Pediatric pulmonary and extrapulmonary tuberculosis: immunological and diagnostic perspectives

Abstract

Background

Pediatric tuberculosis (TB) remains a significant global health issue due to its diagnostic complexity and paucibacillary nature. This study aimed to evaluate the clinical, diagnostic, and immunological characteristics of pulmonary TB (PTB) and extrapulmonary TB (EPTB) in children.

Methods

A retrospective review was conducted on 88 pediatric TB cases (38 PTB, 50 EPTB) at a referral children’s hospital in Iran from 2020 to 2024. Diagnostic methods included nucleic acid amplification tests (NAATs), interferon-gamma release assay (IGRA), tuberculin skin test (TST), mycobacterial culture, and imaging studies such as chest X-ray (CXR) and computed tomography (CT). Flow cytometry was used to assess immune profiles.

Results

Among the 88 pediatric TB cases included in the study, 82 (93%) were caused by Mycobacterium tuberculosis (MTB), while 6 cases (7%) were attributed to nontuberculous mycobacteria (NTM). EPTB accounted for the majority of cases (57%). Drug susceptibility testing of MTB showed 87% sensitivity to all antibiotics; rifampicin and isoniazid resistance were found in 10% and 4%, respectively. Diagnostic test results for PTB and EPTB caused by MTB showed that NAATs had the highest positivity rates (100% for PTB and 96% for EPTB). IGRA was positive in 47% of PTB and 25% of EPTB cases. TST showed high positivity for both groups (80% PTB, 75% EPTB). Culture results were positive in 67% of PTB and 76% of EPTB cases. CXR was positive in 82% of PTB and 55% of EPTB cases, and CT scans showed 86% positivity for PTB and 58% for EPTB. A significant difference in immune profiles was observed between PTB and EPTB patients. The EPTB group exhibited notably reduced levels of CD3⁺ and CD8⁺ T cells, along with elevated CD19⁺, CD16⁺, and myeloperoxidase (MPO) levels.

Conclusion

EPTB and PTB in children display distinct immunological patterns. PTB is characterized by a stronger T-cell response, whereas EPTB exhibits greater humoral immune activation and T-cell suppression. The distinct immunological profiles suggest that different immune pathways contribute to disease localization. Integrating molecular diagnostics and immune profiling can enhance early detection and guide better clinical management of childhood TB.

Background

Pediatric tuberculosis (TB) has been neglected, primarily due to the challenges related to its diagnosis and management [1–3]. This is partly due to the non-specific presentation of the disease in children, the low sensitivity of available molecular tests, and mycobacterial culture to detect mostly paucibacillary forms of the disease. In addition, specimen collection for microbiological testing is challenging, particularly in young children who are often unable to spontaneously expectorate sputum [4].

Globally, approximately 8.2 million new cases of TB were reported in 2023, with a rise from previous years. Alarmingly, 12% of these cases were among children and adolescents [5]. Despite these statistics, childhood TB remains underreported in many regions, including Iran [6, 7]. In Iran, the incidence of TB was reported at 11 per 100,000 people in 2023, with children aged 0–14 years accounting for about 3% of the total cases. However, given the country’s intermediate TB burden, increasing diagnostic capacity, and diverse population, Iran provides a strategically important setting for studying pediatric TB. Insights gained from such a setting can help shape regional and global strategies for TB detection and management in children.

Despite advancements in diagnostic technology, the accurate and timely diagnosis of both pulmonary TB (PTB) and extrapulmonary TB (EPTB) remains a challenge due to reliance on traditional methods that struggle to detect the elusive bacilli. Addressing the persistent gaps in TB detection, treatment, and prevention is essential to achieving global TB elimination goals [8].

The immune response to TB infection is highly complex, involving multiple immune cell populations and molecular pathways that play critical roles in disease progression and control. Both innate and adaptive immunity contribute to host defense against M. tuberculosis (MTB), but the bacterium has developed strategies to evade immune detection and persist within the host [9]. However, there is a lack of data on how immune responses differ between PTB and EPTB in pediatric populations—especially in low- and middle-income countries.

This study aimed to analyze the clinical, epidemiological, and immunological features of pediatric PTB and EPTB at an Iranian referral hospital from 2020 to 2024. We hypothesized that pediatric PTB and EPTB differ significantly in their immune response profiles, and that integrating conventional and molecular diagnostics with immune profiling may enhance diagnostic accuracy and disease management.

Method

The study included all patients ≤ 16 years old who were hospitalized and diagnosed with TB during 2020 to 2024. Patient records were retrieved from both physical files and electronic medical records. This study was approved by the Ethical Committee of Tehran University of Medical Sciences, Tehran, Iran (IR.TUMS.CHMC.REC1403.045).

