To compare the laboratory data of patients with breast tuberculosis and non-tuberculosis diseases, and explore the optimal diagnostic method for breast tuberculosis.

The clinical data of 147 patients with suspected breast tuberculosis in the Wuhan Pulmonary Hospital from December 2014 to March 2019 were collected for a retrospective analysis. According to the final pathological diagnosis, there were 48 cases of breast tuberculosis (tuberculosis group) and 99 cases of other breast diseases (non-tuberculosis group). The baseline data and laboratory data were compared between two groups. The quantitative data of normal distribution (age) were expressed as ±s, and compared by t test. The quantitative data of skewed distribution [erythrocyte sedimentation rate(ESR)and hypersensitive C-reactive protein(hCRP)] were expressed as M (P₂₅-P₇₅) and compared by Wilcoxon signed rank test. Count data [lesion position, results of purified protein derivative(PPD), T cell spot detection(T-SPOT), tuberculosis antibody(TB-Ab), tuberculosis deoxyribonucleic acid (TB-DNA), tuberculosis ribonucleic acid (TB-RNA), gene X-pert multidrug-resistant tuberculosis/rifampicin (X-pert MTB/RIF), smear of tubercle bacillus (STB), culture of tubercle bacillus (CTB)] were expressed as cases (%) and compared with paired χ² test between two groups. In breast tuberculosis patients, the Kappa test was used to analyze the consistency between the results of different diagnostic methods.

There was no significant difference between tuberculosis group and non-tuberculosis group in age, gender and lesion position (t=0.427, P=0.567; P=1.000; χ²=0.844; P=0.933). No significant difference was observed in ESR and hCRP levels in peripheral blood between tuberculosis group and non-tuberculosis group [13.5 (11.0-30.7) mm/h vs 13.0 (10.0-25.0) mm/h, Z=1.945, P=0.163; 8.5 (5.0-20.7) mg/L vs 9.0 (6.0~22.0) mg/L, Z=0.480, P=0.488). The results of three non-invasive examinations (PPD, T-SPOT and TB-Ab) presented no significant difference between two groups (P=1.000, 0.664, 0.272), while the results of five etiological methods (X-pertMTB/RIF, TB-DNA, TB-RNA, CTB and STB) showed a significant difference (P=0.031, 0.016, <0.001, <0.001, <0.001). In breast tuberculosis patients, the result of TB-Ab showed a poor consistency with that of T-SPOT (Kappa=-0.339, P<0.001); the results of TB-RNA, CTB and STB showed a poor consistency with those of X-pert MTB/RIF (Kappa=-0.200, P=0.005; Kappa=-0.152, P=0.012; Kappa=-0.159, P=0.003). The specific pathological changes in breast tuberculosis were granulomatous lesions and caseous necrosis.

It is difficult to distinguish breast tuberculosis from other diseases by ESR and hCRP. Clinically, the combination of etiological and histopathologic examinations is recommended for differential diagnosis of breast tuberculosis.