-
A total of 174 patients with puncture-sampled lesions under ultrasound guidance were enrolled in this study. Fifty-nine cases were excluded due to incomplete sampling, leaving 115 cases for analysis. Puncture samples from the 115 participants were tested using culture, smears, Xpert, and histopathology. Of the 115 participants, 72 (62.61%) were male and 43 (37.39%) were female, whereas 46 (40%) were HIV-positive and 69 (60%) HIV-negative (Table 1, Figure 1). Liquid culture of the 115 puncture samples revealed 42 positive cases (36.52%), of which 39 were positive for MPT64 (confirmed TB), and 73 cases (63.48%) were culture-negative. CRS led to a clinical diagnosis of TB in 87 of the 115 patients (75.65%) and of non-TB in 28 cases (24.35%; 1 case of Burkitt lymphoma, 4 cases of NTM, 2 cases of reactive hyperplasia, 1 case of epididymitis, 13 cases of lymphadenitis, 1 case of Staphylococcus aureus infection, 3 cases of lymphoma, and 3 cases of fungal infection).
Characteristics No. of total case No. of TB cases No. of case without TB P value Gender, n (%) Male 72/115 (62.61) 53/87 (60.92) 19/28 (67.86) 0.65 Female 43/115 (37.39) 34/87 (39.08) 9/28 (32.14) Age (Median [IQR]) 33 (27−44) 25 (31−43) 42 (30−55) 0.22 HIV, n (%) Positive 46/115 (40.00) 32/87 (36.78) 14/28 (50.00) 0.27 Negative 69/115 (60.00) 55/87 (63.22) 14/28 (50.00) Lymphatic diameter (mm) (Median [IQR]) 31 (21−45) 32 (22−45) 28 (18−49) 0.39 Pus volume (mL) (Median [IQR]) 0 (0−1) 0 (0−1) 0 (0−0) 0.92 Sample length (mm) (Median [IQR]) 25 (16−30) 25 (18−30) 20 (15−30) 0.14 Note. TB, tuberculosis; IQR, interquartile range; HIV, human immunodeficiency virus. Table 1. Characteristics of study participants
-
When CRS was used as the reference standard, detection using histopathological examination had the highest sensitivity (85.06%, 95% CI: 76.10%–91.05%), followed by Xpert (70.11%, 95% CI: 59.81%–78.72%), smear (51.72%, 95% CI: 41.38%–61.93%), and culture (44.83%, 95% CI: 34.82%–55.28%), but modestly lower specificity (82.14%, 95% CI: 64.41%–92.12%) than the other methods (Table 2). McNemar’s test was used to evaluate the diagnostic performance of the four diagnostic methods using CRS as the reference standard. Culture (χ2 = 39.71, P < 0.001), smear (χ2 = 33.80, P < 0.001), and Xpert (χ2 = 21.33, P < 0.001) showed statistical differences compared with CRS, but no significant difference was reached compared with pathological examination (χ2 = 2.72, P = 0.099). Further, the diagnostic concordance between CRS and the four diagnostic methods was evaluated using a Kappa test. CRS showed modest but significant consistency with culture (Kappa = 0.221, Z = 3.261, P < 0.001), smear (Kappa = 0.279, Z = 3.828, P < 0.001), and Xpert (Kappa = 0.511, Z = 6.144, P < 0.001), as well as a higher degree of consistency with pathological evaluation (Kappa = 0.613, Z = 6.670, P < 0.001).
Method Result TB (n) Non-TB (n) Sensitivity (%) (95% CI) Specificity (%) (95% CI) PPV (%) NPV (%) Culture Positive 39 3 44.83 89.29 92.86 34.25 Negative 48 25 (34.82−55.28) (72.80−96.29) Smear Positive 45 3 51.72 89.29 93.75 37.31 Negative 42 25 (41.38−61.93) (72.80−96.29) Xpert Positive 61 1 70.11 96.43 98.39 50.94 Negative 26 27 (59.81−78.72) (82.29−99.82) Pathology Positive 74 5 85.06 82.14 93.67 63.89 Negative 13 23 (76.10−91.05) (64.41−92.12) Note. CRS, composite reference standard; EPTB, extrapulmonary tuberculosis; CI, confidence interval; PPV, positive predictive value; NPV, negative predictive value. Table 2. Contribution of different methods for detection of EPTB defined by CRS
A similar trend was observed when liquid culture was used as the standard. TB histopathological detection had superior sensitivity (92.86%, 95% CI: 80.99%–97.54%), which was higher than that of Xpert (83.33%, 95% CI: 69.40%–91.68%) and smear (66.67%, 95% CI: 51.55%–78.99%). However, the specificity of all tests was low, ranging from 45.21% to 72.60% (Table 3). Further, McNemar’s test showed significant differences between culture and both Xpert (χ2 = 10.62, P = 0.001) and pathology (χ2 = 31.84, P < 0.001) but no significant difference with smear (χ2 = 0.74, P = 0.391). Diagnostic concordance between liquid culture and the other three diagnostic methods was evaluated using a Kappa test. Culture showed statistical consistency with smear (Kappa = 0.381, Z = 4.112, P < 0.001) and Xpert (Kappa = 0.421, Z = 4.801, P < 0.001), but a lower degree of consistency with pathology (Kappa = 0.321, Z = 4.238, P < 0.001).
