Exploratory Multivariate Analysis of Mediator Organization in Canine Platelet-Rich Gel Under NSAID Exposure

Platelet-rich gel (PRG) is a fibrin-based biobased biomaterial generated by activating platelet-rich plasma (PRP), yet its biological characterization has commonly relied on univariate measurements of isolated mediators. This study aimed to define the multivariate biological organization of PRG and related hemocomponents (PRP, chemically induced platelet lysate (CIPL), and plasma) in a canine model under single exposure to non-steroidal anti-inflammatory drugs (NSAIDs). In a randomized crossover design (n = 6 dogs), hemocomponents were produced at baseline (0 h) and 6 h after administration of carprofen or firocoxib. Platelet and white blood cell (WBC) counts, growth factors (platelet-derived growth factor-BB (PDGF-BB) and transforming growth factor beta-1 (TGF-β1)), and cytokines (tumor necrosis factor alpha (TNF-α), interleukin-1 beta, and interleukin-10) were integrated using linear mixed-effects modeling, principal component analysis (PCA), and hierarchical clustering. PRG was derived from a leukocyte-poor PRP precursor with moderate platelet enrichment (~1.6-fold vs. whole blood) and a marked WBC reduction (~8-9-fold). In mixed-effects modeling, hemocomponent type significantly influenced the PDGF-BB:TNF-α log-ratio, with PRG (estimate -1.12; 95% CI -1.34 to -0.90) and plasma (-2.06; 95% CI -2.28 to -1.84) lower than PRP, while CIPL did not differ. Time and NSAID effects were not supported. PCA identified two orthogonal axes explaining 61.3% of total variance (PC1 = 43.7%, PC2 = 18.6%), separating a platelet/trophic dimension (log(PDGF-BB), log(TGF-β1), platelet count, PDGF-BB:TNF-α log-ratio) from an inflammatory dimension (log(TNF-α), log(IL-1β)). Overall, hemocomponent composition emerged as the primary determinant of mediator organization, supporting the interpretation of PRG as a structured, biomaterial defined by coordinated mediator networks.