The Role of Altered Mineral Metabolism in Kidney Stone Disease: Is Vitamin D the Key?

Kidney stones are common, affecting 10%-15% of the population.¹ The foundation for the current therapy to prevent kidney stones is based on the principle that supersaturation of solutes in the urine predisposes to crystal and ultimately stone formation. One or more risk factors for kidney stone formation are frequently identified from 24-hour urine studies. Our therapies, therefore, have depended on interventions to alter the solute content and concentration of the urine through lifestyle and/or pharmacologic means with the ultimate goal of decreasing stone formation.²

In fact, there are several limitations of current interventions aimed at the prevention of kidney stones. While these interventions have resulted in some success, the current outcomes for stone formers are far from ideal. Sometimes the 24-hour urine studies do not point toward an obvious risk factor for stone formation. A possible explanation is that urine studies are frequently obtained weeks to months after an episode of renal colic, at a time when a stone former has already altered diet and fluid intake. Moreover, while individuals with familial hypocalciuric hypercalcemia (FHH) are not hypercalciuric, up to 30% will develop kidney stones.^3,4 Sometimes the stone type does not seem to correspond to the urine findings. An example of this phenomenon is the occurrence of calcium oxalate stones in individuals whose only “abnormal” urine finding is hyperuricosuria.⁵ Often, “correction” of a presumed risk factor does not have an impact on recurrent stone formation. A good example is the recently published NOSTONE trial examining the recurrence of kidney stones during treatment with different doses of hydrochlorothiazide, a standard approach to hypercalciuric kidney stone disease based on our understanding of renal physiology.⁶ The surprising result showing no significant impact of hydrochlorothiazide on symptomatic stone recurrence at any dose has been the subject of considerable debate, with questions about design, failure to control for other parameters, and the like.^7,8 Perhaps, however, what the NOSTONE study demonstrated is that not all hypercalciuric calcium kidney stone formers form stones for the same reason. Finally, there is no convincing explanation for why some individuals with highly supersaturated urine do not develop kidney stones. These observations suggest that our current approach to the evaluation of kidney stone disease and our interventions are treating a phenotype, but not the underlying pathogenesis. The path forward in treating kidney stone disease needs to move forward from our current physical chemistry approach and embrace newer methods for identifying the pathophysiology of this complex phenomenon.

What Does This Important Study Show?

Lovegrove et al⁹ proposed to identify genetic variants that increase kidney stone risk by altering calcium and phosphate homeostasis, based on their prior work¹⁰ demonstrating increased stone risk in individuals with higher serum calcium and lower serum phosphate levels. They first performed a genome-wide association study (GWAS) using the UK Biobank comparing 11,186 individuals with kidney stones and 390,488 non–stone formers and a meta-analysis of the FinnGen R12 database of 12,981 stone formers and 486,185 non–stone formers. They narrowed down the 79 independent risk signals across 71 genomic loci identified from these analyses to focus on sites that linked serum calcium and phosphate concentrations to kidney stone formation. They identified likely causal gene variants associated with alterations in either calcium or phosphate concentrations using a cis-Mendelian approach to explore the chromosomal region around the previously identified variants of interest. In this technique, the investigators compare the occurrence of alterations in calcium or phosphate concentrations in individuals who express or who do not express a specific gene variant. They followed this by performing colocalization studies using the GWAS data to link gene variants that impact calcium or phosphate concentrations with variants associated with kidney stones. Of their initial 79 loci, they identified 3 common variants: a predicted transcription binding site for the gene DGKD, which encodes for diacyl glycerol kinase δ, a component of the signaling cascade activated by the calcium sensing receptor (CaSR), associated with an increase in serum calcium; an intergene site upstream of the gene SLC34A1, the type IIa sodium phosphate cotransporter responsible for reabsorption of filtered phosphate by the proximal tubule, associated with low serum phosphate; and an intergene site downstream of the gene CYP24A1, which codes for the 24-hydroxylase enzyme responsible for breakdown of 1,25-dihydroxyvitamin D, associated with an increase in serum calcium. Variants in CYP24A1 and SLC34A1 cause the autosomal recessive disorder called idiopathic infantile hypercalcemia types 1 and 2, respectively, while those in CaSR signaling result in FHH. Returning to the UK Biobank data, the authors found that the genomic variations in DGKD and SLC34A1 were associated with lower serum phosphate and higher serum calcium while the those in CYP24A1 were associated with higher serum calcium.

