CLINICAL RESEARCH www.jasn.org

NPHP1 (Nephrocystin-1) Deletions Cause Adult- Onset ESRD

Rozemarijn Snoek,1,2 Jessica van Setten,3 Brendan J. Keating,4,5 Ajay K. Israni,6 Pamala A. Jacobson,7 William S. Oetting,7 Arthur J. Matas,8 Roslyn B. Mannon,9 Zhongyang Zhang ,10,11 Weijia Zhang,12 Ke Hao,10,11 Barbara Murphy,12 Roman Reindl-Schwaighofer,13 Andreas Heinzl ,13 Rainer Oberbauer,13 Ondrej Viklicky,14 Peter J. Conlon ,15,16 Caragh P. Stapleton,15 Stephan J.L. Bakker,17 Harold Snieder,18 Edith D.J. Peters,1,2 Bert van der Zwaag,1 Nine V.A.M. Knoers,1,2,19 Martin H. de Borst ,17 and Albertien M. van Eerde1,2

Due to the number of contributing authors, the affiliations are listed at the end of this article.

ABSTRACT Background Nephronophthisis (NPH) is the most prevalent genetic cause for ESRD in children. However, little is known about the prevalence of NPH in adult-onset ESRD. Homozygous full gene deletions of the NPHP1 gene encoding nephrocystin-1 are a prominent cause of NPH. We determined the prevalence of NPH in adults by assessing homozygous NPHP1 full gene deletions in adult-onset ESRD. Methods Adult renal transplant recipients from five cohorts of the International Genetics and Translational Research in Transplantation Network (iGeneTRAiN) underwent single-nucleotide polymorphism genotyping. After quality control, we determined autosomal copy number variants (such as deletions) on the basis of median log2 ratios and B-allele frequency patterns. The findings were independently validated in one cohort. Patients were included in the analysis if they had adult-onset ESRD, defined as start of RRT at $18 years old. Results We included 5606 patients with adult-onset ESRD; 26 (0.5%) showed homozygous NPHP1 deletions. No donor controls showed homozygosity for this deletion. Median age at ESRD onset was 30 (range, 18–61) years old for patients with NPH, with 54% of patients age $30 years old. Notably, only three (12%) patients were phenotypically classified as having NPH, whereas most patients were definedashavingCKDwithunknownetiology(n=11; 42%).

Conclusions Considering that other mutation types in NPHP1 or mutations in other NPH-causing were not analyzed, NPH is a relatively frequent monogenic cause of adult-onset ESRD. Because 88% of patients had not been clinically diagnosed with NPH, wider application of genetic testing in adult-onset ESRD may be warranted.

J Am Soc Nephrol 29: 1772–1779, 2018. doi: https://doi.org/10.1681/ASN.2017111200

Nephronophthisis (NPH) is a Mendelian genetic dis- functionally aberrant cilia, making NPH a ciliop- ease and a classic pediatric kidney disease. Although athy.3,4 To date, causative mutations in many genes NPH is considered a rare disorder (incidence of encoding these have been reported, but the 0.1–0.2 per 10,000 live births), NPH is the most prev- alent genetic cause for ESRD in children, with a fre- quency of 15%.1,2 The most common variant is NPH Received November 20, 2017. Accepted March 19, 2018. type 1 (OMIM 256100), in which patients generally Published online ahead of print. Publication date available at present at age 13 years old with ESRD.3 www.jasn.org. The etiology of NPH lies in the primary , Correspondence: Dr. Albertien M. van Eerde, University Medical which functions as a sensory organelle in the renal Center Utrecht, Utrecht University, Lundlaan 6, 3584 EA Utrecht, cell.3,4 Mutations in genes coding for proteins es- The Netherlands. Email: [email protected] sential to the primary cilium lead to structurally or Copyright © 2018 by the American Society of Nephrology