Data were systematically collected from patient files, focusing on several key parameters. These included demographic information such as age, sex, and any underlying health conditions. Clinical features were also examined, including the symptoms presented by the patients and the sites of TB involvement, distinguishing between pulmonary and extrapulmonary forms of the disease. Blood samples were analyzed for complete blood count (white blood cell (WBC), platelet, hemoglobin, neutrophil, and lymphocyte counts), inflammatory markers (ESR, CRP), and liver and kidney function tests (Aspartate aminotransferase (AST), alanine aminotransferase (ALT), alkaline phosphatase (ALK), total and direct bilirubin, creatinine, blood urea nitrogen (BUN), sodium, and potassium levels). Microbiological assessments included mycobacterial culture on clinical samples from various sources such as gastric aspirates, bronchoalveolar lavage (BAL), cerebrospinal fluid (CSF), abscesses, lymph nodes, and pleural fluid. Smear microscopy using Ziehl-Neelsen staining and nucleic acid amplification tests (NAAT) for MTB DNA detection were performed.

Immunological evaluations included interferon-gamma release assay (IGRA), lymphocyte subset analysis using flow cytometry (CD3⁺, CD4⁺, CD8⁺, CD16⁺, CD19⁺, and CD56⁺ markers), and complement system assessments (C3, C4, CH50), along with immunoglobulin levels (IgG, IgA, IgM, IgE). Flow cytometry analyses were based on a defined gating strategy using appropriate controls. Only percentages of cell populations were analyzed. Additionally, oxidative burst function was measured using the nitroblue tetrazolium (NBT) test, while myeloperoxidase (MPO) levels were assessed to evaluate neutrophil activity.

The diagnosis of TB was established based on a comprehensive assessment that included clinical evaluation, laboratory tests, and imaging studies. The clinical assessment involved reviewing symptoms, family history of TB, and contact history with known TB cases. Laboratory tests included positive TST results and culture-positive specimens obtained from sputum, alveolar lavage, or other biological samples. Microscopic examination or nucleic acid amplification tests for MTB detection were also utilized. For imaging studies, chest radiography and CT scans were analyzed for lesions consistent with TB. EPTB was diagnosed based on culture-positive specimens from extrapulmonary sites (e.g., pleural effusion or ascites) or through pathological analysis confirming tissue involvement. Patients with inconclusive or conflicting diagnostic data (e.g., clinical suspicion without radiologic or laboratory confirmation) were excluded from the analysis to ensure diagnostic accuracy and consistency. Only cases with a minimum threshold of diagnostic certainty—either bacteriological confirmation or strong clinical/radiologic evidence consistent with national TB guidelines—were included.

All isolates were tested using phenotypic drug susceptibility testing (DST) and/or genotypic methods.

Statistical analysis

Categorical variables were summarized as frequencies and percentages, while continuous variables were presented as mean ± standard deviation (SD) or median with interquartile range (IQR), depending on the data distribution. Statistical analyses were conducted using SPSS version 27.0 and GraphPad Prism version 9.0.

Group comparisons for categorical variables were made using the Chi-squared test or Fisher’s exact test, as appropriate. For continuous variables, the Mann–Whitney U test or Student’s t-test was applied, based on normality. For comparisons across multiple groups, the Kruskal–Wallis test or ANOVA was used accordingly. Multivariable linear regression was conducted to assess the independent effects of age and underlying comorbidities on immune cell subset levels, adjusting for these covariates in group comparisons. Pearson correlation analysis was used to evaluate associations between CD marker percentages and serum immunoglobulin levels. A p-value < 0.05 was considered statistically significant.

Results

Between 2020 and 2024, a total of 88 pediatric TB cases were reported, of which 82 (93.2%) were caused by MTB and 6 (6.8%) by NTM. Among them, 53 cases were males (60%). A mean age of the patients was 6.6 ± 4.8 years (range: 3 months to 16 years). All children were BCG vaccinated. Forty-seven cases (53%) were from urban areas, 69% were Iranian, and 31% were Afghan. Underlying diseases were present in 65% (35 out of 54) cases, including osteomyelitis (10%), arthritis (7%), blood cancer (4.5%), cystic fibrosis (3%), asthma (2%), and HIV (1%). A total of 13 patients (15%) required ICU admission. The annual distribution of TB cases showed fluctuations, with peaks in 2024 for both PTB at 24% (n = 9) and EPTB at 24% (n = 12) (Table 1).

Table 1.