To characterize the effect of the DGKD variants further, the authors identified DGKD variants associated with kidney stone disease in 2 additional databases, the Genomics England 100,000 Genomes Project and DiscovEHR. They then expressed DGKD variants in cells that also expressed the CaSR and demonstrated loss of CaSR response to increasing calcium concentrations. To identify the mechanism for the loss of function, the authors modeled the predicted DGKδ protein variants, revealing alterations in oligomerization, appropriate intracellular localization, protein binding, and ATP binding, all of which would reduce function, leading to impaired CaSR signaling. The mechanisms for the alterations in phosphate and calcium levels for the CYP24A1- and SLC34A1-associated variants were not evaluated. The specific locus near CYP24A1 is thought to decrease the expression of this gene by decreasing promoter activity and is strongly associated with parathyroid hormone concentration.¹¹ The specific locus near SLC34A1 is actually an expression quantitative trait locus for RGS14, that is, a noncoding site that regulates the expression of the protein RGS14, which in turn regulates the function of SLC34A1, and has been previously implicated in the regulation of serum phosphate concentration.¹²

Overall, the authors concluded that the identified variants that produce only modest changes in serum calcium and phosphate were responsible for 11%-19% of kidney stone cases. They calculated that interventions to decrease serum calcium by 0.32 mg/dL and/or increase phosphate by 0.5 mg/dL could reduce stone risk by as much as 90%.

How Does This Study Compare With Prior Studies?

The high heritability of kidney stone disease has prompted numerous studies to determine genetic causes of kidney stones.^13-15 These studies have identified clinical clues that would suggest an inherited cause of kidney stones, such as young age of onset and high recurrence rate. It is increasingly recognized that genetic variants identified in those high-risk individuals may also contribute to the development of kidney stones in the general population of kidney stone formers when combined with additional risk factors, including lifestyle factors, underlying chronic diseases, and environmental influences.¹⁰ Several recently published GWAS have identified an expanding list of loci associated with kidney stones, including the present study, which added 33 new loci. Many of these genomic sites have been identified in studies using different biobanks and different approaches, increasing the confidence in a causal role for these gene variants.^16-19

The current study confirms a role for calcium and phosphate homeostasis in kidney stone disease and the specific association of the genes that regulate calcium and phosphate, the CaSR, SLC34A1, and CYP24A1. The authors went a step further in determining the effect of one of the gene variants (DGKD) on CaSR signaling and ultimate function, linking serum calcium to kidney stone risk. This observation marks something of a departure from our usual approach to kidney stone disease, which is highly urine chemistry–centric. The variants in SCL34A1 and CYP24A1 would result in higher calcitriol levels, indirectly in the case of SLC34A1 by decreasing serum phosphate and decreasing FGF23 and directly for the CYP24A1 variants by impaired inactivation of calcitriol, leading to increased gut calcium absorption. The mechanism of the DGKδ variant leading to CaSR loss of function is not immediately apparent, as a loss of function in the CaSR is associated with the syndrome of FHH and hypocalciuria. A prior study, however, has demonstrated that the DGKD risk allele rs838717 is associated with hypercalciuria in males but not females.¹⁹ The authors of that study did not demonstrate an association with higher calcium or lower phosphate, in contrast to the current study under discussion, where a slightly lower serum phosphate was associated with this variant, which would be expected to increase calcitriol levels. An additional mechanism for increased calcitriol production with the DGKD risk allele is suggested by a recent in vitro study demonstrating that activation of the CaSR in HEK-293 cells increases CYP27B1 activity.²⁰

The final common pathway for these 3 identified variants that associate with kidney stone disease appears to be enhanced-activation vitamin D (Fig 1). The identification of the effects of the variant on protein interactions and protein functions in this study is a major advance, suggesting that it might be possible to target these pathways in the interest of a more precision-based approach to the treatment of kidney stone disease. In fact, the authors show that cinacalcet, a CaSR modulator, ameliorates the defect in CaSR signaling produced by the DGKD variant. Whether this intervention would lead to diminished stone formation in this particular population needs further study.

Figure 1.

Mechanisms for vitamin D–mediated hypercalciuria in kidney stone disease.

What Are the Implications for Nephrologists?

Studies such as these raise awareness that the current phenotypic-based approach to kidney stone disease centered on analysis of urinary contents can be significantly enhanced by a more pathophysiologic, mechanistic-based approach. For the practicing nephrologist, identifying serum calcium and phosphate as risk factors for kidney stones is a reminder that urine calcium is a function of both filtered load and renal tubular handling and that reliance on urine studies is not sufficient to understand kidney stone disease. The findings from this study also highlight the fact that our phenotypic classification of kidney stone risk by findings such as hypercalciuria can obscure the recognition that many pathophysiologic conditions can produce hypercalciuria. Finally, the identification of genetic variants that do not result in protein variants but do result in altered regulation of protein expression also emphasizes the heterogeneous nature of the pathophysiology of kidney stone disease and the power of genetic analysis to increase our understanding. The suggestions by the authors that decreasing serum calcium concentrations and increasing phosphate concentrations to prevent kidney stones are intriguing but will require extensive clinical investigation, as acknowledged by the authors, as therapies targeting calcium and phosphate concentrations per se could have adverse effects on bone health while raising serum phosphate could increase cardiovascular risk.^21,22