Differences in pulmonary cavity features among drug-sensitive pulmonary tuberculosis and multidrug/extensively-resistant pulmonary tuberculosis: a multi-national multi-center computed tomography-based study

Introduction

Tuberculosis (TB) remains a major infectious threat to global health, affecting about 10.7 million people which caused 1.2 million people died in 2024, making it the world’s leading cause of death from a single infectious agent (Mycobacterium TB) (1). The emergence of drug-resistant TB (DR) increases the burden of TB control. Multidrug-resistant TB (MDR) refers to TB infection resistant to at least two first-line anti-TB drugs, isoniazid and rifampicin. Extensively drug-resistant TB (XDR) is defined as TB that has evolved resistance to rifampin and isoniazid, as well as to any members of the quinolone family and at least one of the second-line injectable drugs: kanamycin, amikacin, and capreomycin. Of all MDRs, XDR accounts for 4–20% of these infections (1-3). When resistant mutants arise during treatment with anti-TB drugs, it is considered as acquired resistance (previously treated MDR-TB). Patients infected with an already drug-resistant strain develop primary resistance (new MDR-TB), which is observed in newly diagnosed TB patients. It has been estimated that globally 3.5% (which can be much higher in some regions) of newly diagnosed TB patients, and 20.5% of previously treated patients, are MDR-TB (1,4). There have been interests to use chest imaging as a supporting tool to suggest the diagnosis of DR (5). A number of published articles described the potential chest imaging feature differences between drug-sensitive TB (DS) and DR (5-18). It has been suggested that MDR cases tend to have more extensive disease (5). XDR overall appears even more aggressive than MDR, with a greater number of cavities, larger cavities, and cavities of thicker wall (5,15).

However, the reported radiological feature differences between DS and MDR are likely having been confounded by that MDR cases tend to have a longer history prior to being diagnosed as MDR. In a computed tomography (CT) study conducted in China, Li et al. (19) studied 212 patients with MDR and 180 patients with DS. Previously treated patients accounted for 75.9% of the MDR cases and 35.0% of the DS cases. The duration of previous anti-TB treatment (months) was 8.0±12.0 for MDR cases and 1.0±2.0 for DS cases. In a chest radiograph study in Indonesia, Zuhriyyah et al. (20) compared the chest radiograph findings of children (<18 years old) with 38 DS patients and 31 DR patients [rifampicin-resistant (RR) =6, MDR =20, pre-XDR =4, XDR =1]. Calcification, a sign for chronicity, was observed in 23% (7/31) of DR patients, while in only 5% (2/38) of DS patients. Fibrosis, also a sign for chronicity, was observed in 42% (13/31) of DR patients, while in only 13% (5/38) of DS patients. In a CT study conducted in Indonesia with 36 DS and 34 MDR patients, Messah et al. (21) reported that, bronchiectasis, a sign of chronicity, was more common in MDR patients (88.2%) than in DS patients (58.3%). Fibrosis distribution was also wider among MDR patients than among DS patients. In a CT study conducted in Korea, Shin et al. (22) investigated 90 patients with new MDR patients and 90 age- and sex-matched patients with new DS. Fibrotic scar was noted in 46.7% of MDR patients and in 36.7% of DS patients. Bronchiectasis was noted in 32.2% of MDR patients and in 20% of DS patients. In a CT study conducted in Iran, Mehrian et al. (23) compared CT findings of MDR (n=28) and XDR (n=17) patients. Patients with XDR had more parenchymal calcification (64.7% vs. 28.6%). In a study on MDR cases in India, Jain et al. (24) described that, with the introduction of molecular diagnostic tools for the upfront diagnosis of all TB cases, there was a substantial reduction in the median time from the onset of symptoms to diagnosis between 2015 and 2020. In 2015, a higher frequency of cases exhibited cavitations, bronchiectasis, and fibrosis on chest radiograph compared to the findings in 2020. A higher occurrence of significantly advanced cases was noted in 2015 in contrast to 2020. The mean cavity size in 2015 measured 6.73 cm, while in 2020 it averaged 4.06 cm. Bronchiectasis was noted in 90% of the 2015 cohort, and in 64.3% of the 2020 cohort. All these reports described in this paragraph suggest that the more extensive lesions seen in MDR patients might be at least partly due to the fact that they had a longer TB disease history. Most existing literature cannot clearly establish whether MDR patients intrinsically have more extensive pathologies, or the higher prevalence and extent of lung lesions in MDR patients are purely due to that MDR cases are associated with a longer disease history and less responsive to first line anti-TB treatment.