Characteristics of the patients with pulmonary and extrapulmonary TB

Variable		Pulmonary (n, %)	Extrapulmonary (n, %)	Total (n, %)
Sex	Male	25 (66)	28 (56)	53 (60)
Nationality	Iranian	26 (68)	35 (70)	61 (69)
	Afghan	12 (32)	15 (30)	27 (31)
Residence	Urban	19 (51)	28 (57)	47 (55)
	Rural	18 (49)	21 (43)	39 (45)
Year	2020	12 (32)	14 (28)	26 (30)
	2021	8 (21)	8 (16)	16 (18)
	2022	4 (10.5)	5 (10)	9 (10)
	2023	5 (13)	11 (22)	16 (18)
	2024	9 (24)	12 (24)	21 (24)
Underlying disease		17 (77)	18 (56)	35 out of 54 (65)
Asthma		0 (0)	2 (4)	2 (2)
CF		3 (8)	0 (0)	3 (3)
Arthritis		2 (5)	4 (8)	6 (7)
Lymphadenitis		3 (8)	21 (42)	24 (27)
HIV		1 (3)	0 (0)	1 (1)
Blood Cancer		4 (10.5)	0 (0)	4 (4.5)
Bronchiectasia		5 (13.5)	0 (0)	5 (6)
Smoking exposure		18 (47)	14 (29)	32 (37)
Narcotic exposure		6 (16)	6 (12)	12 (14)
Contact with high-risk TB group		12 (80)	16 (61.5)	28 out of 41 (68)
ICU admission		3 (8)	10 (20)	13 (15)

CF: Cystic Fibrosis

HIV: Human Immunodeficiency Virus

Symptoms

The most common symptom in pediatric TB patients was fever (n = 60, 68%), followed by weight loss (n = 47, 53%). Cough was present in 36% (n = 32), and night sweats in 24% (n = 21). Chest pain occurred in 18% (n = 16). Less common symptoms included vomiting (12.5%, n = 11), headache (8%, n = 7), loss of consciousness (8%, n = 7), seizures (7%, n = 6), focal neurological complaints (6%, n = 5), and hemoptysis (3%, n = 3).

Diagnostic test results

MTB culture was positive in 71% of samples (62 out of 87). Smear microscopy showed positivity in 28% (24 out of 87), lower than culture-positive results (Table 2). Culture positivity varied by sample type: gastric aspirates (18%, n = 16), BAL (9%, n = 8), CSF (8%, n = 7), abscess (9%, n = 8), lymph node (13%, n = 11), and colon/pleural samples (3%, n = 3 each). Brain abscess and lymph node cultures were positive in 1 case each (1%).

Table 2.

Laboratory and imaging results for pediatric TB diagnosis

Variables	Frequency	%
Immunological tests
IGRA	14 out of 40	35
TST	18 out of 23	78.3
Microbiological tests
Smear Positive	24 out of 87	27.6
Culture Positive	62 out of 87	71.3
Specimen types
Gastric Aspirate	16	18.4
BAL	8	9.2
CSF	7	8
Abscess	8	9.2
Sputum	2	2.3
Auxiliary lymph nodes	11	12.6
Colon	3	3.4
Pleural	3	3.4
Brain tumor	1	1.1
Cervical lymph node	1	1.1
Others	2	2.3
Molecular test
NAATs Positive	86 out of 88	97.7
Radiological findings
Abnormal CXR	52 out of 80	65
Abnormal CT	20 out of 28	71.4

IGRA: Interferon-gamma release assay

TST: Tuberculin skin test

BAL: Bronchoalveolar Lavage

CSF: Cerebrospinal Fluid

CXR: Chest X-ray

CT: Computed tomography

NAATs showed high sensitivity (98%, 86 out of 88). BAL had the highest positivity rate (2%, n = 21), followed by abscess samples (22%, n = 19) and CSF (9%, n = 8). Sputum had a lower positivity rate (3.5%, n = 3). Gastric aspirates were positive in 17% (n = 15), and colon samples in 6% (n = 5). Bone, lymph node, and brain abscess samples each showed positivity in 5 cases (6%), while pleural and lung abscess samples had 2 (2%) and 1 case (1%), respectively.

TST was performed on a total of 23 individuals, of whom 18 (78%) were positive. Specifically, 8 out of 9 individuals without underlying disease (89%) tested positive, while only 3 out of 6 individuals with underlying disease (50%) showed TST positivity (p-value = 0.23).

Additionally, IGRA tests were positive in 14 out of 40 cases (35%). These results show a comparable IGRA positivity rate between individuals with and without underlying conditions (6 out of 17 (35%) vs. 3 out of 9 (33%)). Among 40 children tested, 3 out of 7 (43%) aged ≤ 5 years were IGRA positive, compared to 11 out of 33 (33%) aged > 5 years. Among 23 children tested with TST, 80% (4 out of 5) of those aged ≤ 5 years were TST positive, compared to 78% (14 out of 18) of children older than 5 years (p-value > 0.0.5).

In total, 16 individuals underwent both TST and IGRA testing. Among the 12 individuals who were TST positive, only 3 (25%) were also IGRA positive. Similarly, among the 4 individuals who were TST negative, 1 (25%) tested IGRA positive and 3 (75.0%) were IGRA negative.

