ABSTRACT
Background Molecular mechanisms of kidney stone formation remain unknown in most patients. Previous studies showed high heritability of nephrolithiasis, but data on prevalence and characteristics of genetic disease in unselected adults with nephrolithiasis are lacking.
Methods We performed whole exome sequencing in 787 participants of the Bern Kidney Stone Registry, an unselected cohort of adults with ≥ 1 past kidney stone episode (KSF), and 114 non- stone-forming individuals (NKSF). A total of 34 established nephrolithiasis genes were analyzed and variants assessed according to ACMG criteria. Pathogenic (P) or likely pathogenic (LP) variants were considered diagnostic.
Results Mean age of KSF was 47±15 years, and 18 % were first time KSF. A Mendelian kidney stone disease was present in 2.9% (23 of 787) of KSF. The most common genetic diagnoses were cystinuria (SLC3A1, SLC7A9; n=13), Vitamin D-24 hydroxylase deficiency (CYP24A1; n=5) and primary hyperoxaluria (AGXT, GRHPR, HOGA1; n=3). 8.1% (64 of 787) of KSF were monoallelic for LP/P variants predisposing to nephrolithiasis, most frequently in SLC34A1/A3 or SLC9A3R1 (n=37), CLDN16 (n=8) and CYP24A1 (n=8). KSF with Mendelian disease had a lower age at the first stone event (30±14 years vs. 36±14 years, p=0.003), were more likely to have cystine stones (23.4 % vs. 1.4 %) and less likely to have calcium oxalate monohydrates stones (31.9 % vs. 52.5 %) compared to KSF without genetic diagnosis. The phenotype of KSF with variants predisposing to nephrolithiasis was subtle and showed significant overlap with KSF without diagnostic variants. In NKSF, no Mendelian disease was detected, and LP/P variants were significantly less prevalent compared to KSF (1.8 % vs. 8.1%).
Conclusion Mendelian disease is uncommon in unselected adult KSF, yet variants predisposing to nephrolithiasis are significantly enriched in adult KSF.
INTRODUCTION
Nephrolithiasis is a common global healthcare problem 1. Kidney stones recur frequently and cause substantial morbidity, reduced quality of life and enormous cost 2-4. Most kidney stones are classified as “idiopathic”, indicating that the pathogenesis is unclear. Consequently, undifferentiated dietary and pharmacological preventive measures are initiated, but many patients continue to form stones 5,6. Hence, our current strategies in managing this very common and relapsing disease are clearly inadequate. There is an unmet need for novel diagnostic and therapeutic approaches.
Kidney stone formation is strongly influenced by genetic factors: a positive family history is present in 30-60% of individuals with kidney stones 7, and both twin 8 and genealogy 9 studies revealed a high heritability of nephrolithiasis. More than 30 monogenic, Mendelian causes of nephrolithiasis have been described thus far 10. Recent studies have also highlighted the importance of intermediate effect size / incomplete penetrance variants in increasing the risk for nephrolithiasis11-14. Identification of patients with Mendelian forms of nephrolithiasis remains a major challenge, and no clear consensus exists on clinical parameters to guide genetic testing, although extreme clinical phenotypes usually prompt genetic testing. As a result, many cases of Mendelian forms of nephrolithiasis are missed and incorrectly labelled “idiopathic”. Whole exome sequencing (WES) is widely applied as diagnostic tool for rare diseases and the detection of pathogenic variants in cancer 15,16. In contrast, the diagnostic utility of WES has not been established for most constitutional disorders, such as nephrolithiasis. In small, selected cohorts of early-onset or familial nephrolithiasis, a Mendelian cause was identified in 6.8 - 29.4% of cases 10,17-19. However, the frequency and phenotypic spectrum of genetic disease in sporadic adult-onset nephrolithiasis, by far the most common type encountered in clinical routine, is unknown. To address these important knowledge gaps, we employed WES in a deeply phenotyped, unselected European cohort of 787 adult kidney stone formers (KSF) and in 114 non-stone forming individuals (NKSF).