Method Result Culture + (n) Culture − (n) Sensitivity (%) (95% CI) Specificity (%) (95% CI) PPV (%) NPV (%) Smear Positive 28 20 66.67 72.60 58.33 79.10 Negative 14 53 (51.55−78.99) (61.44−81.51) Xpert Positive 35 27 83.33 63.01 56.45 86.79 Negative 7 46 (69.40−91.68) (51.55−73.18) Pathology Positive 39 40 92.86 45.21 49.37 91.67 Negative 3 33 (80.99−97.54) (34.31−56.58) Note. CI, confidence interval; PPV, positive predictive value; NPV, negative predictive value. Culture +, Culture positive; Cuture −, Culture negative. Table 3. Potential contribution of supplementary diagnostic tests of the CRS method to diagnosis with culture negative EPTB biopsy samples
-
To further study the utility of lesion pathology in diagnosing EPTB, the results of the histopathological assessment were combined with those from culture, smear, and Xpert. In the combined diagnostic criteria, a positive result was determined when one of the test results was positive, whereas a negative result required both test results to be negative. The sensitivity and specificity of the combined tests are presented in Table 4. The combination of pathological analysis and Xpert led to a sensitivity of 90.80% and specificity of 82.14% (Table 4). Sensitivity was highest when the specificity of the combined detection methods was equal at 82.14%. In addition, we found that the combination of the four methods led to an increase in detection sensitivity to 91.95% (Table 4).
Method Result TB (n) Non-TB (n) Sensitivity (%) (95% CI) Specificity (%) (95% CI) PPV (%) PNV (%) Culture+Pathology Positive 77 5 88.51 82.14 93.90 69.70 Negative 10 23 (80.12−93.64) (64.41−92.12) Smear+Pathology Positive 76 5 87.36 82.14 93.83 67.65 Negative 11 23 (78.76−92.79) (64.41−92.12) Xpert+Pathology Positive 79 5 90.80 82.14 94.05 74.19 Negative 8 23 (82.89−95.27) (64.41−92.12) Culture+Smear+Xpert Positive 66 4 75.86 85.71 94.29 53.33 Negative 21 24 (65.90−83.64) (68.51−94.30) Cuture+Smear+
Xpert+PathologyPositive 80 5 91.95 82.14 94.12 76.67 Negative 7 23 (84.31−96.05) (64.41−92.12) Note. TB, tuberculosis; CI, confidence interval; PPV, positive predictive value; NPV, negative predictive value. Table 4. Different detection methods combined with pathology to diagnose EPTB in puncture samples
-
Higher detection rates were observed in the samples from the neck and axilla than in those from other sites. Among all methods, histopathological assessment resulted in the highest detection rate, followed by Xpert. Notably, the detection rates from histopathological assessment of neck, chest, and axillary lesions were 81.13% (43/53), 66.67% (4/6), and 92.86% (13/14), respectively (Table 5).
Site Incidence of culture (%) Incidence of smear (%) Incidence of xpert (%) Incidence of pathology (%) Neck (n = 53) 23/53 (43.40) 25/53 (47.17) 36/53 (67.92) 43/53 (81.13) Chest (n = 6) 2/6 (33.33) 2/6 (33.33) 4/6 (66.67) 4/6 (66.67) Abdomen (n = 3) 1/3 (33.33) 2/3 (66.67) 2/3 (66.67) 3/3 (100.00) Armpit (n = 14) 7/14 (50.00) 10/14 (71.43) 11/14 (78.57) 13/14 (92.86) Joint (n = 6) 3/6 (50.00) 3/6 (50.00) 5/6 (66.67) 6/6 (100.00) Collarbone (n = 4) 2/4 (50.00) 3/4 (75.00) 2/4 (50.00) 4/4 (100.00) Table 5. Different detection methods performance in various puncture sites for the diagnosis of EPTB
Effectiveness of Histopathological Examination of Ultrasound-guided Puncture Biopsy Samples for Diagnosis of Extrapulmonary Tuberculosis
doi: 10.3967/bes2024.017
- Received Date: 2023-02-23
- Accepted Date: 2023-08-24
-
Key words:
- Extrapulmonary tuberculosis /
- Diagnosis /
- Biopsy /
- Histopathological examination /
- Puncture samples
Abstract:
The authors declare that the research was conducted without any commercial or financial relationships that could be construed as a potential conflict of interest.
&These authors contributed equally to this work.
Citation: | GU Wen Fei, SHI Xia, MA Xin, YU Jun Lei, XU Jin Chuan, QIAN Cheng Cheng, HU Zhi Dong, ZHANG Hui. Effectiveness of Histopathological Examination of Ultrasound-guided Puncture Biopsy Samples for Diagnosis of Extrapulmonary Tuberculosis[J]. Biomedical and Environmental Sciences, 2024, 37(2): 170-177. doi: 10.3967/bes2024.017 |