Imaging studies indicated abnormalities, with CXR showing positive results in 52 out of 80 cases (65%), and CT scans were abnormal in 20 out of 28 cases (71%).

Type of TB

Of the 88 TB cases, 38 (43%) were PTB and 50 (57%) EPTB. A large proportion of patients had known contact with high-risk individuals—80% of PTB cases (12 out of 15) and 61.5% of EPTB cases (16 out of 26). Most TB cases were Iranian, comprising 68% (26 out of 38) of PTB and 70% (35 out of 50) of EPTB cases, while Afghan patients made up 32% (12 out of 38) and 30% (15 out of 50), respectively. Underlying diseases were present in 77% (n = 17) of PTB cases and 56% (n = 18) of EPTB cases (p-value = 0.15). Males accounted for 66% (n = 25) of PTB and 56% (n = 28) of EPTB cases (p-value = 0.39). Urban residents made up 51% (n = 19) of PTB and 57.1% (n = 28) of EPTB cases, while rural residents constituted 49% (n = 18) and 43% (n = 21), respectively (p-value = 0.66).

There were no significant differences in age and gender between the two groups. The median age of PTB patients was 7years (IQR: 3–11 years), while EPTB patients had a median age of 5.5 years (IQR: 1.9–10 years). Among children with PTB (n = 38), 18 (47%) were under 5 years and 20 (53%) were 5 years or older. In EPTB cases (n = 50), 25 (50%) were in each age group. A Chi-square test showed no significant association between age group and TB type (p-value = 0.83).

Symptoms among PTB and EPTB cases

Fever was significantly more common in PTB cases, with 36 out of 38 patients (95%) reporting it, compared to 24 out of 50 EPTB cases (48%) (p-value < 0.0001). Cough was predominantly associated with PTB, reported in 25 out of 38 cases (66%), whereas it was far less common in EPTB, occurring in only 7 out of 50 cases (14%) (p-value < 0.0001). Headache was a relatively rare symptom among PTB patients, with only 1 out of 37 cases (3%) reporting it, compared to 12.5% EPTB cases (6 out of 48).

Diagnostic test performance in PTB and EPTB

Table 3 shows the diagnostic tests result for PTB and EPTB caused by MTB. NAATs exhibited a 100% positivity rate for PTB (37 out of 37) and 96% in EPTB cases (43 out of 45). For the IGRA test, 9 out of 16 (47%) PTB cases and 5 out of 20 (25.0%) EPTB cases tested positive. The TST demonstrated a high positivity rate, with 8 out of 10 (80%) PTB cases and 9 out of 12 (75%) EPTB cases testing positive. The culture test revealed that 25 out of 36 (67%) PTB cases and 34 out of 45 (76%) EPTB cases were positive. The CXR showed 28 out of 34 (82%) PTB cases and 24 out of 40 (55%) EPTB cases were positive. NAATs positivity rates were 100% in the PTB and 96% in the EPTB group. CT showed 86% positivity for PTB (12 out of 14) and 58% for EPTB (7 out of 12).

Table 3.

Laboratory and imaging findings in pulmonary and extrapulmonary TB

Test	Type	Positive (n, %)	p- value
IGRA	Pulmonary	9 out of 16 (47%)	0.19
	Extrapulmonary	5 out of 20 (25%)
TST	Pulmonary	8 out of 10 (80%)	1.0
	Extrapulmonary	9 out of 12 (75%)
Culture	Pulmonary	24 out of 36 (67%)	0.46
	Extrapulmonary	34 out of 45 (76%)
CXR	Pulmonary	28 out of 34 (82%)	0.012
	Extrapulmonary	22 out of 40 (55%)
CT	Pulmonary	12 out of 14 (86%)	0.19
	Extrapulmonary	7 out of 12 (58%)
NAATs	Pulmonary	37 out of 37 (100%)	0.5
	Extrapulmonary	43 out of 45 (96%)

IGRA: Interferon-gamma release assay

TST: Tuberculin skin test

CXR: Chest X-ray

CT: Computed tomography

NAATs: Nucleic acid amplification tests

Antibiotic resistance patterns of MTB and NTM isolates

Among 82 MTB, 71 (87%) were sensitive to all tested antibiotics, including ethambutol, pyrazinamide, amikacin, isoniazid, and rifampicin. However, 8 (10%) exhibited resistance to rifampicin, and 3 (4%) showed resistance to isoniazid. All NTM isolates were sensitive to all mentioned antibiotics.

Laboratory results in PTB and EPTB

Hematological and biochemical parameters

The median WBC count was higher in PTB patients (12,945 [8,352.5–16,192.5]×10⁹/L) than in EPTB patients (9,6 [6,605–15,822.5]×10⁹/L). Platelet levels were also elevated in PTB cases, with a median of 427,0 [242,25–607,0]×10⁹/L compared to 357,0 [299,0–478,5]×10⁹/L in EPTB. Inflammatory markers, including ESR and CRP, were elevated in PTB patients, with ESR at 52.0 [24.75–77.25] mm/hr and CRP at 44.0 [19.75–88.5] mg/L, compared to 30.0 [12.5–64.0] mm/hr and 24.0 [7.5–65.5] mg/L in EPTB cases, respectively (Table 4).