1772 ISSN : 1046-6673/2906-1772 JAmSocNephrol29: 1772–1779, 2018 www.jasn.org CLINICAL RESEARCH largest proportion of NPH (20%–25%) is caused by homozy- Significance Statement gous full gene deletions of the NPHP1 gene (OMIM 607100).5,6 These full gene deletions are recurrent in the gen- Nephronophthisis (NPH) is a classic pediatric Mendelian kidney eral population, a result of recurrent complex rearrangements disease, generally leading to ESRD around age 13 years old. With a – at this locus due to flanking low copy repeats (elements with frequency of 20% 25%, homozygous full gene deletions of the NPHP1 gene (encoding nephrocystin-1) are a prominent cause of 7–9 highly similar sequence identities). Mutations in NPHP1 NPH. However, little is known about the prevalence in the adult- are completely penetrant; thus, the presence of the mutation onset ESRD population. This manuscript describes an analysis of always leads to an NPH phenotype.5,6 For the other NPH genomic data from 5606 patients with adult-onset ESRD (un- genes, no such recurrent full gene deletion is known.5,6 selected for etiology), making it the first to show that these dele- Clinically, NPH generally starts around age 6 years old with tions have a prevalence of 0.5% in adult-onset ESRD. The age at fi onset is also older than described before, ranging up 61 years old. nonspeci c and mild symptoms due to an impaired ability to Only 12% of the patients with homozygous NPHP1 full gene de- concentrate urine and retain water.10 This leads to early symp- letion were clinically diagnosed as having NPH. Therefore, this toms, such as polyuria, polydipsia, and secondary enuresis.10 If paper concludes that NPH is a relatively frequent monogenic cause performed, renal ultrasound may show echogenicity with loss of of adult-onset ESRD. This has many clinical implications: for in- corticomedullary differentiation.11 The disease always progresses stance, in patients with living related family transplantation. Con- sidering that this is only one type of mutation in only one gene to ESRD, in general around age 13 years old, with a need for RRT known to cause monogenic kidney disease, these results warrant (dialysis or renal transplantation).3 In some patients with NPH, wider application of genetic testing in adult-onset ESRD. renal ultrasound shows corticomedullary cysts, although the ma- jority display small atrophic kidneys at the ESRD stage.11 Some patients with NPH (15%) also display additional extrarenal ab- positions across the whole genome (on the basis of human refer- 20 normalities, such as neurologic anomalies () or ence genome build GRCh37). Of these, approximately 22,000 ophthalmologic dysplasias (Senior–Løken syndrome), which can variants were used to cover approximately 2200 manually curated 20 guide diagnosis.12 However, isolated NPH may prove difficult to CNV regions. The GoCAR cohort was genotyped with the fi diagnose clinically, because the phenotype is generally nonspe- In nium HumanOmniExpressExome-8 v1B and OmniExpressEx- cific, can be variable, and often only becomes clinically apparent ome-8 v1.1A (Illumina, San Diego, CA), providing whole-genome 19 in the ESRD stage.3 Therefore, genetic testing is the sole method to coverage with markers for approximately 960,000 SNPs. diagnose NPH with certainty.6 Quality control was performed by excluding all low-quality . – Because it almost always presents in midchildhood, NPH is SNPs, leaving only high-quality SNPs (call rate 0.99; Hardy . , primarily considered a pediatric kidney disease. Nevertheless, Weinberg equilibrium P 0.001; minor allele frequency 0.1) anecdotal cases of patients with adult-onset ESRD have been that were linkage disequilibrium pruned to leave no pairs with 2. reported.13–15 Although these patients suggest that the diag- r 0.2. Also, we removed SNPs in regions with known long nosis of NPH is rare in adults, little is known about the overall prevalence of NPH in the adult-onset ESRD population.13–18 We set out to investigate the prevalence of NPH in adults by analyzing, as a proxy, the prevalence of homozygous NPHP1 full gene deletions in patients with adult-onset ESRD. We pro- posed to do this by assessing autosomal copy number variants (CNVs; e.g., large deletions) in genomic data generated for a genome-wide association study in multiple cohorts of renal transplant recipients and (corresponding donor) controls from the International Genetics and Translational Research in Transplantation Network (iGeneTRAiN) Consortium.19

METHODS

CNV Analyses We genotyped five iGeneTRAiN Consortium cohorts, all con- sisting of renal transplant recipients and (donor) controls.19 Figure 1. Homozygous NPHP1 full gene deletions have an overall prevalence of 0.5% in adult onset ESRD patients, while The DeKAF Genomics, Gen03, TransplantLines-Genetics, and the prevalence is even 0.9% in patients aged 18–50 at ESRD Vienna cohorts were genotyped with the custom Affymetrix onset. The number of renal transplant recipients (n) in the five 20 Axiom Tx GWAS Array (Affymetrix Inc., Santa Clara, CA). combined cohorts is displayed by age decade of start of RRT. This array, designed for the iGeneTRAiN Consortium, contains The number of homozygous NPHP1 full gene deletion carriers single-nucleotide variants, single-nucleotide polymorphisms (n) and the percentage per specific age category are shown (SNPs), and monomorphic markers for approximately 782,000 above each bar.

