Bronchoalveolar lavage fluid polymerase chain reaction for invasive pulmonary aspergillosis among high-risk patients: a diagnostic meta-analysis

Polymerase chain reaction (PCR) assays are perceived to facilitate the diagnosis of fungal infections. However, due to lack of standardization, the value of bronchoalveolar lavage (BAL) fluid PCR in diagnosis of invasive pulmonary aspergillosis (IPA) remains unclear. We conducted a systematic meta-analysis to evaluate the accuracy of BAL fluid PCR in IPA diagnosis among high-risk patients. All studies involving patients at risk for IPA were included. The sensitivity, specificity, positive and negative likelihood ratios of BAL fluid PCR were summarized for diagnosis of proven/probable IPA, or proven IPA only. Potential heterogeneity was assessed by subgroup analyses and meta-regression. Forty-one studies involving 5668 patients were analyzed. The summary sensitivity, specificity, positive and negative likelihood ratios of BAL fluid PCR for proven/probable IPA were 0.75 (95% CI = 0.67-0.81), 0.94 (95% CI = 0.90-0.96), 11.8 (95% CI = 7.7-18.1) and 0.27 (95% CI = 0.20-0.36), respectively. Whereas for proven IPA only, sensitivity and specificity were 0.91 (95% CI = 0.68-0.98) and 0.80 (95% CI = 0.74-0.85) in fourteen studies involving 2061 patients. Significant heterogeneity was present due to the underlying disease, antifungal treatment and differences in DNA extraction techniques and choice of PCR assay. Compared to patients with hematological malignancies (HM) and hematopoietic stem cell/solid organ transplantation (HSCT/SOT), sensitivity was higher in the population with disease such as chronic obstructive pulmonary disease, solid tumor, autoimmune disease with prolonged use of corticosteroids, etc. (0.88 vs. 0.68, P