Indonesia, one of the countries with the highest TB burden in Southeast Asia, faces substantial challenges in terms of TB control and management. Along with Bangladesh and the Philippines, Indonesia accounted for 18% of global TB incidence in 2017, recording an estimated 969,000 new or relapsed TB cases in 2021.^16,17 Indonesia faces a substantial detection and notification gap—55% of estimated TB cases in 2021 were undiagnosed or unreported, higher than India (33%) and Bangladesh (18%) but lower than Myanmar (67%).¹⁷ Only 30% of infectious TB cases were detected, similar to the Philippines (22%) and far below Bangladesh’s near-total detection.¹⁶ Drug-resistant TB poses a major obstacle: the Southeast Asia (SEA) region accounted for 38% of the global multidrug-resistant (MDR)/RR-TB burden in 2021, and Indonesia is classified among the high-burden countries for MDR/RR-TB, struggling with one of the highest notification gaps regionally (70%).¹⁷ The Philippines, classified within the SEA/WPR region, was estimated to have 9000 MDR-TB cases in 2019, while Vietnam had 6000. Moreover, research has projected that the proportion of MDR-TB cases in the Philippines will increase to 8.9% by 2040.¹⁸ In terms of workplace TB policies, the Philippines leads the region with the most comprehensive coverage (13 of 17 WHO/International Labour Organization-recommended components), ahead of Myanmar, Malaysia, and Cambodia.¹⁹ Overall, Indonesia shares key challenges with the Philippines—high infectious TB prevalence and wide notification gaps—while also contending with one of the highest MDR-TB burdens in the region.¹⁷

The present study provides a comprehensive overview of the demographic, clinical, radiological, and molecular characteristics of patients with RR-TB treated at the Dr. Soetomo General Hospital, Surabaya. Most patients were male and within the productive age range (20-50 years), aligning with global epidemiological trends that associate this group with a higher TB risk due to increased social mobility, occupational exposure, and broader interaction networks.^20-22 The observed male predominance is consistent with previous findings and may be attributed to behavioral factors, such as smoking and treatment adherence,^20,23 although disparities in access to healthcare may place women at increased risk in specific settings.²⁴

While most patients had a normal BMI, cases of malnutrition (BMI <18.5 kg/m²) and anemia were notable. Both conditions are established risk factors for TB, and are associated with impaired immunity and poor outcomes. Anemia alone may increase TB susceptibility up to four-fold.²⁵ Nutritional interventions—including individualized dietary plans, micronutrient supplementation, and ongoing counseling—have demonstrated efficacy in improving treatment outcomes.

Interestingly, the majority of cases were newly diagnosed, even though previous TB treatment history remains a strong predictor of RR-TB, with an estimated five-fold increased risk.²³ This trend suggests that the transmission of resistant MTB strains is increasingly driven by community spread rather than acquired resistance, potentially facilitated by compensatory mutations that enhance bacterial fitness.^26-28 These findings emphasize the need for early drug-resistance screening, even in treatment-naןve patients.

Diabetes mellitus emerged as the most common comorbidity, consistent with its known role in increasing both TB-specific and all-cause mortality (OR 1.90 and 4.54, respectively), largely due to impaired immune responses and reduced therapeutic efficacy.²⁹ Although HIV co-infection was rare in this cohort, it remains clinically significant given its association with a more than 28-fold increase in TB mortality and substantial treatment complexity.

Most patients were managed using the WHO-recommended, short-course BPaLM regimen (bedaquiline, pretomanid, linezolid, and moxifloxacin), which has demonstrated high effectiveness and favorable adherence profiles in other high-burden settings.^30,31 The average treatment duration was 26 weeks, which was extended to 39 weeks based on clinical or radiological response(s). Adverse effects, primarily anemia, neuropathy, and gastrointestinal symptoms, were generally mild to moderate and managed without treatment discontinuation, underscoring the importance of structured pharmacovigilance to ensure adherence.

Laboratory data revealed that many smear-negative patients had low IS6110–IS1081 CT values, indicating a substantial bacillary load detectable only via molecular methods. This highlights the limitations of smear microscopy—particularly in RR-TB—and supports the superior sensitivity of digital polymerase chain reaction (dPCR) as demonstrated by Li et al.¹⁴ Clinical symptoms, such as cough, fever, and dyspnea, are common, whereas radiological findings are dominated by cavitary and nodular lesions, fibrosis, and pleural effusion, which are the hallmarks of advanced pulmonary damage.

Although no significant associations were observed between sputum smear scores and clinical, demographic, or radiological features, a modest but consistent inverse correlation was found between smear scores and CT values (r = -0.342; p = 0.011). This finding supports the biological plausibility that a higher bacillary load corresponds to lower CT values because lower CTs indicate higher concentrations of MTB DNA. Smear-negative and scanty cases demonstrated the highest mean CT values (20.5 ± 4.7 and 20.3 ± 3.2), whereas patients with +2 and +3 smear scores exhibited notably lower mean CT values (16.7 ± 0.8 and 16.4 ± 0.7). This trend aligns with the quantitative principle of PCR, in which amplification occurs more rapidly in samples with higher DNA content.³² Similarly, Gota, Shenoy, and Kamath¹⁴ reported a stronger polychoric correlation (σ = -0.8681) between smear grade and CT, reinforcing the robustness of this inverse relationship. Although Li et al.³³ did not directly analyze smear correlation, their work supports the high sensitivity of dPCR for detecting MTB DNA, even in smear-negative specimens.³³

CT values were further associated with clinical severity: lower values were significantly associated with higher Modified Bandim TB scores and the presence of chest pain. Multivariate analysis confirmed that both variables were independently correlated with CT values (p = 0.036; β = -0.091) despite the limited explanatory power of the model (R² = 0.080). This suggests that smear scores and CT values, while related, are influenced by distinct factors, such as bacterial viability, DNA degradation, and host immune status. CT overlaps across smear categories and instances of low CT values in culture-negative samples,¹⁴ further indicating that PCR may detect nonviable DNA, thereby limiting its utility in assessing infectivity.

Importantly, treatment failure was independently associated with prolonged therapy and the presence of anemia, both of which exhibited statistically significant ORs. These results underscore the multifactorial nature of RR-TB treatment outcomes and the need for integrated clinical decision making that considers biological, treatment-related, and host-specific variables.