We tried to investigate whether DR TB is intrinsically ‘more aggressive’ than DS TB. In the ‘Dalian study’ conducted in 2022 (25), we analysed the lung CT feature differences of DS vs. DR patients from a well-defined urban region in Dalian, China. There were 33 consecutive new DR cases (inclusive of RR and MDR cases), with 19 cases having a history of <1 month and 8 and 6 cases having a history of 1–6 and >6 months respectively. To pair the MDR cases according to the disease history length, disease history length matched 33 DS patients were included. The first CT exams prior to treatment were analyzed. In the ‘Guangzhou study’ conducted in China in 2024 (26), we retrieved CT data of consecutive new DR cases (n=46, inclusive of RR and MDR cases), and according to the electronic case archiving system records, the TB history was ≤3 months till the first CT scan was taken. To pair the MDR-pulmonary TB (PTB) cases with assumed equal disease history length, we retrieved data of 46 DS patients. The ‘Dalian study’ and the ‘Guangzhou study’ showed higher prevalence and wider distribution of lesions of pulmonary nodular (PN) consolidation and pulmonary cavity (PC) among DR patients than among DS patients, even though DR patients and DS patients had similar TB history length (25,26). Most recently, we reported an analysis utilizing the TB patient data in the NIAID (National Institute of Allergy & Infectious Diseases, USA) TB Portals Program (TBPP, https://tbportals.niaid.nih.gov/) dataset (27), and added 135 TB patients from Shenzhen, China (28). Calcified lesions in the lungs, as a sign of chronicity, were recorded. In new patients, there was no difference in lung lesion calcification prevalence among DS and MDR. In previously treated patients, lung calcification prevalence was higher among MDR and XDR patients. It was noted that, (even) for new patients, the PN prevalence was higher for MDR/XDR cases than for DS cases (around 70% vs. around 39%). For new patients, the mean PN number for positive cases was DS: 2.38, MDR: 2.89, XDR: 2.72. The number of lung fields with PN lesion was higher for MDR cases than for DS cases. PN lesions were even more widely spread in XDR cases than in MDR cases, even for new patients (28). Our additional analysis of literature suggests that a trend may exist that the frequency of lung lesions is: DS < RR < MDR < XDR, even when the TB history length is similar between DR and DS patients (28). We thus conclude that DR TB are intrinsically ‘more aggressive’ than DS TB. The more extensive lesions commonly seen among DR patients might be due to a combination of DR TB being intrinsically ‘more aggressive’ and MDR cases being commonly associated with a longer TB disease history and less responsive to first line anti-TB treatment.

Following the of emergence these new evidences, and utilizing the same NIAID (USA) TB Portals Program data and the 135 Chinese TB patients as described in our recent study (28), the current study investigates the PC prevalence and distribution pattern among DS, MDR, and DXR patients, and studies how PC pattern can suggest the diagnosis of DR TB.

Methods

Data resource

This is a retrospective analysis of available convenience patient data. We used the same patient data as our recent report (28). Eastern European TB patient data were from the NIAID (USA) TB Portals Program (TBPP, https://tbportals.niaid.nih.gov/) dataset (all patients registered in the database before January 2019 and with lung CT data). In this article, Eastern European countries refer to Belarus, Moldova, Romania, Azerbaijan, and Georgia. Chinese patients were from The Third People’s Hospital of Shenzhen and the Shenzhen Center for Chronic Disease Control, Shenzhen, China, treated between April 2017 and February 2019. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by ethics boards of The Third People’s Hospital of Shenzhen and the Shenzhen Center for Chronic Disease Control. Informed consent was waived due to the retrospective nature of this study. We excluded patients with unsatisfactory images or without drug susceptibility test (DST). We also excluded patients with human immunodeficiency virus (HIV) (+) status. Finally, 743 eligible patients (489 men and 254 women; mean age, 40.0±14.8 years) were included. According to the history of the previous anti-TB treatment, the patients were categorized into new cases (446 cases) and previously treated cases (297 cases). The Chinese patients were all new cases. For new patients, the interval days between the first CT chest scan and anti-TB treatment starting date were recorded, and the new patients were further classified into three groups: (I) the first CT was done before anti-TB treatment; (II) the first CT was done after anti-TB treatment started but the interval between CT scan and treatment initiation was ≤14 days; (III) the first CT was done after ant-TB treatment started and the interval was >14 days. The patient enrolment process and patient baseline are shown in Figure 1 and Table 1.

Figure 1 PTB patient data utilized in the current study [re-used with permission from Tang et al. (28)]. Pre-treated: previously treated. tr: anti-TB treatment. tr-CT: the interval time in days (d) between the anti-TB treatment initiation and the first CT scan was acquired. CT before tr: first CT scan was acquired before the anti-TB treatment started. CT after tr: first CT scan was acquired after the anti-TB treatment started. DS, drug-sensitive; DST, drug susceptibility test; HIV, human immunodeficiency virus; MDR, multidrug-resistant; pts, patients; PTB, pulmonary tuberculosis; TB, tuberculosis; XDR, extensively drug resistant.