Table 4.

Laboratory results in the patients with pulmonary and extrapulmonary TB

Variable	PTB (Median [IQR])	EPTB (Median [IQR])
White Blood Cells (×10⁹/L)	12,945 [8,352.5–16,192.5]	9,600 [6,605–15,822.5]
Platelets (×10⁹/L)	427,0 [242,250–607,0]	357,0 [299,0–478,5]
Hemoglobin (g/dL)	11.1 [9.15–12.65]	9.85 [8.9–11.53]
Erythrocyte Sedimentation Rate (mm/hr)	52.0 [24.75–77.25]	30.0 [12.5–64.0]
C-Reactive Protein (mg/L)	44.0 [19.75–88.5]	24.0 [7.5–65.5]
Neutrophils (%)	5,447 [3,173.25–10,370]	5,424.5 [3,140–9,240.25]
Lymphocytes (%)	3,523.5 [2,281–6,089.75]	2,630.5 [1,756.5–4,584.0]
Aspartate Aminotransferase (IU/L)	25.0 [22.0–45.0]	32.0 [23.0–47.0]
Alanine Aminotransferase (IU/L)	23.0 [16.0–33.0]	19.0 [14.0–35.0]
Alkaline Phosphatase (IU/L)	320.0 [195.0–386.0]	329.0 [247.0–430.0]
Blood Urea Nitrogen (mg/dL)	12.0 [7.5–15.0]	8.0 [6.0–14.0]
Creatinine (mg/dL)	0.5 [0.4–0.7]	0.5 [0.4–0.6]
Sodium (mmol/L)	137.0 [133.5–141.0]	136.0 [132.0–138.0]*
Potassium (mmol/L)	4.1 [3.9–4.3]	4.1 [3.75–4.5]
Bilirubin Total (mg/dL)	0.8 [0.65–1.5]	0.5 [0.4–0.6]*
Bilirubin Direct (mg/dL)	0.2 [0.2–1.0]	0.2 [0.2–0.3]*
Albumin (g/dL)	3.3 [2.8–3.6]	3.9 [3.1–4.1]

* The differences observed between PTB and EPTB were statistically significant (p-value < 0.05)

PTB: Pulmonary TuberculosisEPTB: Extrapulmonary Tuberculosis

Liver and kidney function

Liver enzymes varied between groups, with AST higher in EPTB (32 IU/L) than PTB (25 IU/L) and ALT lower in EPTB (19 IU/L) compared to PTB (23 IU/L). Total bilirubin was higher in PTB (0.8 mg/dL) than EPTB (0.5 mg/dL). Kidney function, including creatinine and BUN, was normal, though BUN was lower in EPTB (8 mg/dL) than PTB (12 mg/dL) (Table 5).

Table 5.

Immune cell profiles in pediatric PTB and EPTB cases

Variable	PTB (Median [IQR])	EPTB (Median [IQR])	p-value
CD3+ (%)	69.5 [52.5–71.25]	53.0 [39.2–62.25]	0.034
CD4+ (%)	36.0 [28.25–41.0]	35.0 [17.0–41.0]	0.690
CD8+ (%)	26.0 [20.25–29.5]	18.0 [12.0–24.0]	0.016
CD16+ (%)	8.0 [7.0–9.5]	14.0 [5.0–19.5]	0.034
CD19+ (%)	17.0 [13.0–28.75]	36.0 [19.5–49.0]	0.036
CD56+ (%)	7.0 [6.0–10.5]	10.5 [5.0–18.25]	0.250
IgG (mg/dL)	874.0 [524.5–1,189.5]	1021.0 [606.25–1,651.5]	0.649
IgA (mg/dL)	89.0 [34.0–303.0]	68.0 [45–246.0]	0.857
IgM (mg/dL)	130.0 [91.5–181.5]	148.0 [97.0–167.0]	1.000
IgE (IU/mL)	27.0 [8.5–109.5]	98.0 [31.5–338.0]	0.076
MPO (ng/mL)	70.0 [62.25–71.75]	80.0 [74.25–83.75]	0.085
HLA-DR (%)	22.5 [13.5–46.5]	13.5 [12.0–35.75]	0.559
C3 (mg/dL)	135.5 [106.75–175.0]	146.0 [120.5–169.5]	0.739
C4 (mg/dL)	29.0 [21.5–33.0]	30.0 [27.5–35.5]	0.569
CH50 (U/mL)	131.0 [110.5–139.5]	125.0 [97.5–134.0]	0.752

MPO: Myeloperoxidase

Immune cell profiles

NTM cases were excluded from the immunological analyses. CD3⁺, CD8⁺, CD16⁺, CD19⁺ percentage, and MPO levels showed significant differences between PTB and EPTB cases. CD3⁺ levels were lower in EPTB (53.5%) compared to PTB (69%). CD8⁺ was also reduced in EPTB (18%) compared to PTB (26%), while CD19⁺ was significantly higher in EPTB (33.5%) compared to PTB (18%). MPO levels were elevated in EPTB patients (80 ng/mL) compared to PTB (70 ng/mL) (Table 5; Fig. 1).

