Background/Purpose: Persistent subclinical inflammation from monosodium urate (MSU) crystals is present in gout and asymptomatic hyperuricemia (AH). The underlying mechanism is poorly understood. This study analyzes the articular inflammatory proteome linked to MSU crystal deposition in hyperuricemic patients without clinically active inflammation.

Methods: Observational, cross-sectional study. 100 participants with either AH (hyperuricemia as serum urate ≥7mg/dl) [n=77] or intercritical gout (2015 ACR/EULAR criteria, hyperuricemia, untreated) [n=23] underwent clinical assessment, blood testing, and lower limbs ultrasonography. In the case of effusion, synovial fluid (SF) was obtained (n=81). Samples underwent polarized microscopy for microcrystals (MSU, calcium pyrophosphate –CPP) and leukocytes (using a Neubauer chamber). The SF proteome was studied using Olink Inflammation panel in cell extracts from samples with enough volume after centrifugation (n=60). For these cases, paired serum proteome assessments were intended (n=28). Comparisons were conducted based on microcrystal deposition, identifying differential expression analysis of proteomics using Principal Component Analysis (PCA), represented by volcano plots. Also, we explored predictive models using SF features (proteome and leukocytes) to discriminate microcrystal deposition.

Results: Out of 100 participants, 81 had SF samples available for polarized microscopy; MSU and CPP crystals were detected in 5 (6.2%) and 11 samples (13.6%), respectively, while the others showed no crystals (65, 80.2%). Leukocyte counts were significantly higher in samples with MSU (640/mm3, 245-875) compared to CPP (55/mm3, 43-203) or no crystals (40/mm3, 20-70), p< 0.001. Phagocytosis was seen in 32.4% (0-61.8%) and 8.3% (0-53.3%) of MSU and CPP samples [p=0.743]. SF leukocytes showed no correlation to serum urate, total leukocytes, or acute phase reactants.Sixty samples underwent proteomic analysis (3 with MSU, 8 with CPP, 49 with no crystals). Figure 1 shows the volcano plots comparing MSU versus CPP (left) and versus no crystals (right). Both comparisons yielded similar results: SF containing MSU crystals showed an increased relative expression of several inflammatory proteins. These proteins differentiated samples among the three groups (71.3% of explained variances [Figure 2]). Among different predictive models, lightGBM with variable selection by KBest showed the best performance (accuracy 0.83, AUC 0.8182). Figure 3 shows the 20 features considered most relevant by the previously mentioned optimal model for prediction.Only IL-8 significantly correlated between SF and serum expression (r=+0.44, p=0.019) in paired sera [n=28].

Conclusion: Subclinical inflammation associated with MSU crystal deposition involves elevated leukocyte counts and a distinctive proteomic profile, which can distinguish the presence of microcrystals. Our data also suggest that during asymptomatic stages, the subclinical inflammation appears localized and confined to the joint, the relevance of which requires further investigation.

Figure 1. Volcano plot of differentially expressed proteins in SF samples (to the right favors MSU; to the left favors no crystals [left] or CPP [right]).

Figure 2. Left, PCA of statistically significant inflammatory proteins in synovial fluid, comparing the three patient groups, showing the distribution of patients’ samples across dimensions 1 (dim1) and dim2 [red=no crystals; green=MSU; blue=CPP]. Middle, percentages of explained variances by dimensions. Right, protein distributions to dim1 and dim2 and their contribution.

Figure 3. SF individual features and role in distinguishing the presence of microcrystals, using the lightGBM decision tree-based algorithm with KBest variable selection.

« Back to ACR Convergence 2025

ACR Meeting Abstracts - https://acrabstracts.org/abstract/a-proteomic-analysis-of-subclinical-synovial-inflammation-associated-with-monosodium-urate-crystal-deposition/