Diagnostic value of the combinations of bronchoalveolar lavage fluid galactomannan test and serum galactomannan test in invasive pulmonary aspergillosis / 中华危重病急救医学

ABSTRACT

Objective To evaluate the accuracy and diagnostic value of bronchoalveolar lavage fluid galactomannan test (BALF-GM) combined with serum GM test on invasive pulmonary aspergillosis (IPA). Methods 190 cases of BALF-GM and 4 787 cases of serum GM specimens suspected of fungal infection in patients admitted to Affiliated Hospital of Jining Medical University from January 2016 to June 2018 were enrolled and analyzed. All patients were classified into clinically confirmed IPA, clinically diagnosed IPA, suspected IPA and excluded IPA according to the classification standard of Expert consensus on diagnosis and treatment of pulmonary mycosis. The coincidence rate of BALF and serum GM test results with clinical diagnosis was analyzed. Receiver operating characteristic (ROC) curve was performed, and the diagnostic value of BALF and serum GM test alone or in combination for IPA was evaluated. Subgroup analysis was performed in patients with normal or abnormal immune function, and the sensitivity and specificity of BALF and serum GM test were compared separately or jointly. Results The positive rate of BALF-GM was 46.8% (89/190), and 10.4% (497/4 787) on serum GM. Among them, 156 patients were both tested on BALF and serum GM. There were 44 cases with both positive in BALF and serum GM, the coincidence rate of clinical definite was 93.2% (41/44). There were 34 cases with positive BALF-GM and negative GM test in serum, and the coincidence rate of clinical definite was 64.7% (22/34). There were 56 cases positive in serum GM and negative in BALF-GM, and the coincidence rate of clinical definite was 48.2% (27/56). BALF and serum GM tests were both negative in 22 cases, and the coincidence rate of exclusion diagnosis was 90.9% (20/22). ROC curve analysis showed that the diagnostic value of BALF-GM test combined with serum GM test for IPA was better than that of BALF-GM test or serum GM test alone [area under ROC curve (AUC) 0.992 vs. 0.983, 0.976]. The sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were 95.3%, 87.0%, 93.2% and 90.9%, respectively. Subgroup analysis showed that among 89 patients with positive BALF-GM test, 85 cases (95.5%) had normal immune function and 4 cases (4.5%) had unknown condition. Among 497 patients with positive serum GM test, 12 cases (2.4%) had normal immune function, 372 cases (74.9%) had abnormal immune function and 113 cases (22.7%) were uncertain. It was shown by ROC curve analysis that the sensitivity of positive BALF-GM test in diagnosis of IPA in patients with normal immune function was higher than that of positive serum GM test (95.6% vs. 88.9%), while the sensitivity of positive serum GM test in patients with abnormal immune function was higher than that of positive BALF-GM test (91.8% vs. 89.9%). Conclusion The results of BALF and serum GM tests are in good agreement with clinical diagnosis, and the combined detection of BALF and serum GM is more valuable for IPA diagnosis than single detection, especially for patients with unknown immune function.