Fig. 1.

Median immune cell profiles in pediatric PTB and EPTB cases. This figure illustrates the distribution of immune cell profiles in pediatric PTB and EPTB cases. The square points represent PTB cases, while the circle points represent EPTB cases. Data for immune cell subsets, are presented as median percentages

We performed multivariable analyses to assess whether immune cell subset levels (CD3⁺, CD4⁺, CD8⁺, CD16⁺, CD19⁺, and CD56⁺) were influenced by age and underlying comorbidities. After adjusting for age and underlying disease, no statistically significant effects of these covariates were observed on any of the immune markers (p-values > 0.05).

To assess the relationship between immune cell subset percentages and serum immunoglobulin levels (IgG, IgA, IgM), Pearson correlation analysis was performed. CD8⁺ and CD56⁺ cell percentages were positively correlated with IgG levels (r = 0.451, p = 0.003 and r = 0.571, p = 0.001, respectively). Similarly, CD16⁺ cells showed a strong positive correlation with IgM levels (r = 0.572, p = 0.001).

Complement system and other markers

HLA-DR levels were lower in EPTB cases (13.5%) compared to PTB (22.5%). Other complement markers, including C3, C4, and CH50, showed minimal variation between PTB and EPTB.

Discussion

Diagnosing pediatric TB is challenging due to the often paucibacillary nature of the disease. Without timely detection and treatment, TB can progress rapidly, frequently resulting in severe and potentially life-threatening complications [10]. The findings from this study provide valuable insights into the clinical and laboratory findings of pediatric TB, highlighting both similarities and differences between PTB and EPTB.

In recent years, EPTB incidence in Iran has slightly increased, highlighting its growing clinical importance [11]. Our study found that 57% of children had EPTB, a higher rate than reported in China (46%) [12], Colombia (34.4%) [13], India (43.1%) [14], and Brazil (14.2%) [15]. However, our findings were consistent with studies from China [16], where more than 50% of children hospitalized with TB had EPTB. This variation may result from referral bias at our tertiary center, regional differences in TB strains and transmission, host genetic factors, and disparities in healthcare access and diagnostic capacity. Environmental and socioeconomic factors like malnutrition and co-infections may also influence EPTB occurrence.

This study highlights the variable sensitivity of TB diagnostics in children. Culture, though the gold standard, showed a 71% positivity rate (62 out of 87). In addition, smear microscopy had low sensitivity (28%) (24 out of 87), reflecting challenges in detecting paucibacillary disease in children [17]. NAATs demonstrated the highest positivity rate (98%, 86 out of 88), underscoring their superior sensitivity and rapid results. These findings support global recommendations favoring molecular diagnostics for pediatric TB, where bacteriological confirmation is often difficult [17].

In our study, IGRA had a 35% positivity rate (14 out of 40), while TST positivity was higher 78% (18 out of 23). IGRA and TST positivity showed no significant difference by age group (≤ 5 vs. >5 years) (p > 0.05). Notably, among 12 TST-positive patients, only 3 (25%) were also IGRA positive. Similarly, just 1 of 4 TST-negative cases was IGRA-positive. Discordance between TST and IGRA results has been reported in various studies [18, 19]. In our study, only 19% showed concordant positive results, highlighting poor agreement between the two tests and reinforcing the need for additional diagnostics when TB is suspected [20].

The surprisingly low IGRA sensitivity (35%) observed in our study may be influenced by several factors. Immunosuppression can impair the immune response, leading to false-negative IGRA results. The IGRA test shows similar positivity rates in both individuals with and without underlying diseases (33% vs. 35%). This suggests that the presence of underlying diseases does not significantly affect IGRA results. Other possible reasons include the young age of the participants, as immune immaturity in children can reduce IGRA responsiveness. Additionally, technical factors related to test administration, or variability in laboratory procedures could contribute to reduced sensitivity. On the other hand, IGRA may inherently have lower sensitivity in pediatric populations compared to adults, highlighting the need for complementary diagnostic methods when evaluating TB in children. Since EPTB frequently involves immune responses that are localized to specific tissue compartments—such as the lymph nodes, pleura, central nervous system, or bones—systemic blood tests like IGRA may fail to detect the full extent of T-cell activation occurring at the actual site of infection. Consequently, IGRA performed on blood samples can yield false-negative results in EPTB cases, underestimating the immune response and limiting its diagnostic utility [21, 22].

