Aims
To evaluate the diagnostic yield, accuracy, and safety of Endoscopic Ultrasound-Guided Fine-Needle Biopsy (EUS-FNB) for isolated periportal lymphadenopathy of unclear etiology. A secondary objective was to identify specific EUS morphological features predictive of underlying pathology, particularly tuberculosis (TB), a common cause in our setting.
Methods
This prospective study, conducted at a single tertiary care centre, analyzed data from 68 patients who underwent EUS-FNB for isolated periportal lymphadenopathy between January 2021 and July 2022. Patients with a known primary malignancy were excluded. All procedures were performed by an experienced endosonographer using a 22-gauge core biopsy needle. Samples were systematically processed for histopathological examination and microbiological analysis, including GeneXpert ULTRA and MGIT culture. Final diagnoses were established using a composite gold standard of positive tissue analysis or clinical-radiological follow-up. Logistic regression was used to identify EUS predictors of TB
Results
Of 710 EUS procedures performed during the study period, 82 patients were referred for EUS-FNB of periportal lymph nodes, with 68 ultimately enrolled for analysis. The final diagnosis was tuberculosis in 47 patients (69.1%), malignancy in 7 (10.3%), and reactive lymphadenopathy in 10 (14.7%). On multivariate logistic regression, an irregular lymph node shape (adjusted OR: 4.71, p=0.039) and the presence of necrosis (adjusted OR: 3.28, p=0.049) were identified as independent predictors of a tubercular etiology. The integration of microbiological testing significantly enhanced diagnostic yield for tuberculosis; GeneXpert ULTRA and MGIT combined confirmed TB in 35 of 47 cases (74.5%), proving superior to histopathology alone, which was diagnostic in only 18 cases (40.4%) (p=0.00084). Overall, EUS-FNB demonstrated excellent diagnostic accuracy, with 94% sensitivity and 100% specificity for tuberculosis, and 87.5% sensitivity with 100% specificity for malignancy. The procedure was very safe, with no major complications reported among the patients.
|
Parameter |
Univariate OR (95% CI) |
p-value |
Multivariate OR (95% CI) |
p-value |
|
Irregular shape |
6.94 (2.10 – 22.95) |
0.0018 |
4.71 (1.08 – 20.57) |
0.039 |
|
Hypoechoic |
2.36 (0.25 – 22.60) |
0.452 |
- |
- |
|
Hyperechoic |
1.42 (0.47 – 4.28) |
0.543 |
- |
- |
|
Heterogeneous |
0.58 (0.21 – 1.62) |
0.299 |
- |
- |
|
Necrosis |
4.43 (1.51 – 13.00) |
0.006 |
3.28 (1.01 – 10.62) |
0.049 |
|
Calcification |
2.00 (0.54 – 7.40) |
0.294 |
- |
- |
|
Matting |
4.06 (1.12 – 14.63) |
0.033 |
3.03 (0.75 – 12.29) |
0.118 |
|
Long-axis diameter |
1.39 (0.74 – 2.60) |
0.303 |
- |
- |
|
Short-axis diameter |
1.33 (0.80 – 2.22) |
0.272 |
- |
- |
|
Diagnostic Category |
Sensitivity (%) |
Specificity (%) |
PPV (%) |
NPV (%) |
Accuracy (%) |
AUROC |
|---|---|---|---|---|---|---|
|
Tuberculosis |
94.0 |
100.0 |
100.0 |
85.7 |
95.6 |
0.97 |
|
Malignancy |
87.5 |
100.0 |
100.0 |
98.4 |
98.5 |
0.93 |
|
TB + Malignancy |
93.1 |
100.0 |
100.0 |
71.4 |
94.1 |
0.96 |
Conclusions
EUS-FNB is a highly accurate, safe, and effective first-line diagnostic modality for isolated periportal lymphadenopathy. In regions where TB is prevalent, it is the most common etiology. Our findings strongly support a multi-modal diagnostic strategy, combining EUS imaging, histopathology, and advanced microbiological testing (GeneXpert and MGIT) to maximize diagnostic yield, particularly for lymph node tuberculosis. Furthermore, EUS features such as irregular shape and necrosis are valuable independent predictors that can guide clinical suspicion and ensure timely patient management. This integrated approach should be considered the standard of care