METHODS
Study cohort
The study was conducted with participants enrolled in the Bern Kidney Stone Registry (BKSR), an. observational cohort of adult kidney stone formers described previously and in the supplemental information20-23. Inclusion criteria for the BKSR are i) written informed consent, ii) age ≥ 18 years, and iii) ≥ 1 past kidney stone episode.
Inclusion criteria for NKSF were i) written informed consent, ii) age ≥ 18 years, and iii) no history of past kidney stones and no evidence of asymptomatic nephrolithiasis or nephrocalcinosis on ultrasound at enrolment.
Whole exome sequencing and variant calling
Isolation of genomic DNA, exome capture, high-throughput sequencing and bioinformatic analysis including joint variant calling using Genome analysis Toolkit v3.8 (GATK) according to GATK best practices recommendations24,25 and annotation using Ensembl Variant Effect Predictor Release 106 (VEP) 26 was performed using established methods, described in more detail in Supplemental Methods. After filtering for predicted consequences and genome aggregation database (gnomAD) minor allele frequency (<1% in all populations) 27, variants in 34 genes 10,17-19,28 previously implicated in Mendelian kidney stone disease were examined (Fig. 1, lower panel). Further variant stratification was performed according to the recommendations of the American College of Medical Genetics and Genomics (ACMG)29 after manual review of phenotype data. Only variants classified as likely pathogenic (LP) or pathogenic (P) applying the ACMG criteria were defined as diagnostic variants leading to a genetic diagnosis, as Mendelian or as LP/P variants predisposing to nephrolithiasis (LP/P variants) depending on evidence available from this study and/or from published case/control studies.11,12,30,31 (Fig. 1 and Supplemental Methods).
Data of 1152 individuals recruited into the BKSR were analyzed. After exclusion of individuals without genetic information available, declined consent or multiple inclusions, the analyzed cohort consisted of 901 individuals (787 stone formers (KSF) and 114 non stone forming controls (NKSF). Before inclusion in the BKSR, NKSF underwent ultrasound imaging to exclude nephrolithiasis (NL) and/or nephrocalcinosis (NC). Genetic analysis was performed in a standardized prioritization pathway in 34 known kidney stone genes. N: number of individuals. Genetic Diagnosis: Likely pathogenic or pathogenic variant according to ACMG criteria. LP/P Variant: monoallelic LP/P variant, predisposing to nephrolithiasis. WES: whole exome sequencing.
Lower Panel: Kidney stone disease gene panel used for genetic analysis. Panel of 34 known nephrolithiasis genes used with their inheritance mode, as accepted for classification as “Mendelian disease” in this manuscript, grouped by phenotypes. AD: autosomal dominant, AR: autosomal recessive, XLR: X-linked recessive, MODY: maturity- onset diabetes of the young, FRTS: Fanconi renal tubular syndrome.
Statistical Analysis
Continuous variables were reported as medians with 25th-75th percentiles or means with standard deviations and categorical variables were reported as counts with percentages, as appropriate. Different statistical methods were employed to analyze differences between independent groups, depending on the nature of the variables involved. Mann-Whitney U test was used for non-normally distributed continuous variables, while Student’s t-test was applied for normally distributed variables. For categorical variables, the Fisher’s Exact Test was implemented. Findings from these analyses were summarized and presented as descriptive tables. Statistical tests were two-sided and a p-value < 0.05 was considered statistically significant. Statistical analyses were performed using Stata, version 16 (Statacorp, TX, USA). Pie-charts and dot plots were generated with GraphPad Prism, version 8.4.3 (GraphPad Software, San Diego, CA, USA).
RESULTS
Study cohort
We performed WES in a total of 901 individuals, including 787 KSF and 114 NKSF (Fig. 1). Clinical characteristics of the study cohort, separated in KSF and NKSF, are shown in Table 1. Mean age ± standard deviation (SD) was 46.6±14.5 years in KSF and 42.4±14.9 years in NKSF, respectively. The majority of KSF were male (71.8 %, n=565), whereas the percentage of men was lower in NKSF (54 %, n=61). KSF were significantly more likely to report a positive family history of kidney stones compared to NKSF (43.6 % vs. 6.8 %, Table 1).
