A retrospective narrative review of kidney stone composition in patients from two race groups having different stone occurrence profiles: a hypothesis-generating study of stone formation based on stone composition alone.

Studies of kidney stone composition have been ongoing for over 60 years but have failed to establish a compelling relationship between stone composition and stone pathophysiology. As such, a focussed investigation involving a different approach to those employed previously is warranted. The stone anomaly in South Africa provides an ideal scenario for such an investigation. In this country, stones occur in the white population group (W) but historically they have been relatively rare in their black African counterparts (B). Accounting for this anomaly in terms of urine composition has not been convincingly described. We undertook to achieve better understanding of this anomaly and the riddles of stone formation in general by comparing compositions and their proportionate presence in stones formed in these groups. Composition of 3783 stones were retrieved from stone composition files in our research laboratory. Each was analysed by x-ray powder diffraction and was classified according to the race and gender of the patients from whom they were obtained and according to the presence of a main component, defined as constituting ≥ 50% of the overall composition. Chi-squared tests were used to test differences in proportions between the groups and for correlation p-values. Results were adjusted for multiple-comparisons using a Bonferroni correction of the p-values. Phi coefficients and p-values were calculated using functions from the R-library. Correlation tables were generated using the R library ‘flextable’, while correlation plots were generated using the Corrplot function from the R library ‘DescTools’. Statistical significance was assigned to correlation coefficients ≥ 0.25, p ≤ 0.05. Systematic literature searches were conducted to identify compositional, clinical and pathophysiological associations. The proportion of calcium oxalate dihydrate (COD) stones was significantly lower in B than in W (7.2% vs. 18.8%, p = 0.006). Uric acid dihydrate (UAD), sodium urate and ammonium acid urate stones were common in W, but were absent in B. Intergender differences and correlation analyses were unremarkable as they were in alignment with global observations. Literature searches highlighted weaker crystal-cell adhesive forces associated with COD compared to the monohydrate (COM) and relatively slower transformation kinetics of the former to the latter. Associations between UAD, type 2 diabetes, obesity, metabolic syndrome and Randall’s plaque were also featured in these searches. We speculate that smaller proportions of COD in B might be indicative of a mechanism involving the transformation kinetics of COD to COM and the presence of weaker crystal-cell adhesion forces associated with COD. We also hypothesize that the absence of UAD, NaUr and AAU in B is indicative of a loss of promotive capacity for the formation of Randall’s plaque, substantiated by an apparent lower incidence of the latter in B, and a concomitantly lower stone occurrence rate in this group. Finally, we suggest that the South African stone anomaly might be related to the reported lower metabolic syndrome occurrence in B.