Chest CT reading

For all the patients, the first CT scan images were used and read in consensus by a radiology trainee (S.N.T.) and a specialist radiologist (Y.X.J.W.). This study focused on the CT feature of PC. As described earlier (25,26), a PC was a gas-filled space, seen as a lucency area within pulmonary consolidation or a nodule. PC was counted only for those with a lumen diameter >5 mm. Multiple cavities in a single consolidation was counted as one cavity. Moth-eaten cavities are characterized by multiple, small, irregularly shaped cavities within an area of lung consolidation. Moth-eaten cavities are more likely to be numerous, these usually small cavities in one consolidation were together counted as one cavity. A PC with a PN was counted as both one PC and one PN. PC was also differentiated from bulla and cyst with a thin wall. The number was counted for each patient. The diameter for each PC was measured on axial CT images showing the largest size. When a nodule and a cavity coexist, we consider it to be a thick-walled cavity when the cavity accounts for more than 60 per cent of the volume of the nodule; otherwise, we consider it a nodule combined with cavity. Thick-walled cavities and thin-walled cavities are distinguished based on a boundary of 3 mm (wall thickness measurement is based on the thickest wall of the cavity). A non-walled cavity was a gas-filled space often in a large area of opacification lacking a visible surrounding wall. Examples of the PC are shown in Figures 2,3. Location of each PC was also recorded based on the natural anatomy of lung lobes. In addition to the comparison of PC features among DS, MDR, and XDR patients, a comparison was made between DS patients from China and DS patients from Eastern Europe.

Figure 2 Examples of PC morphology classification in tuberculosis patients (A-G). Thick-walled cavity (A,E,F), thin-walled cavity (B,G), moth-eaten cavity (C), no wall cavity (D). PC, pulmonary cavities.

Figure 3 Examples of PC in tuberculosis patients with maximum diameter labeled. (A-H) Cavities of various sizes. PC, pulmonary cavities.

Cases with apparent calcified lesions in the lungs were also recorded in our earlier analysis (28) and re-used in the current study.

Statistical analysis

Data analysis was processed using GraphPad Software (GraphPad Software Inc., San Diego, CA, USA). Categorical and continuous variables were analyzed by the Chi-squared test and Mann-Whitney U test, respectively. Receiver operating characteristic (ROC) curve analysis was used to determine the diagnostic performance, reporting the area under the ROC and optimal cut-off values with sensitivity and specificity.

Results

As reported earlier (28), for new patients, lung calcification prevalence on lung CT was slightly and statistically non-significantly more common among Eastern European cases than among Chinese patients (16.2% vs. 13.9%), suggesting that Eastern European cases could have a slightly longer disease history. For DS patients, the PC prevalence and PC number per positive case were slightly and statistically non-significantly higher among Chinese patients than among Eastern European patients (Figure 4). Due to the statistical insignificance, Chinese patients and Eastern European cases were grouped together for further analysis.

Figure 4 DS patient comparison of Eastern European cases and Chinese cases for PC prevalence according to PC number per positive case (A) and the lung calcification prevalence (B). All the patients were new patients (Chinese patients did not have previously treated patients). The PC prevalence and PC number per positive cases were slightly and statistically non-significantly higher among Chinese patients than among Eastern European patients (A). (B) is re-used with permission from Tang et al. (28), showing with the prevalence of lung calcification prevalence is slightly higher among Eastern European patients which suggests that the Eastern European patients might have had a longer disease history. Overall, this graph shows the CT data of Eastern European patients and Chinese patients are broadly comparable. P_a=0.62, P_b=0.74, P_c=0.21. CT, computed tomography; DS, drug-sensitive; PC, pulmonary cavities.

For the new DS patients, the PC prevalence and mean PC number per PC positive case were almost the same for cases without anti-TB treatment, cases with ≤14 days anti-TB treatment, and all new DS cases, being between 24% and 25% for prevalence and around 1.65 for mean PC number (Table 2, Figure 5). However, cases without anti-TB treatment had smaller cavity diameter, probably suggesting patients without anti-TB treatment had a shorter TB history. For the new DR cases, the PC prevalence was 27.13% for cases without anti-TB treatment, 29.66% cases with ≤14 days of anti-TB treatment, and 59% for cases with >14 days of anti-TB treatment, suggesting patients without anti-TB treatment had a shorter TB history and administered anti-TB treatment was not effective for the treated patients.