Characteristics are indicated separately for participants with and without kidney stone disease. Categorical variables are described by number of participants N (%), continuous variables are described by their mean (SD) or median (25th-75th percentile). eGFR, estimated glomerular filtration rate; BSA, body surface area; PTH, parathyroid hormone; SD, standard deviation.
* The Non kidney stone former (NKSF) with a diagnostic variant (#5948) has been removed for phenotypic analyses.
Genetic analysis
Among 787 unrelated, unselected adult KSF, we discovered in 23 patients (2.9 %) a Mendelian form of nephrolithiasis, i.e., pathogenic (P) or likely pathogenic (LP) variants in genes associated with Mendelian forms of nephrolithiasis respecting the relevant mode of inheritance (autosomal dominant and/or autosomal recessive and X-linked recessive) (Table 2, Fig. 1, Fig. 2). Mendelian disease was detected in nine of 34 analyzed genes (n= number of patients): AGXT (n=1), ATP6V1B1 (n=1), CYP24A1 (n=5), GRHPR (n=1), HOGA1 (n=1), SLC3A1 (n=9), SLC4A1 (n=1), SLC7A9 (n=4), SLC12A1 (n=1).
(A) Diagnostic yield (Mendelian vs. LP/P variants predisposing to nephrolithiasis)) in kidney stone formers (KSF), (B) Diagnostic yield in non-stone forming controls (NKSF), C) Overview of Mendelian diagnoses in KSF, sorted by phenotype groups, (D) Overview of LP/P variants predisposing to nephrolithiasis in KSF, sorted by phenotype groups.
Acc.N°: RefSeq accession number, ACMG: American College of Medical Genetics, AF: allele frequency, AR: autosomal recessive, AD: autosomal dominant, dbSNP: Reference SNP number of variant, het: heterozygous, hom: homozygous, LP: likely pathogenic, P: pathogenic, tot: total, Novel: mutation detected for the first time in this study/not previously described, NS: nonsense, Family history: N= no, Y=yes, M: male, F: female, mo: months, NL: Nephrolithiasis, NC: Nephrocalcinosis, dRTA: distal renal tubular acidosis, PHPTH: primary hyperparathyroidism, CKD: chronic kidney disease. Stone analysis: CaOx: Calcium oxalate, COM: Calcium oxalate monohydrate, COD: Calcium oxalate dihydrate, CaP: Calcium phosphate, CA: Carbonate apatite, Brushite: Calcium hydrogen phosphate dihydrate, MAP: Magnesium ammonium phosphate (=Struvite), UA: Uric acid.
A total of 66 individuals (8.4%), 64 (8.1%) of which were KSF, presented with LP/P variants not fulfilling our stringent criteria for Mendelian disease, but predisposing for nephrolithiasis (Supplementary Table 1). LP/P variants were detected in nine of 34 genes (n=number of patients): ADCY10 (n=1), CASR (n=4), CLDN16 (n=8), CYP24A1 (n=8), SLC4A1 (n=1), SLC7A9 (n=3), SLC34A1 (n=15), SLC34A3 (n=17) and SLC9A3R1 (n=8). Of these variants, 18 % (12 of 66) were novel, previously unreported in ClinVar or HGMD (Supplementary Table 2).
Prior to WES, a Mendelian form of nephrolithiasis was known or suspected in 18 of 23 individuals (78 %) with a post WES Mendelian diagnosis, despite each case being reviewed by a kidney stone expert. Therefore, in five of 23 individuals (22 %) genetic analysis established a new or corrected a suspected a priori diagnosis. The diagnostic yield was similar between first-time and recurrent stone formers (Supplementary Table 3), and between men and women (Table 3, Supplementary Table 4).
Characteristics are indicated separately for stone formers with Mendelian disease and with/without LP/P predisposing gene variants. Categorical variables are described by number of participants N (%), continuous variables are described by their mean (SD) or median (25th-75th percentile). eGFR, estimated glomerular filtration rate; BSA, body surface area; PTH, parathyroid hormone; SD, standard deviation.
