Genetic and Clinical Characterization of a Large Cohort with Suspected Monogenic Stone Disease

Abstract

BACKGROUND:: Urinary stone disease with a clear genetic cause, monogenic stone disease (MSD), is increasingly recognized as a significant proportion of the total population. When MSD is suspected, genetic testing provides a firm diagnosis that can alter management and treatment. Here we present testing results from a large cohort with suspected MSD.

METHODS:: Subjects with features suggestive of MSD (early onset, family history, frequent stones, nephrocalcinosis [NC], and/or CKD) were recruited by the Rare Kidney Stone Consortium and genotyped for up to 160 known or candidate MSD genes via a targeted massively parallel sequencing (tMPS) panel. We compared clinical and biochemical features between genetically resolved MSD and unresolved individuals.

RESULTS:: Of 426 families (657 patients) enrolled, 145 (34%) were resolved with identified disease associated variants in 22 known MSD genes. Ninety-nine families were biallelic, 37 monoallelic, and 2 digenic. An additional 21 of the 231 screened family members were resolved. Genes identified in 10 or more families were: AGXT, HOGA1, SLC34A3, CYP24A1, SLC3A1, and CLCN5. Compared to the unresolved group, MSD probands had a lower baseline and last visit estimated glomerular filtration rate (eGFR), earlier age of stone presentation, and more stone events and procedures/year of life. The resolve rate was higher in those less than 16 years, and NC was seen earlier in the MSD group. Overall, NC was a risk factor for lower eGFR. Among the specific disorders, primary hyperoxaluria patients had the earliest age of stone and NC diagnosis, and as expected, the highest urinary oxalate level.

CONCLUSIONS:: Our study emphasizes the value of selecting patients enriched for factors associated with MSD, and comprehensive genetic testing to achieve a high yield of genetic diagnoses. Significant clinical and biochemical characteristics of MSD patients were defined. A definitive MSD diagnosis facilitates individualized management and strategies to delay disease progression in probands and affected family members.

Keywords: CKD; genetics and development.

Figure 1.. Flowchart showing the design of the study with the genetic results for the resolved families

The study included 426 families that in addition to the proband included 231 family members. From tMPS testing, 34% of families were resolved, and 9% of additional family members were proven genetically affected. Individuals were divided into six phenotypic groups based on initially available clinical and biochemical data: Hypercalciuria, Hyperoxaluria, Cystinuria, NC/NL without urine abnormalities, Low Molecular Weight Proteinuria (LMWP), or Others. The number of patients resolved in each phenotypic group are shown (Supplemental Figure S2 shows the details of the genes identified in each group).NPVD: no pathogenic variant detected; NC: nephrocalcinosis; NL: nephrolithiasis.

Figure 2.. A summary of the associated genes in the resolved biallelic (left) and monoallelic (right) families.

A. In the whole cohort, 103 families were biallelic with 15 different genes identified. The number of families due to each gene is indicated and shading shows genes identified with similar disorders: blue, PH^–; yellow, cystinuria^,–[; green, phosphate transport disorders^–, and purple, renal tubular acidosis^–, plus additional diseases due to pathogenic variants to: ABCC6, APRT, CLDN16, and KCNJ1. B. Similar data is shown for the 40 families with monoallelic inheritance with 12 different genes identified. Genes not also detected biallelically included: grey, DD (denoted with an asterisk as X-linked, CLCN5 and OCRL. plus, the X-linked HPRT1. Autosomal loci include: ADCY10^,, ALPL, HNF4A, PKHD1. C. Descriptions of the MSD, genes, first reference, inheritance pattern, and phenotypes. Disease – APRT, Adenine phosphoribosyltransferase; HHRH, hereditary hypophosphatemic rickets with hypercalciuria; PH, primary hyperoxaluria; Inheritance - AD: autosomal dominant; AR: autosomal recessive; XLR: X-linked recessive. Disease phenotypes - dRTA: renal tubular acidosis; distal FS: Fanconi syndrome; HCM: hypercalcemia; HCU: hypercalciuria; LMWP: low molecular weight protein; NC: nephrocalcinosis; NL: nephrolithiasis; PTH: Parathyroid hormone; Vit D: vitamin D.

Figure 3.. Comparison of eGFR between the resolved and unresolved groups at the first and last visit.

The resolved group had a lower eGFR at both the 1st (P=0.02) and last (P=0.01) measurement.

Figure 4.. Analysis of nephrocalcinosis (NC).

A. NC was found at an earlier age in the resolved vs. unresolved groups (P=0.003), B. Analysis of the whole cohort (resolved and unresolved) showed lower eGFR in the NC compared to the no NC group, with values at the mean age of 35.16y (indicated with a vertical dashed line) of 63.4ml/min/1.73m2 and 81.4 ml/min/1.73m2, respectively (P = 0.009). The rates of eGFR decline did not differ between the NC and no NC groups (P=0.382).

Figure 5.. Three interesting families analyzed as part of the study.

A. Digenic family (OTH4) with pathogenic variants in both SLC34A1 (c.1416+5G>A) and SLC34A3 (c.1336–2A>C) detected in the proband. She had hypercalciuria, bilateral NC, recurrent kidney stones starting at 6y that required multiple surgical interventions, and stage 4 CKD at 22y. There were two other possibly affected family members (mother and brother) who had recurrent kidney stones, but samples were not available for genetic testing. B. In the proband with an APRTd phenotype in pedigree AP-5, a pathogenic variant to APRT (c.289_290del; p.Leu97fs) was detected plus 3 variants in APRT IVS4, a 16bp deletion and two substitutions (c.401–20C>G, c.401–21A>T, c.401–40_401–25del), that are predicted to disrupt the branch point. To determine phase, a 700bp region of the APRT gene from exon 3, before the known pathogenic FS deletion, and with the reverse primer binding on the deleted 16bp intronic region was amplified. Sanger sequencing of the PCR product without intronic deletion showed that the known pathogenic FS deletion was present, confirming the pathogenic variants are in trans.