Figure 5 A comparison for new patients with first CT image taken (I) before anti-TB treatment, (II) with anti-TB treatment ≤14 days (≤14 d pts); (III) all-inclusive new patients (CT scanned before tr: n=301, tr began ≤14 days: n=29, tr >14 days: n=116). (A) Cavity number per cavity-positive patient; (B) cavity size per cavity-positive patient. For the new DS patients, the PC prevalence and mean PC number per PN positive case were almost the same for cases without anti-TB treatment, cases with ≤14 days anti-TB treatment, and all new DS cases. For the new DR cases, cases had anti-TB treatment during the current course had higher cavity number and slightly larger diameter. CT, computed tomography; DR, drug-resistant tuberculosis; DS, drug-sensitive tuberculosis; MDR, multidrug-resistant tuberculosis; No., number; PC, pulmonary cavities; PN, pulmonary nodular consolidation; pts, patients; TB, tuberculosis; tr, anti-TB treatment; XDR, extensively drug-resistant tuberculosis.

The prevalence of PC for DS and DR cases are shown in Table 3 and Figure 6. The PC prevalence was higher among DR cases than among DS cases, both for new patients and previously treated patients. Previously treated patients had a higher PC prevalence than new patients. New DS patients had the lowest PC prevalence of around 25%, while previously treated XDR patients had the highest PC prevalence of around 71%.

Figure 6 Prevalence of PC among DS and among DR patients. (A) New cases and pre-treated cases presented separately. (B) New cases and pre-treated cases presented together. The prevalence of PC was higher for previously treated cases than for new cases. DR cases had a higher PC prevalence than DS cases. DR, drug-resistant tuberculosis; DS, drug-sensitive tuberculosis; MDR, multidrug-resistant tuberculosis; PC, pulmonary cavities; XDR, extensively drug-resistant tuberculosis.

The median number, mean number of PC for DS and DR cases are shown in Table 3 and Figure 7. The mean and median PC numbers were higher among DR cases only among new cases. For PC positive cases, there was no difference in mean and median PC numbers for DS previously treated cases and DR previously treated cases. For PC positive cases, the median PC number among DS new patients was 1, while the median PC number among other categories of patients was all 2. The maximum PC number was 7 among DS cases, and PC number ≥8 was only seen among DR patients.

Figure 7 Number of PC per PC positive case among DS and among DR patients [(A-C) median and scatter plot; (D-F) mean and 95% confidence interval]. The mean and median PC numbers were higher among DR cases only among new cases. There was no difference in mean and median PC numbers for DS previously treated cases and DR previously treated cases. There were great variations in PC numbers for each group. “Total cases” means both new patients and previously treated (pre-treated) cases. DR, drug-resistant tuberculosis; DS, drug-sensitive tuberculosis; MDR, multidrug-resistant tuberculosis; PC, pulmonary cavities; XDR, extensively drug-resistant tuberculosis.

The mean diameter and maximum diameter of PC for DS and DR cases are shown in Table 3, Figures 8,9. For both mean diameter and maximum diameter, there was a trend that: DS < MDR < XDR. For previously treated patients, calcification was more prevalent in DR cases (DS =36.36%, MDR =44.87%, XDR =45.38%, Table 3), suggesting that the severity of PC lesion (i.e., cavity size) could be partially explained by that the DR patients had a longer disease history. The maximum PC diameter was mostly <70 mm, and a maximum PC diameter ≥75 mm was only seen among DR patients.

Figure 8 Mean and maximum diameter of PC among DS and DR patients [(A-F) mean diameter, (G-L) maximum diameter; (A-C,G-I) mean and 95% confidence interval; (D-F,J-L) median and scatter plot]. For both mean diameter and maximum diameter, there is a trend that: DS < MDR < XDR. However, there were great variations in PC diameters for each group. “Total cases” means both new patients and previously treated (pre-treated) cases. DR, drug-resistant tuberculosis; DS, drug-sensitive tuberculosis; MDR, multidrug-resistant tuberculosis; PC, pulmonary cavities; XDR, extensively drug-resistant tuberculosis.

Figure 9 Prevalence of PC of certain sizes in DS, MDR, and XDR patients. X-axis: PC diameter in mm. Y-axis: percentage prevalence of PC number of ≥1 (A1-C1), ≥2 (A2-C2), ≥3 (A3-C3), and ≥4 (A4-C4). (A1-A4) New cases. (B1-B4) Previously treated cases. (C1-C4) All cases. “All cases” means both new patients and previously treated (pre-treated) cases. DS, drug-sensitive tuberculosis; MDR, multidrug-resistant tuberculosis; PC, pulmonary cavity; XDR, extensively drug-resistant tuberculosis.

Differences in PC patterns between DS, MDR, and XDR patients are shown in Table 4 and Figure 10. Thick-walled PC was the most common pattern for all DS, MDR, and XDR cases, followed by thin-walled PC. In proportion, thin-walled PC proportion decreased from DS to MDR, and to XDR, while thick-walled PC proportion increased from DS to MDR, and to XDR.

