Diagnostic test accuracy of the Fungitell serum (1→3)-β-D-glucan assay for the diagnosis of Pneumocystis jirovecii pneumonia: a systematic review and meta-analysis

Background: The diagnosis of Pneumocystis jirovecii pneumonia (PCP) can be challenging because of diagnostic tests that are imperfect and/or invasive. The Fungitell serum (1 → 3)-β-D-glucan (BDG) assay is a noninvasive blood test studied for PCP; however, the manufacturer-recommended cut-off of 80 pg/mL is not well validated for this disease.

Objectives: We conducted a systematic review and meta-analysis to determine the diagnostic test accuracy of the Fungitell BDG assay for the diagnosis of PCP.

Data sources: A search strategy of MEDLINE and Embase from a previous meta-analysis on BDG was updated to 31 January 2024.

Study eligibility criteria: Observational studies.

Participants: Patients with risk factors for PCP. TEST: Fungitell BDG assay.

Reference standard: One or more of lung biopsy, bronchoalveolar lavage, induced sputum, or nasopharyngeal swab specimens tested for PCP by histopathology, microscopy using immunofluorescence or staining, or PCR.

Assessment of risk of bias: The Quality Assessment of Diagnostic Accuracy Studies-2 tool.

Methods of data synthesis: Diagnostic test accuracy data of the Fungitell serum BDG assay across all reported cut-offs were pooled by meta-analysis. We then evaluated a categorical approach using <80 pg/mL as a rule-out threshold and ≥400 pg/mL as a rule-in threshold.

Results: A total of 26 articles were included comprising 5111 patients and 1150 PCP cases. At the conventional cut-off of 80 pg/mL, the overall pooled sensitivity and specificity were 83.5% (95% 95% CI, 72.8-90.6) and 75.5% (95% CI, 66.0-83.0), respectively. At a pretest probability of <20% and a BDG <80 pg/mL, the post-test probability would be <5% (negative predictive value > 95%). At 400 pg/mL, sensitivity was reduced to 63.5% (95% CI, 45.8-78.1) with specificity increased to 93.6% (95% CI, 88.6-96.5). At a pretest probability of 47.5%, a BDG >400 pg/mL would have a post-test probability of >90%.

Discussion: A categorical approach using <80 pg/mL to rule-out and >400 pg/mL to rule-in PCP may allow for a more nuanced interpretation based on pretest probability. More accurate estimates of pretest probability and further external validation are required.