Genotype-phenotype correlation
Mean age of first kidney stone event (23.5±12.3 years vs. 35.9±13.8 years, p<0.001) and mean age of presentation for metabolic-work up (39±15.8 years vs. 47.2±14.3 years, p=0.009) were lower in KSF with Mendelian diagnosis compared to KSF without, with intermediate results for patients with LP/P variants predisposing for nephrolithiasis (Table 3, Supplemental Fig. 3). Furthermore, cystine stones were significantly more common (64.7 % vs. 1.4 %) and calcium oxalate monohydrates stones less common (23.5 % vs. 52.5 %) in KSF with Mendelian diagnosis compared to KSF with LP/P variants or without genetic diagnosis, in line with higher urinary cystine and urine pH in KSF with Mendelian diagnosis (Table 3, Supplemental Fig. 2). The prevalence of a positive family history of kidney stone disease was similar in all three groups (43.1 % vs. 49.2% vs. 46.0 %). Additionally, KSF with Mendelian disease had more past kidney stone events [5 (2,5) vs. 3 (2,4), p=0.052] than KSF with LP/P variants. In contrast, KSF with LP/P variants had significantly lower plasma phosphate (Table 3).
Mendelian disease
The most common Mendelian diagnosis was cystinuria (n=13) due to biallelic diagnostic variants in SLC3A1 (type A cystinuria, n=9) or SLC7A9 (type B cystinuria, n=4). Urine cystine excretion, where available, was increased in all KSF with genetic diagnosis of cystinuria. One patient had biallelic diagnostic variants in SLC3A1 and a monoallelic LP/P variant in SLC7A9. KSF with biallelic diagnostic SLC3A1 or SLC7A9 variants all had cystine stones.
Five patients had biallelic diagnostic variants in CYP24A1, which encodes the Vitamin-D inactivating enzyme Vitamin D-24 hydroxylase. Interestingly, all KSF with biallelic variants in CYP24A1 (but no KSF with a monoallelic CYP24A1 variant) displayed cystic kidney disease, as reported previously 31,32. Common misdiagnoses for patients with biallelic variants in CYP24A1 (n=3) were polycystic kidney disease or medullary sponge kidney. Two patients with biallelic CYP24A1 variants underwent parathyroidectomy for suspected primary hyperparathyroidism (PHPT), without phenotype alleviation postoperatively.
We furthermore identified three patients with primary hyperoxaluria (PH) in our cohort. The diagnosis was already established in the two patients with primary hyperoxaluria type 1 (PH1) and primary hyperoxaluria type 2 (PH2), respectively. The diagnosis was previously unknown in the patient with primary hyperoxaluria type 3 (PH3) due to a homozygous pathogenic HOGA1 variant. This patient had his first kidney stone episode in his 30s. Metabolic work-up in his late 60s showed CKD stage 2 and a urine oxalate in the upper normal range on free-choice diet, but a strong increase after an instructed one-week diet low in calcium and sodium.
No Mendelian form of nephrolithiasis was detected in the 114 NKSF included in this study (Fig. 1, Fig. 2, Supplementary Table 5).
LP/P variants predisposing to nephrolithiasis
The most common identified genetic predisposition to nephrolithiasis was renal phosphate wasting due to monoallelic variants in the genes encoding the renal sodium/phosphate co- transporters NaPi-2a and NaPi-2c, SLC34A1 (n=15), SLC34A3 (n=18) or in the gene SLC9A3R1, encoding the regulatory interaction protein for NaPi-2a/2c, NHERF, (n=7). LP/P variants in SLC34A1, SLC34A3, but not in SLC9A3R1, were enriched in KSF compared with controls (NKSF and gnomAD, Supplementary Table 5). KSF with LP/P variants in SLC34A1/A3 or SLC9A3R1 had similar plasma phosphate and TmP/GFR compared to KSF without genetic diagnosis, but urine calcium in individuals with LP/P variants in SLC34A3 was higher compared to KSF without genetic diagnosis (Supplemental Figure 4).