Figure 10 Proportional difference in PC patterns between DS, MDR, and XDR cases. (A) New cases; (B) pre-treated cases; (C) all cases. n: total number of cavities assessed. “All cases” means both new patients and previously treated (pre-treated) cases. DS, drug-sensitive tuberculosis; MDR, multidrug-resistant tuberculosis; PC, pulmonary cavity; XDR, extensively drug-resistant tuberculosis.

The lung field distribution of PC is shown in Table 5. For both DS and DR patients, PC was most commonly seen in the upper right lobe. In relative terms, PCs were more commonly seen in the lower lobes for DR cases than for DS cases. PCs were more commonly seen in the left upper lobe for DR cases than for DS cases. The number of lung fields with PC lesion was higher for DR cases than for DS cases. DR cases were more likely to have bilateral PC lesions.

Taking together the PN results from our earlier report (28), the correlation between PC and PN is shown in Figure 11. For DS patients, the Pearson r for correlation between PC and PN was 0.03, not statistically significant. For DR patients, the Pearson r for correlation between PC and PN was 0.15 (P=0.001).

Figure 11 Correlation between PN and PC numbers in DS and DR (MDR + XDR) tuberculosis patients (both new patients and previously treated patients). For DS patients (n=244), the Pearson r for correlation between PC and PN is 0.03, not statistically significant. For DR patients (n=499), the Pearson r for correlation between PC and PN is 0.15 (P=0.001). Note that one dot in this graph may represent multiple patients. DR, drug-resistant tuberculosis; DS, drug-sensitive tuberculosis; MDR, multidrug-resistant tuberculosis; PC, pulmonary cavity; PN, pulmonary nodular consolidation; XDR, extensively drug-resistant tuberculosis.

Based on the data in this article and the PN results from our earlier report (28), the area under the receiver operating characteristic curve (AUROC) are shown in Figure 12, the ROC analysis results for using PC numbers to suggest a diagnosis for DR are shown in Table 6. For new patients, PC number ≥2 had a specificity of around 92.3% and a sensitivity of 21.9% suggesting the diagnosis of MDR/XDR. For previously treated, PC number ≥2 had a specificity of around 81.0% and a sensitivity of 33.0% suggesting the diagnosis of MDR/XDR.

Figure 12 Singular and multi-factorial ROC analyses for predicting drug-resistant tuberculosis based on imaging characteristics and treatment history. AUROC for PC num (A1-C1), PN num (A2-C2), and a combination of PC num and PN num (A3-C3), as well as a combination of PC num, PN num, and previous TB treatment history (C4). (A1-A3) are for new patients. (B1-B3) are for previously treated patients. (C1-C4) are for all patients. “All cases” means both new patients and previously treated (pre-treated) cases. The probability (P) of a patient being DR is given according to: (A1) ; (A2) ; (A3) ; (B1) ; (B2) ; (B3) ; (C1) ; (C2) ; (C3) ; (C4) . New patient, treatment =0; previously treated patient, treatment =1. AUC, area under the curve; AUROC, area under the receiver operating characteristic curve; DR, drug-resistant tuberculosis; num, number; PC, pulmonary cavities; PN, pulmonary nodular consolidation; ROC, receiver operating characteristic; TB, tuberculosis; trt his, treatment history.

Discussion

This study quantified lung PC prevalence, size, pattern, and distribution among DS and MDR/XDR patients. For our new patients, there was no difference in lung lesion calcification (a sign of chronicity) prevalence among DS (13.51%), MDR (14.36%), and XDR (13.89%) patients. In previously treated patients, lung calcification prevalence was 36.36% for DS, 44.87% for MDR, and 45.38% for XDR. Thus, for previously treated cases, DR cases likely had a longer disease history than DS cases. For new patients, the PC prevalence was around 25% for DS cases, and around 40% for DR cases. For PC positive new patients, the mean PC number was 1.66 for DS cases, and 2.69 for MDR cases; and mean diameter cavity was 15.4, 16.9, and 17.5 mm, for DS, MRD, and XDR, respectively. The number of lung fields with PC lesion was higher for MDR cases than for DS cases (1.69 vs. 1.36). DR cases were more likely to have bilateral PC lesions (30.5% vs. 19.7%). For both DS and DR patients, PCs were most commonly seen in the right upper lobe. In relative terms, PCs were more commonly seen in the lower lobes for DR cases than for DS cases. It is interesting to note that, relatively, both PNs and PCs were more commonly seen in the left upper lobe for DR cases than for DS cases (MDR PN 46.6%, MDR PC 54.6%, DS PN 36%, DS PC 39.4%) (28). These results further support that, despite the complication that the MDR patients we see in clinics tend to have a longer disease history, DR patients intrinsically have more extensive lung lesions than DS patients.

