C-reactive protein in bronchoalveolar lavage fluid as a potential biomarker of bacterial bronchopneumonia in dogs

Diagnosis of canine bacterial bronchopneumonia (bBP) is challenging and mostly based on bronchoalveolar lavage fluid (BALF) cytology and bacterial culture (BC) results. However, bacterial contamination of BALF is frequent, intracellular bacteria are not always visualized on BALF cytology, and BC results may not be related to antimicrobial requirements. Based on local expression of C-reactive protein (CRP) by alveolar macrophages, we aimed to evaluate whether CRP was detectable in canine BALF, and to determine whether BALF-CRP detection could assist in differentiating bBP from non-bacterial bronchopneumopathy (nbBP) in dogs. We used surplus BALF material collected from 24 dogs. Final diagnosis of bBP (<i>n</i> = 3) versus nbBP (<i>n</i> = 21) was retrospectively achieved by 2 board-certified internists. CRP was measured in BALF using an Idexx Catalyst One analyzer. Agreement between a detectable BALF-CRP (≥1 mg/L) and a diagnosis of bBP was assessed using the Cohen kappa coefficient. Five dogs had detectable BALF-CRP, including 3 dogs with bBP and 2 dogs with nbBP. Substantial agreement (κ = 0.7) was found between BALF-CRP and the final diagnosis. BALF-CRP had a test sensitivity of 100% (95% CI [29, 100]) and a specificity of 90% (95% CI [69, 98]) for bBP. Only a fair and not significant correlation was found between plasma CRP and BALF-CRP concentration (rho = 0.3; <i>p</i> = 0.3). Our results suggest that CRP is detectable in canine BALF. The BALF-CRP may be a potential in-house biomarker for bBP in dogs, without being influenced by the plasma CRP concentration.