In our study, 87% of MTB were sensitive to all tested antibiotics, including ethambutol, pyrazinamide, amikacin, isoniazid, and rifampicin. However, 10% exhibited resistance to rifampicin, and 4% showed resistance to isoniazid. In a previous study conducted in India, 5.7% of the MTB isolates showed resistance to isoniazid, while none exhibited resistance to rifampicin or ethambutol [23]. An analysis of isoniazid resistance among rifampicin-sensitive PTB in children and adolescents found an estimated resistance rate of 9.8% (95% CI: 8.7–11.1%) in new PTB cases, 6.8% (95% CI: 5.4–8.5%) in EPTB, and 14.6% (95% CI: 11.8–17.9%) in previously treated PTB [24]. Globally, it has been estimated that 12.1% (95% CI: 9.8%–14.8%) had isoniazid-resistant disease [25]. In another study conducted in India, 7 cases (5.5%) were resistant to rifampicin either singly or in combination, while 11 cases (8.7%) were resistant to isoniazid either singly or in combination [26].

Children with TB exhibit distinct immune response profiles compared to adults [27]. The immune response to MTB extends beyond the classical Th1/Th17 cellular pathways, with growing evidence suggesting that B-cells and antibodies (Abs) can complement and enhance cell-mediated immunity (CMI). While Abs and B-cells alone may not be sufficient to eliminate MTB, their supportive role in immunity should not be overlooked [28].

Interestingly, while Ig levels such as IgG were higher in EPTB patients, the differences were not statistically significant. Some studies suggest that serum Ig levels increase proportionally with TB severity and organ involvement [29]. However, our findings indicate that while humoral immunity may be more active in EPTB, it is not definitive enough to distinguish between the two disease forms.

Multiparametric flow cytometry enables detailed profiling of immune responses to MTB and shows promise as a diagnostic tool by identifying TB-specific immune signatures [3, 30]. In this study, EPTB cases showed a significantly lower CD3⁺ T-cell percentage than PTB (53% vs. 70%, p- value = 0.034), indicating greater cellular immune suppression. This may be due to downregulation of the CD3-ζ chain, a critical component of TCR signaling [31]. Reduced CD3-ζ expression impairs T-cell activation and has been associated with T-cell dysfunction and MTB immune evasion, contributing to global T-cell depletion observed in active TB, particularly among CD4⁺ cells [32].

MTB has evolved mechanisms to downregulate CD4⁺ T-cell activation, leading to T-helper cell (Th1/Th2) imbalance [3] and low CD4⁺ expression [4, 5]. However, the effect on CD8⁺ T-cells remains controversial, with some studies reporting CD8⁺ increases [6, 7] while others suggest CD8⁺ depletion [8, 9]. Further investigation is required to determine whether CD8⁺ modulation contributes to EPTB pathogenesis. However, in our study, CD4⁺ T-cell percentages did not significantly differ between PTB and EPTB, suggesting that CD4⁺ T-cells play a crucial role in TB pathogenesis regardless of localization. Previous studies on CD4⁺ and CD8⁺ T-cell changes in TB have reported inconsistent results [32–36]. Some studies indicate CD4⁺ depletion with stable CD8⁺ levels [36], while others report CD8⁺ increases or decreases depending on disease severity [35].

The negative correlation between CD4⁺ T-cell counts and bacterial load suggests that as MTB burden increases, CD4⁺ levels decrease, contributing to worsening disease outcomes [37]. This is consistent with reports that low CD4⁺ and CD8⁺ T-cell counts are associated with greater lesion extent in TB patients [35].

A key finding was the higher CD19⁺ percentage in EPTB patients compared to PTB, which was statistically significant (p-value = 0.036). While previous studies found no significant difference in CD19⁺ levels between active TB and latent TB infection [32], our study shows that within active TB cases, EPTB exhibits a distinct B-cell profile.

Our study indicates that certain immune cell subsets are closely linked to humoral immune responses in TB patients. The significant positive correlations between CD8⁺ and CD56⁺ cells and IgG levels suggest that increased cytotoxic T-cell and natural killer (NK) cell activity may be associated with elevated IgG production. This could reflect a coordinated adaptive and innate immune response against infection. Likewise, the positive correlation between CD16⁺ cells—commonly associated with NK cells and monocytes—and IgM levels suggests that early immune activation involving these cells may be related to heightened initial antibody responses. Overall, these findings imply that specific immune cell subsets may influence or reflect patterns of immunoglobulin production in TB, potentially offering insights into disease activity or immune regulation.

This suggests a stronger B-cell response in EPTB, possibly due to more pronounced immune activation in disseminated disease. Increased CD19⁺ levels may indicate heightened humoral immunity in EPTB, which could contribute to its pathogenesis. However, further research is needed to explore the role of this immune response in disease progression.