This study shows the proportion of thick-wall cavities was higher among DR patients, and there was a trend for this that: DS < MDR < DXR. The portion of thick-walled cavities was 69% among XDR patients (Table 4). PTBs with cavitary lesion are also associated with the development of XDR-TB during MDR-TB treatment (29,30). The thick lining of the cavity reduces the amount of drug that can penetrate from the bloodstream, thus it is likely there is an increased bacillary burden within cavitary lesions, in which the likelihood of spontaneous mutations associated with drug resistance is greater, and/or the existence of subpopulations of bacilli that survive either due to metabolic dormancy or exposure to sub-inhibitory drug concentrations. If there are cavities which are very thick-walled or too big which may inhibit anti-TB drug penetration, or no response of cavities to treatment, then individualized treatment such as aggressive medication or even surgery, rather than standard regimens, may be considered (31).

This study shows that, for new patients, there was no significant correlation between PC count and PN count in this study. We have earlier showed that PN is an earlier sign for lung TB, while it takes time for PC to develop (25). For DR patients, correlation between PC count and PN count in this study was weak yet statistically significant (Pearson r=0.15, P=0.001). This is understandable since DR patients had a higher prevalence and a higher count for both PN and PC, while DS patients had lower prevalence and lower counts for both PN and PC. It is anticipated that a longer history will be associated with both higher prevalence and higher counts for both PN and PC.

The general perception of many practicing radiologists is that it is impossible to differentiate MDR/XDR from DS based on chest imaging. This can be partially explained by that, the PC pattern differences are not ‘substantial’ between DS and MDR patients. For example, in this study, 52% of the new MDR patients did not have PC, and 74% of the new DS patients did not have PC. For PC positive new cases, PC number ranged from 1 to 7 (median: 1) for DS patients, while ranged from 1 to 15 for MDR cases (median: 2). Based on the data in this article, for new patients, PC number ≥2 had a specificity of around 92.3% and a sensitivity of 21.9% suggesting the diagnosis of MDR/XDR. For previously treated patients, PC number ≥2 had a specificity of around 81.0% and a sensitivity of 33.0% suggesting the diagnosis of MDR/XDR. We compared the prevalence and number of PC for the data in this study and other earlier literature reports (Tables 7,8). The prevalence of PC and the PC number per positive cases were comparatively low in this study. It is possible that the patients in the earlier results had a longer TB disease history. For Guangzhou study with new patients (26), when cutoff PC number was ≥4, specificity was 93% and sensitivity was 36% for diagnosing MDR patients. For the Dalian study with new patients (25), when cutoff PC number was ≥4, specificity was 84.9% and sensitivity was 39.4% for diagnosing MDR patients (if PC number ≥3, specificity: 72.7%, sensitivity: 54.6%). Taking these results together, PC number ≥3 may offer a reasonable specificity for suggesting the diagnosis of MDR, though the corresponding sensitivity would be low. Furthermore, based on the data in this study, the maximum PC number was 7 among DS cases, and PC number ≥8 was only seen among DR patients. The maximum PC diameter was mostly <70 mm, and maximum PC diameter ≥75 mm was only seen among DR patients.

The strength of this study is that the data were from multi-national multi-centers with large sample size, while there are a number of limitations to this study. This is a retrospective study of convenience sampling, and the included patients were unevenly distributed among countries. Chinese patients were dominantly DS cases (80%, 108/135), and East European patients were dominantly from Belarus (76%, 456/608) and Romania 16.9% (103/608). Prevalence of lung calcification prevalence was slightly higher among Eastern European patients than among Chinese patients (16.2% vs. 13.9%, P=0.62), which suggests that the Eastern European patients might have had a longer disease history. Another limitation of the current study is that it was not possible for us to precisely quantify the disease history length for each patient. On the other hand, it is also possible that our sample may represent the real-world scenario that we might encounter. Disease history length is a subjective measure by patients themselves. Patients from rural/remote areas might only present to the hospital until their discomforts reach a certain degree of severity, or their discomforts have protracted for a long period of time. This study applied calcification as a ‘biomarker’ as the comparative ‘chronicity’ of the TB disease history. Though this is not a reliable biomarker, since the same criterion was applied to different groups of the patients, this sign may allow a reasonable intra-study comparison.