We also identified eight patients carrying the same previously described monoallelic LP/P variant (c.458A>G, p.Asn153Ser) in CLDN16, encoding the tight junction protein claudin-16. Biallelic pathogenic variants in CLDN16 cause familial hypomagnesemia with hypercalciuria and nephrocalcinosis (FHHNC)33, and an increased prevalence of nephrolithiasis has been observed in monoallelic carriers of pathogenic CLDN16 variants 30,34. Individuals with the monoallelic CLDN16 variant presented with hypercalciuria and calcium oxalate stones but normal renal function and plasma magnesium. The identified CLDN16 variant was more prevalent in KSF vs. NKSF and vs. all LP/P-variants in CLDN16 in gnomAD-NFE, respectively (Supplementary Table 5).
Additionally, eight patients carried monoallelic LP/P variants in CYP24A1, and the respective variants were five to eight times more prevalent in KSF vs. NKSF and vs. gnomAD-NFE, respectively (Supplementary Table 5). Overall (i.e., considering the prevalence of all variants), CYP24A1 variants were not enriched in our cohort compared to gnomAD-NFE (Supplementary Table 5). KSF with mono- and biallelic CYP24A1 variants both displayed hypercalciuria, but hypercalcemia and/or suppressed parathyroid hormone (PTH) or nephrocalcinosis were only present in KSF with biallelic variants. Urine calcium was higher in KSF with biallelic variants, but KSF with monoallelic variants had a higher number of past stone events (Supplemental Fig. 5).
We furthermore identified six patients with monoallelic LP/P variants in SLC7A9. Risk variants were either previously described to cause autosomal-dominant cystinuria (c.544G>A p.(Ala182Thr)) 35,36 37,38, or enriched in KSF vs NKSF and gnomAD (c.313G>A p.(Gly105Arg)). In contrast to biallelic variants, the six KSF with monoallelic SLC7A9 variants presented with calcium-containing kidney stones without cystine content, similar to previously reported cases 39-41. KSF with monoallelic SLC7A9 variants had elevated urine cystine, albeit lower levels compared to KSF with biallelic variants in SLC3A1/SLC7A9. No significant differences in stone number or age at first stone event were detected between the two groups of patients (Supplemental Fig. 5). In seven individuals with cystine-containing kidney stones and two individuals with increased urine cystine, no diagnostic variants in SLC3A1 or SLC7A9 could be identified (Supplementary Table 6).
Four KSF had LP/P variants in the gene encoding the calcium-sensing receptor, CASR, which is associated with familial hypocalciuric hypercalcaemia type 1 (FHH1), or autosomal- dominant hypocalcemia. The frequency of CASR LP/P variants in KSF was significantly higher than LP/P variants in all gnomAD-NFE participants (Supplementary Table 5). The phenotype of these patients was variable: two patients presented with hypercalciuria but normal plasma calcium, phosphate and PTH. One patient had hypercalcemia with a low-normal PTH and low urine calcium. Another patient presented with elevated PTH, hypercalcemia, pronounced hypophosphatemia, hypercalciuria and calcium oxalate dihydrate stones, compatible with the diagnosis of PHPT. Parathyroidectomy led to a complete normalization of the phenotype.
In the 114 NKSF included in this study, exome sequencing revealed a LP/P variant in two individuals, corresponding to a prevalence of variants predisposing to nephrolithiasis of 1.8 % (compared with 8.13% for KSF) (Fig. 1, Fig. 2, Supplementary Table 5).
