Exploratory Diagnostic Performance of On-Admission Soluble CD40 Ligand for Distinguishing Acute Pulmonary Embolism from Hospitalization-Requiring Community-Acquired Pneumonia: A Single-Center Observational Study

Background/Objectives: Acute pulmonary embolism (PE) and hospitalization-requiring community-acquired pneumonia (CAP) may present with overlapping clinical, laboratory, and radiological features. Soluble CD40 ligand (sCD40L) is a platelet-derived thrombo-inflammatory mediator that may be influenced by both thrombotic and inflammatory processes. This study retrospectively compared on-admission serum sCD40L concentrations between selected hospitalized patients with established acute PE and selected patients with hospitalization-requiring CAP. Methods: This single-center retrospective exploratory comparative biomarker study included 82 hospitalized adults: 48 with computed tomography pulmonary angiography (CTPA)-confirmed acute PE and 34 with hospitalization-requiring CAP defined using CURB-65-supported admission criteria. Stored admission serum samples were used for sCD40L measurement. Between-group comparison was the primary analysis; receiver operating characteristic (ROC) analysis was performed as a secondary exploratory description of the apparent within-sample discriminatory signal. Results: sCD40L was higher in acute PE than in hospitalization-requiring CAP (median 821.3 vs. 629.0 pg/mL; p < 0.001). ROC analysis demonstrated a strong exploratory within-sample discriminatory signal (AUC 0.951, 95% CI 0.905–0.997). After excluding five patients with recorded antiplatelet or rivaroxaban exposure, the apparent signal remained similar (AUC 0.945; bootstrap 95% CI 0.891–0.984), and sCD40L remained associated with PE in a Firth-penalized model adjusted for platelet count and COPD (OR 3.39 per 50 pg/mL, 95% CI 2.00–7.71; p < 0.001). Conclusions: In this retrospective selected two-group comparison, on-admission serum sCD40L concentrations were higher in established acute PE than in hospitalization-requiring CAP. ROC-derived estimates should be interpreted only as apparent within-sample discrimination and not as a replacement for D-dimer, clinical probability assessment, or imaging-based PE diagnosis. Prospective validation in unselected suspected-PE cohorts is required before any diagnostic or clinical use can be considered.