Additionally, there was a significantly higher percentage of CD16⁺ cells in EPTB, which is consistent with previous reports [38, 39]. This observation may suggest differences in NK cell activity or monocyte function between PTB and EPTB patients.

In PTB patients, the decreased levels of CD56^brightCD16^- in peripheral blood suggest an impairment in the immune system’s ability to mount an effective cytokine-mediated response [40]. Since this NK subset is typically less mature but highly efficient at producing cytokines, its reduction may contribute to weakened early immune signaling and impaired control of TB infection. In contrast, Choreño-Parra et al. reported that patients with tuberculous meningitis exhibited an increased frequency of cytotoxic CD56^dimCD16 + NK cells in circulation [39], highlighting potential differences in NK cell dynamics between PTB and EPTB.

Elevated MPO levels in EPTB may reflect increased neutrophil-driven inflammation, consistent with the more disseminated nature of the disease. MPO is an enzyme released by activated neutrophils during the innate immune response and plays a key role in microbial killing through the generation of reactive oxygen species. In EPTB, the involvement of deeper or multiple tissue sites likely triggers stronger neutrophil recruitment and activation, leading to elevated MPO release. This heightened response may indicate a more aggressive inflammatory process aimed at controlling widespread infection, but it can also contribute to tissue damage.

This study has several limitations. Its retrospective design may have led to missing or incomplete clinical data. The relatively small sample size, especially within PTB and EPTB subgroups, limits generalizability. The absence of longitudinal follow-up prevents analysis of long-term immune responses. Diagnostic tests like IGRA and TST were not uniformly applied, hindering direct comparison, and notable discordance between their results could not be fully explored. As a single-center study, selection bias is possible. Immune profiling was limited in scope, and confounding factors such as co-infections or prior TB exposure were not accounted for, potentially affecting immune and diagnostic outcomes. In addition, immune profiling was not performed for all patients, which may introduce selection bias and affect the generalizability of subgroup comparisons.

In conclusion, this study reveals immunological distinctions between pediatric PTB and EPTB. PTB is associated with a stronger and more effective T-cell–mediated response, while EPTB shows a profile dominated by humoral activation, innate inflammatory signaling, and features of T-cell suppression or exhaustion. These findings suggest that the clinical manifestation of pediatric TB reflects the balance between successful T-cell–driven containment and a dysregulated immune response characterized by impaired T-cell function and compensatory B-cell and innate activation. Such profiles may also help identify children at greater risk of disease dissemination. From a diagnostic perspective, integrating immune profiling into clinical workflows may improve differentiation between PTB and EPTB, particularly in pediatric patients who often present with non-specific symptoms and are more prone to extrapulmonary involvement. Earlier and more accurate classification of disease type could enhance clinical decision-making and improve outcomes. On the other hand, these findings provide a rationale for host-directed treatment strategies. Children with EPTB and evidence of T-cell exhaustion may benefit from targeted immunomodulation (e.g., anti-PD-1), while modulating pathological B-cell or neutrophil responses could be advantageous in severe extrapulmonary disease. The proposed mechanisms are hypotheses that need validation through functional assays (e.g., measuring T-cell exhaustion markers, cytokine production, and B-reg function) in larger, prospective studies.

Abbreviations

TB

Tuberculosis

PTB

Pulmonary TB

EPTB

Extrapulmonary TB

MTB

Mycobacterium tuberculosis

TST

Tuberculin skin tests

CXR

Chest X-rays

CT

Computed tomography

WBC

White blood cell

AST

Aspartate aminotransferase

ALT

Alanine aminotransferase

ALK

Alkaline phosphatase

BUN

Blood urea nitrogen

BAL

Bronchoalveolar lavage

CSF

Cerebrospinal fluid

NAAT

Nucleic acid amplification tests

IGRA

Interferon-gamma release assay

NBT

Nitroblue tetrazolium test

MPO

Myeloperoxidase

DST

Drug susceptibility testing

NTM

Nontuberculous mycobacteria

Author contributions

BP: conceptualization, supervision, project management, methodology, research, and examination. HRR: conceptualization, data collection, methodology, research, and examination. MS: methodology, validation, analysis, research, examination, and validation. RHS: data collection, methodology, research, and examination. SM1: conceptualization, methodology, validation, analysis, drafting the manuscript, editing, finalization, and visualization. SM2: conceptualization, supervision, project management, methodology, research, examination and validation.

Funding

NA.

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

This research was conducted in accordance with the principles of the Declaration of Helsinki. This study was approved by the Ethics Committee of Tehran University of Medical Sciences, Tehran, Iran (IR.TUMS.CHMC.REC1403.045). In addition, informed consent was obtained from the parents or legal guardians of any participant under the age of 16.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.