DISCUSSION
In this exome-sequencing study in a large, unselected European cohort of 787 adult KSF, we detected a Mendelian kidney stone disease in 23 patients (2.9 %) using stringent diagnostic criteria. This diagnostic yield is significantly lower compared to previous studies conducted in selected groups of KSF 10,17,42. Yet the fraction of additional individuals (8.13%) with LP/P variants predisposing to nephrolithiasis is substantial, especially when considering the broad inclusion criteria, the high prevalence of the disease and the potential rate of false-negatives due to stringent molecular genetic diagnosis criteria and technical limitations of exome sequencing, such as inability to detect deep intronic or difficulty in reliably calling copy-number variants. In fact, if Mendelian disease and variants predisposing to nephrolithiasis are combined, the overall diagnostic yield is 11 % (87 of 787 individuals), comparable to neurometabolic disorders or cancer, where exome sequencing is routinely used 16,43,44. Interestingly, the diagnostic yield was not different between first time and recurrent stone formers (Supplementary Table 3) and did not differ when stratified by sex (Table 3, Supplementary Table 4). In NKSF, no Mendelian kidney stone disease was detected, and LP/P variants in nephrolithiasis genes were significantly less prevalent compared to KSF (1.8 % vs. 8.1 %).
Cystinuria due to biallelic diagnostic variants in SLC3A1 or SLC7A9 was the most common Mendelian disease in our cohort (57 %; n=13 of 23 individuals with Mendelian disease), and an additional six individuals carried monoallelic LP/P variants in SLC7A9. We confirmed a high prevalence of CKD in KSF with cystinuria 45. Yet, this applied only to KSF with biallelic diagnostic variants. We were unable to provide a genetic diagnosis in nine KSF with a cystinuria phenotype, including seven individuals with cystine-containing kidney stones and two individuals with elevated urinary cystine excretion (Supplementary Table 5). While this could be explained at least in part by technical issues, likely additional, yet to be discovered genes, are associated with the clinical phenotype of cystinuria.
Of the LP/P variants detected predisposing to nephrolithiasis, by far the most common (63 %; n=40 of 64) were variants in SLC34A1/3 or SLC9A3R1, highlighting the pathophysiological importance of renal phosphate loss in increasing risk for kidney stone formation 12,46. However, the phenotype of KSF with LP/P variants in these genes on a free choice diet was very subtle and indistinguishable from KSF without predisposing variants. Future studies need to determine if the phenotype of KSF with LP/P variants in these genes can be unmasked by dietary interventions (e.g., by a low phosphate diet) for diagnostic purposes and mitigated for prevention of future stone events (e.g., by phosphate supplementation).
We only detected one KSF with a monoallelic LP/P variant in ADCY10, indicating that variants in ADCY10 are not contributing significantly to the overall nephrolithiasis risk, at least in individuals of European descent 39,47. Further, we identified four patients with CASR LP/P variants in our cohort with very variable phenotypes. Atypical characteristics such as chondrocalcinosis or kidney stones have been described in patients with FHH1 but complicate the separation from primary hyperparathyroidism (PHPT) 48-50. Yet, separation is of critical importance: parathyroidectomy is considered as a first-line treatment in PHPT but surgery is ineffective in FHH1. Additional complexity arises from the fact that PHPT and FHH1 can co- exist, as observed in patient# 6023 51,52.
Genetic diagnoses in KSF can have important prognostic implications and therapeutic consequences. However, while the pathophysiology is well delineated, the effectiveness of therapeutic interventions has not been studied systematically in most forms of Mendelian nephrolithiasis. The widespread use of genetic testing will facilitate the inclusion of KSF with Mendelian disease and LP/P variants predisposing to nephrolithiasis in prospective registries and clinical trials to study disease evolution and evaluate tailored therapeutic interventions. Currently, KSF are only subjected to genetic testing if a specific disease is strongly suspected.
While clinical signs suggesting an inherited disorder have been proposed 53, Mendelian disease was missed or misclassified in five of 23 (22%) KSF in our cohort, despite individual review of each case by a kidney stone expert. The situation is even far more difficult in KSF with LP/P variants predisposing to nephrolithiasis: the associated phenotypes were discrete and variable with a large phenotype overlap with KSF without LP/P variants. Yet, the low prevalence of LP/P variants detected in NKSF suggests a rather high penetrance of LP/P variants predisposing to nephrolithiasis in adults, with the limitation of a small NKSF sample size. Thus, together, these results preclude the definition of clear phenotypic criteria to prioritize genetic testing in adult KSF. In addition, LP/P variants present with their own challenges in clinics and considerations for clinical reporting of LP/P variants and risk alleles have recently been suggested 14.
