High-resolution sonography has been widely used for assessment of lymph nodes in the neck and has been established as a first-line imaging tool.15–17 The main roles of sonography in patients with a clinical suspicion of tuberculous lymphadenitis in the neck are detection and characterization. Furthermore, sonographic guidance is available worldwide for diagnostic and interventional procedures. As a diagnostic tool for tuberculous lymphadenitis in the neck, sonography may be behind sonographically guided fine-needle aspiration or PCR. Recently, one study comparing sonographic diagnosis, sonographically guided fine-needle aspiration, and PCR for detecting tuberculous lymphadenitis demonstrated that the diagnostic values of PCR alone and in combination with sonographically guided fine-needle aspiration were higher than those of other methods.18
Among the 4 individual sonographic features, a predominantly hypoechoic node with a heterogeneous echo pattern and intranodal necrosis showed high specificity and a high PPV but low sensitivity. Two sonographic features (poorly defined anechoic areas in the perinodal soft tissue with or without sinus and abscess formation and a vascular distribution with apparently avascular areas and displaced vascularity on color Doppler sonography) showed high sensitivity and a high NPV but low specificity. Therefore, no sonographic features with both high sensitivity and specificity were found. Among the sonographic features, poorly defined anechoic areas in the perinodal soft tissue have been known to indicate soft tissue edema.4,14 In this study, 13 patients with tuberculous lymphadenitis showed poorly defined anechoic areas in the perinodal soft tissue. Among them, 7 (53.8%) had a minimal amount of pyogenic aspirates on sonographically guided fine-needle aspiration of these sites; thus, we believe that an early-stage abscess is more likely than soft tissue edema. For clarity, further studies may be required.
For sonographic diagnosis, the “2 or more” category was more accurate for diagnosis of tuberculous lymphadenitis than the “1 or more,” “3 or more,” and “4” categories. Of all 79 patients, 62 (78.5%) were classified as “1 or more,” 46 (58.2%) as “2 or more,” and 22 (27.8%) as “3 or more.” Of the 40 patients with tuberculous lymphadenitis, 40 (100%) were classified as “1 or more,” 38 (95%) as “2 or more,” and 20 (50%) as “3 or more.” In comparison with the “2 or more” category, the “1 or more” category had high sensitivity but low specificity, and the “3 or more” category had high specificity and a high PPV but low sensitivity. However, we did not investigate a diagnostic index for sonographic diagnosis of tuberculous lymphadenitis according to different combinations of the 4 sonographic features.
There were several limitations to this study. First, it was a retrospective study. In each case, limited sonograms were used. Second, a single radiologist performed the sonographic examinations, sonographically guided fine-needle aspirations, and retrospective image analyses of lymph nodes in the neck, but the image analyses with blinding to all patient information were done 2 to 3 years after neck sonography and fine-needle aspiration. Third, 5 patients with tuberculous lymphadenitis were confirmed as having a reactivated type, but we did not discriminate between them. Finally, color Doppler sonography was used for evaluation of nodal vascularity in all cases, whereas power Doppler sonography was performed with color Doppler sonography in a minority. However, the sensitivity of color Doppler sonography is not different from that of power Doppler sonography because of advances in technology.12,13
In conclusion, the study results showed that no sonographic feature had both high sensitivity and specificity for identification of tuberculous lymphadenitis, but the diagnostic accuracy of the “2 or more” category was higher than those of other categories. Therefore, our sonographic diagnostic method may be useful for detection of tuberculous lymphadenitis in clinically suspected patients.