Clustering patients with gout based on comorbidities and biomarkers: a cross-sectional study


Fatima Alduraibi, Mohammad Saleem, Karina Ricart, Rakesh P. Patel, Alexander J Szalai, Jasvinder A. Singh

The University Of Alabama At Birmingham, Birmingham


Background / Purpose: Our understanding of various factors and steps involved in the pathogenesis of gout and comorbidities and their obvious links have improved. The current study aimed to identify a cluster of patients with gout based on comorbidities and/or biological markers (inflammatory cytokines and oxidative stress markers) to identify distinctive phenotype clusters in a cross-sectional single-center clinic cohort.

Methods: This is a cross-sectional study with patients with clinically diagnosed gout who were enrolled in the university of alabama at birmingham (UAB) rheumatology arthritis database and repository (RADAR) from january 2018 to december 2019. Demographic, clinical, and laboratory data were collected. Serum and plasma were assayed for key inflammatory markers and oxidative stress pathway metabolites. Renal function tests and other clinical end points were also assessed. Hierarchical cluster analyses were performed to group clinical data and/or biological markers variables with close proximity and to obtain homogenous clusters of individuals in this cohort. We used the analysis of variance (anova test) (continuous) and chi-square (categorical) to calculate statistics between subgroups clusters.

Results: A total of 88 gout patients were enrolled. All patients except one (score of 7; threshold of ≥ 8) met the gout classification criteria of the American College of Rheumatology/EUropean League Against heumatism (ACR/EULAR). Of those, 74% (n = 65) were male, 49% (n = 43) were caucasian, 47% (n = 41) were african american (AA) and 57% (n = 50) had a mean body mass index of > 30 kg/m2. Sixteen percent (n = 14) had tophaceous gout. Three subgroups clusters (c1–c3) among gout patients based on clinical data were identified: cluster 1 (c1, n = 24) was labeled “HTN and dyslipidemia.” Of these, 27% were overweight, 54% had hypertension( HTN), and 67% had dyslipidemia and a longer gout disease duration. Cluster 2 (c2, n = 25) was labeled “tophaceous gout with multiple comorbidities.” Of these, 29% had class 1 obesity, 100% had HTN, 68% had dyslipidemia, 64% had nephrolithiasis, 56% had CKD, 48% had diabetes (DM), and 44% had heart disease. They were the oldest at the time of enrollment in the study. Cluster 3 (c3, n = 39) was labeled “tophaceous gout with morbid obesity.” Of these, 44% had class 2 obesity, 90% had HTN, 51% had dyslipidemia, 49% had DM, 33% had CKD, and 23% had tophaceous gout. They were the oldest at the time of the gout diagnosis and had a shorter gout duration. The following oxidative and inflammatory cytokines markers were higher in cluster 1 (c1): 3-nitrotyrosine; tumor necrosis factor alpha, c-reactive protein, interleukin 1 beta and derived growth factor AA; cluster 2 (c2): were aldosterone; and cluster 3 (c3): carbonyl, nitrite, 8-isoprostane, and interleukin 6.

Conclusion: Together, these results suggest that measuring oxidative stress and inflammatory cytokine levels can be a unique and valuable adjunctive tool for assessing patients with gout with comorbidities. These associations need further study in larger samples.