Strengths of our study include the lack of patient pre-selection with broad inclusion criteria, the large sample size, availability of ethnically matched non-stone forming controls and the detailed phenotype. Our study also has limitations, such as the inability to detect deep intronic and copy- number variants, lack of family recruitment for segregation analyses, overrepresentation of men and a limited ethnic diversity.
In conclusion, our study shows that Mendelian disease is far less common in adult KSF than previously suggested, but simultaneously reveals a high prevalence of LP/P variants predisposing to nephrolithiasis in adult KSF. Our study highlights the potential of genetic testing in unselected adult KSF for accurate assessment of Mendelian disease and identification of variants predisposing to nephrolithiasis to direct patients to tailored therapies and clinical trials.
DISCLOSURES
The authors declare no competing interests.
FUNDING
MAA was supported by a grant for protected research time by the InselGruppe. EGO was supported by Postdoc MobilityStipendien of the Swiss National Science Foundation (grants # P2ZHP3_195181 and P500PB_206851), and Kidney Research UK ((grants # Paed_RP_001_20180925). DGF was supported by the Swiss National Science Foundation (grants # 31003A_172974 and 33IC30_166785/1) and the Swiss National Centre of Competence in Research Kidney.CH. JAS is supported by the Northern Counties Kidney Research Fund (22/01), Kidney Research UK (Paed_RP_001_20180925), LifeArc, and the Medical Research Council (MR/Y007808/1).
AUTHOR CONTRIBUTIONS
DGF, AS and JAS conceptualized the study. DGF and MAA acquired financial support. MAA, EGO and DGF designed data analysis plans. MAA, EGO and MB performed data analysis. MAA, EGO, MB, AS and DGF performed data analysis and data interpretation. MAA wrote the first draft of the manuscript. All authors contributed to discussion and editing of text and approved the final version of the manuscript.
DATA PRIVACY PROTECTION STATEMENT
Any participant identifiers (PIDs) included were solely used for the BKSR and not known to anyone (e.g., hospital staff, patients or participants themselves) outside the research group. Therefore, they cannot be used to identify individuals.
DATA SHARING STATEMENT
The data supporting the findings of this study are available from the corresponding author upon request. Stratified genetic data will be accessible upon publication at European Genome- Phenome Archive (EGA) (https://ega-archive.org/).
SUPPLEMENTARY MATERIAL
Figure S1: Stratification pathway used for whole exome sequencing (WES) data in the Bern Kidney Stone Registry (BKSR)
Figure S2: 24h urinary risk factors stratified by WES results.
Figure S3. Age of first visit and age of first stone event stratified by type of genetic diagnosis (Mendelian vs LP/P predisposing variants).
Figure S4. Genotype–phenotype correlations for LP/P phosphate transporter variants.
Figure S5. Genotype–phenotype correlations (Cystinuria and CYP24A1).
Table S1. Genotype and phenotype of patients with LP/P variants predisposing to nephrolithiasis.
Table S2: List of likely pathogenic (LP) or pathogenic (P) variants in BKSR
Table S3. Characteristics of stone formers stratified by stone events number
Table S4. Characteristics of stone formers with diagnostic variants (Mendelian and LP/P predisposing variants) stratified by sex
Table S5: Case/control prevalence (enrichment) calculation for LP/P variants
Table S6. Patients with cystine stones, clinical cystinuria or elevated urinary cystine but without diagnostic variants in cystinuria genes (SLC3A1/SLC7A9)
ACKNOWLEDGEMENTS
We wish to thank all the patients and staff from all units participating in the study. We also wish to thank Hans Anderegg for technical expertise in database development.
Footnotes
Genetic disease has been separated into clearly Mendelian vs genetic predisposition to nephrolithiasis. This results in revised Fig.1,2, Table 2,3, revised SI and supplementary tables.