Conclusions

In conclusion, this multi-national multi-center study further supports the earlier observation that lesion prevalence and lesion extent are generally higher among MDR/XDR patients than among DS patients. In this study, both for DS and DR patients, PC prevalence was increased in previously treated patients than in new patients. When PC number is ≥3, a possible diagnosis of DR could be suggested with a high specificity, though the associated sensitivity would be low. This study further supports the notion that we should consider patient disease history length when analyzing the chest CT features of TB patients. Following this logic, a new patient with a short TB disease history but showing ≥2 PCs should also raise the suspicion that this patient being a DR case.

Acknowledgments

None.

Footnote

Data Sharing Statement: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-aw-2331/dss

Peer Review File: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-aw-2331/prf

Funding: None.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-aw-2331/coif). Y.X.J.W. serves as an unpaid editorial board member of Journal of Thoracic Disease from April 2024 to June 2026. The other authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by ethics boards of The Third People’s Hospital of Shenzhen and the Shenzhen Center for Chronic Disease Control. Informed consent was waived due to the retrospective nature of this study.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. World Health Organization. Global tuberculosis report 2025. Geneva: World Health Organization; 2025. Available online: https://www.who.int/teams/global-programme-on-tuberculosis-and-lung-health/tb-reports/global-tuberculosis-report-2025
2. Raviglione MC, Smith IM. XDR tuberculosis--implications for global public health. N Engl J Med 2007;356:656-9. [Crossref] [PubMed]
3. Aziz MA, Wright A, Laszlo A, et al. Epidemiology of antituberculosis drug resistance (the Global Project on Anti-tuberculosis Drug Resistance Surveillance): an updated analysis. Lancet 2006;368:2142-54. [Crossref] [PubMed]
4. World Health Organization. Global tuberculosis report. 2014. Available online: https://apps.who.int/iris/handle/10665/137094
5. Wáng YXJ, Chung MJ, Skrahin A, et al. Radiological signs associated with pulmonary multi-drug resistant tuberculosis: an analysis of published evidences. Quant Imaging Med Surg 2018;8:161-73. [Crossref] [PubMed]
6. Xu CJ, Lu PX, Li CH, et al. Chinese expert consensus on imaging diagnosis of drug-resistant pulmonary tuberculosis. Quant Imaging Med Surg 2024;14:1039-60. [Crossref] [PubMed]
7. Huang XL, Skrahin A, Lu PX, et al. Prediction of multiple drug resistant pulmonary tuberculosis against drug sensitive pulmonary tuberculosis by CT nodular consolidation sign. bioRxiv preprint 2019. Available online: https://doi.org/10.1101/833954
8. Chuchottaworn C, Thanachartwet V, Sangsayunh P, et al. Risk Factors for Multidrug-Resistant Tuberculosis among Patients with Pulmonary Tuberculosis at the Central Chest Institute of Thailand. PLoS One 2015;10:e0139986. [Crossref] [PubMed]
9. Chung MJ, Lee KS, Koh WJ, et al. Drug-sensitive tuberculosis, multidrug-resistant tuberculosis, and nontuberculous mycobacterial pulmonary disease in nonAIDS adults: comparisons of thin-section CT findings. Eur Radiol 2006;16:1934-41. [Crossref] [PubMed]
10. Kim HC, Goo JM, Lee HJ, et al. Multidrug-resistant tuberculosis versus drug-sensitive tuberculosis in human immunodeficiency virus-negative patients: computed tomography features. J Comput Assist Tomogr 2004;28:366-71. [Crossref] [PubMed]
11. Kim SH, Min JH, Lee JY. Radiological Findings of Primary Multidrug-resistant Pulmonary Tuberculosis in HIV-seronegative Patients. Hong Kong J Radiol 2014;17:4-8.
12. Yeom JA, Jeong YJ, Jeon D, et al. Imaging findings of primary multidrug-resistant tuberculosis: a comparison with findings of drug-sensitive tuberculosis. J Comput Assist Tomogr 2009;33:956-60. [Crossref] [PubMed]
13. Kim W, Lee KS, Kim HS, et al. CT and microbiologic follow-up in primary multidrug-resistant pulmonary tuberculosis. Acta Radiol 2016;57:197-204. [Crossref] [PubMed]
14. Cha J, Lee HY, Lee KS, et al. Radiological findings of extensively drug-resistant pulmonary tuberculosis in non-AIDS adults: comparisons with findings of multidrug-resistant and drug-sensitive tuberculosis. Korean J Radiol 2009;10:207-16. [Crossref] [PubMed]
