findings correlate with the previous work of Lee et al. [6] and Bombarda et al. [4]. Lee et al. [6] found cavitary lesion in 73% cases which persisted in 35% cases after treatment. In addition, 25% cases in the study by Bombarda et al. [4] showed persistent thin-walled cavitary lesions after treatment. Im et al. [21] also reported persist cavities in 42% cases.
The cavitary lesions are independently related with increased time for acid-fast smears to become negative in patients receiving ATT and their persistence increases risk of relapse after completion of treatment [24].
Ground glass opacity was noted in 42% of our study cases before treatment and 7.3% after treatment. These findings are consistent with the studies of Capone et al. and Lee et al. [6, 19,]. Capone et al. [19] noted ground glass opacities in 37.8% cases before treatment which persisted in 4% patients. In the study of Lee et al. [6] it persisted in 2% patients. However, these findings do not match with the Raj et al. [25] who reported lower percentage of ground glass opacities in their study (20%). Ground glass opacities on HRCT can result from trauma, inflammation, focal interstitial fibrosis, hemorrhage, and malignancy [10]. The higher proportion of patients with hemoptysis in this study could be the possible reason for this discrepancy.
Pleural effusion was seen in nine (18%) patients before treatment and it resolved completely in all these patients, although one patient was diagnosed with empyema after treatment who had no signs of pleural effusion before treatment. Our findings are in line with previous studies. Nachiappan et al. [26] found pleural effusion in only 25% patients with primary TB. Only 10% of the cases in the study by Lee et al. [6] showed pleural effusion before treatment which resolved after treatment in all the patients.
We found emphysematous changes in nine (18%) cases before treatment and eleven (26.8%) cases after treatment. Emphysematous changes were mainly paraseptal type.
These findings are similar to Im et al. [21] who found emphysema in 24% of pretreatment cases.
In this study, fibrosis was found in 8 (16%) pretreatment cases and this number rose to 27 (66%) after treatment. Our findings match with Lee et al. [6] who noted an increase in fibrosis in TB patients after treatment from 42 to 92%. Bombardo et al. [4] in their study also reported increase in fibrotic changes posttreatment from 10 to 70%. Healing of TB lesions is usually followed by development of a fibrotic scar and dystrophic calcification with loss of lung volume. Rarely, if a lesion heals without development of necrosis, no residual fibrotic changes may occur [27].
In the present study, bronchiectasis was found in two cases (4%) before treatment. It significantly increased to 17 cases (41%) after treatment. Our findings match with Bombarda et al. [4] who also found increase in bronchiectasis in posttreatment cases (35%).
Lee et al. [6] encountered increase in bronchiectasis from 29 to 44% in posttreatment cases. Im et al. [21] reported cicatricial changes in the areas of cavitation, which on follow-up CT scans were characterized by increase in fibrotic bands, emphysema, or bronchiectasis. Fibrosis and bronchiectasis are common sequelae of TB. While fibrosis occurs primarily due to scarring of pleura spaces, bronchiectasis occurs due to fibrosis and damage of the lung parenchyma with irreversible dilatation of secondary bronchus [28].
Collapse (volume loss) was seen in only one of our cases before treatment while after treatment it was observed in three patients. Bronchial wall thickening was seen in five cases after the treatment of PTB. Hatipoglu et al. [29] found bronchial wall thickening adjoining to areas of fibrotic changes in few of their patients with inactive disease.
Thus, the collapse and bronchial wall thickening represents sequelae to post-tubercular cicatrization.

HRCT imaging feature after completion of treatment
Resolution to thin-walled cavitary lesions (36.5%), bronchiectasis (41.5%), and fibrotic (parenchymal) bands (66%) were common complications or sequelae which were observed after completion of treatment in our patients. We also observed decrease in nodule burden, disappearance of consolidation, tree-in-bud pattern, pleural effusion, and parenchymal cyst after treatment which indicates effectiveness of the treatment.
These findings are in accordance with previous studies. Lee et al. [6] found increase in fibrotic bands in patients after treatment. Bombarda et al. [4] reported thin-walled cavities, fibrotic bands, and traction bronchiectasis in their patients posttreatment.
Skoura et al. [31] stated that the appearance of fibrosis and resolution to thin-walled smooth cavitary lesions are main features of inactive PTB. Our findings, however, differ from Khan et al. [28] as they reported aspergilloma, bronchogenic carcinoma, and tracheobronchial stenosis as post-TB sequelae. These are mostly rare and late complication of TB and require long-term follow-up. Also, most of these findings require contrast study for optimal evaluation which was not done in the present study.
Thus, this study exhibited that HRCT lung significantly aids in detection of parenchymal and airway lesions of PTB. It can accurately identify features of active TB and can easily differentiate active cases from inactive ones. HRCT not only permits assessment of treatment progress but also allows early diagnosis and management of posttreatment complications thereby reducing mortality and morbidity in these cases.

Limitations
The small sample size of the study may limit statistical relevance. Due to problem of attrition and high mortality in TB, the follow-up patient sample size was even lower.
Some of the patients relapsed after treatment which may have affected the results in posttreatment follow-up category. MDR (multi-drug resistance) and XDR (extensively drug resistant) TB cases were not included in the study and findings in those cases may differ. Some of the post-tubercular sequelae like aspergilloma, bronchogenic carcinoma, and tracheobronchial stenosis were not observed in this study due to short-term follow-up and lack of contrast imaging. Mediastinal imaging features of TB in pre-and posttreatment cases were not analyzed due to lack of contrast study.

Conclusion
HRCT thorax is a sensitive modality for evaluation of parenchymal and airway manifestations in cases of PTB and the most frequent pretreatment HRCT findings include Ill-defined nodules, tree-in-bud appearance, thick-walled cavitary lesions, consolidation, ground glass opacities, and pleural effusion. Ill-defined nodules, tree-in-bud pattern,