15. Cheon H. Comparison of CT findings of between MDR-TB and XDR-TB: A propensity score matching study. Imaging Med 2017;9:125-9.
16. Lee ES, Park CM, Goo JM, et al. Computed tomography features of extensively drug-resistant pulmonary tuberculosis in non-HIV-infected patients. J Comput Assist Tomogr 2010;34:559-63. [Crossref] [PubMed]
17. Fishman JE, Sais GJ, Schwartz DS, et al. Radiographic findings and patterns in multidrug-resistant tuberculosis. J Thorac Imaging 1998;13:65-71. [Crossref] [PubMed]
18. Sulistijawati RS, Icksan AG, Lolong DB, et al. Thoracic Radiography Characteristics of Drug Sensitive Tuberculosis and Multi Drug Resistant Tuberculosis: A Study of Indonesian National Tuberculosis Prevalence Survey. Acta Medica (Hradec Kralove) 2019;62:24-9. [Crossref] [PubMed]
19. Li CH, Fan X, Lv SX, et al. Clinical and Computed Tomography Features Associated with Multidrug-Resistant Pulmonary Tuberculosis: A Retrospective Study in China. Infect Drug Resist 2023;16:651-9. [Crossref] [PubMed]
20. Zuhriyyah SA, Nugraha HG, Setiabudi D, et al. Chest X-Ray Comparison Between Drug-Resistant and Drug-Sensitive Pulmonary Tuberculosis in Children. Clin Respir J 2024;18:e70010. [Crossref] [PubMed]
21. Messah ADV, Darmiati S, Rumende CM, et al. Correlation between Gene polymorphism levels of serum matrix metalloproteinases with cavitary features and pulmonary fibrosis of the Patient tuberculosis multi-drug resistance using high-resolution computerized tomography of the Thorax. Heliyon 2024;10:e33671. [Crossref] [PubMed]
22. Shin HS, Choi DS, Na JB, et al. Low pectoralis muscle index, cavitary nodule or mass and segmental to lobar consolidation as predictors of primary multidrug-resistant tuberculosis: A comparison with primary drug sensitive tuberculosis. PLoS One 2020;15:e0239431. [Crossref] [PubMed]
23. Mehrian P, Farnia P, Jalalvand D, et al. Computerised tomography scan in multi-drug-resistant versus extensively drug-resistant tuberculosis. Pol J Radiol 2020;85:e39-44. [Crossref] [PubMed]
24. Jain S, Sarin R, Vinay V, et al. To investigate the impact of revised diagnostic algorithm on presentation of multidrug-resistant tuberculosis cases at a referral centre in India. J Family Med Prim Care 2024;13:4432-7. [Crossref] [PubMed]
25. Song QS, Zheng CJ, Wang KP, et al. Differences in pulmonary nodular consolidation and pulmonary cavity among drug-sensitive, rifampicin-resistant and multi-drug resistant tuberculosis patients: a computerized tomography study with history length matched cases. J Thorac Dis 2022;14:2522-31. [Crossref] [PubMed]
26. Fang WJ, Tang SN, Liang RY, et al. Differences in pulmonary nodular consolidation and pulmonary cavity among drug-sensitive, rifampicin-resistant and multi-drug resistant tuberculosis patients: the Guangzhou computerized tomography study. Quant Imaging Med Surg 2024;14:1010-21. [Crossref] [PubMed]
27. Gabrielian A, Engle E, Harris M, et al. TB DEPOT (Data Exploration Portal): A multi-domain tuberculosis data analysis resource. PLoS One 2019;14:e0217410. [Crossref] [PubMed]
28. Tang SN, Huang XL, Skrahina A, et al. Differences in pulmonary nodular consolidation features among drug-sensitive pulmonary tuberculosis and multidrug/extensively-resistant pulmonary tuberculosis: a multi-national multi-center study. J Thorac Dis 2025;17:7498-514. [Crossref] [PubMed]
29. Long R. Drug-resistant tuberculosis. CMAJ 2000;163:425-8.
30. Shin SS, Keshavjee S, Gelmanova IY, et al. Development of extensively drug-resistant tuberculosis during multidrug-resistant tuberculosis treatment. Am J Respir Crit Care Med 2010;182:426-32. [Crossref] [PubMed]
31. Bastos ML, Lan Z, Menzies D. An updated systematic review and meta-analysis for treatment of multidrug-resistant tuberculosis. Eur Respir J 2017;49:1600803. [Crossref] [PubMed]
32. Joshi AR, Mishra S, Sankhe AP, et al. HRCT Spectrum of Pulmonary Multidrug-Resistant Tuberculosis in HIV Negative Patients: A Study in Indian Population. International Journal of Science and Research 2017;6:596-600.
33. Zahirifard S, Amiri MV, Bakhshayesh Karam M, et al. The radiological spectrum of pulmonary multidrug-resistant tuberculosis in HIVnegative patients. Iran J Radiol 2003;1:161-6.
34. Zhang L, Pang Y, Yu X, et al. Risk factors for pulmonary cavitation in tuberculosis patients from China. Emerg Microbes Infect 2016;5:e110. [Crossref] [PubMed]