J Am Soc Nephrol 29: 1772–1779, 2018 NPHP1 Deletions in 0.5% of Adult ESRD 1773 CLINICAL RESEARCH www.jasn.org

(GoCAR and Vienna).21,22 All calls with logR ratio SD .0.3, B-allele frequency drift .0.01, waviness factor .0.05, or waviness SD .0.15 were excluded. Loci where no definite copy number call could be made were also excluded from further analysis. To assess NPHP1 full gene deletions, all samples containing a CNV with an overlap of at least one nucleotide with the NPHP1 region ( 2: 110879888– 110962643 on the basis of GRCh37) were identified using R Studio (version 1.0.153 for Windows; RStudio Inc., Boston, MA). All patients with a copy number call of zero in this entire region (i.e.,homozygousfull gene deletions) were selected for further analysis. We additionally assessed the num- ber of recipients and donors with a copy number call of one (i.e.,heterozygousfull gene deletion carriers). The regions of the other known NPH genes were also assessed for CNVs.6 All NPHP1 deletions called by the algo- rithms were manually inspected to ensure that only true calls were made. Furthermore, the dense SNP arrays applied to the cohorts are well established and validated, and the NPHP1 region is large, leading to a near Figure 2. The graphical representation of (A) log2 ratio and (B) B-allele frequency of zero chance of false findings.19–22 To further all single-nucleotide polymorphism markers in the NPHP1 gene region clearly dis- validate this, the TransplantLines-Genetics tinguishes between patients with a NPHP1 homozygous full gene deletion and healthy controls. Displayed here are an example homozygous NPHP1 full gene deletion pa- samples that displayed a homozygous dele- tient (patient 19, Table 1), shown with black squares, and a healthy control sample, tion copy number call for NPHP1 (n=11) shown with gray circles. (A) The log2 ratio represents the normalized signal intensity were independently validated with multiplex for a specific marker; thus, if there is signal for that marker, the ratio is zero, whereas ligation–dependent probe amplification, any value below zero indicates genomic deletion. An average log2 ratio of 22spe- showing 100% concordance.5,9,15 cifically indicates a homozygous deletion. (B) The B-allele frequency displays if the particular marker is present in a homozygous state (either B-allele frequency of ap- Demographic and Phenotypic proximately 0.00 or 1.00) or a heterozygous state (B-allele frequency of approximately Statistics 0.50), such as is clear in the control patient shown here. In patients with a homozygous Demographic and phenotypic information deletion, the algorithm cannot accurately determine the B-allele frequency, because was retrieved per cohort for all patients no markers are present at all. This leads to a ”waterfall“ configuration, with markers with adult-onset ESRD, which was defined being assigned B-allele frequencies anywhere between 0.00 and 1.00 at random.30 asstartoffirstRRTat18yearsoldorolder.For patients with adult-onset NPH, we studied various phenotypical characteristics, namely stretches of linkage disequilibrium and nonautosomal SNPs. the age at ESRD onset (defined as age at first RRT [dialysis or Only data on the 5606 patients with adult-onset ESRD (start of transplantation]), ethnicity, and primary renal disease diagnosis. first RRTat 18 years old or older) were analyzed: n=3192 from If patient consent (discussed in Ethical Constraints below) al- the DeKAF Genomics and Gen03 combined cohorts, n=1230 lowed for data retrieval from the patient file, data on family from the TransplantLines-Genetics cohort, n=500 from the history, hypertension, polyuria, proteinuria, and extrarenal GoCAR cohort, and n=684 from the Vienna cohort. NPH-associated anomalies were retrieved. The data generated were used to determine autosomal CNVs Ethnicity was determined by principal component analysis (i.e., deletions and duplications). CNV calling was performed us- using1000GenomesPhase3data(1092samplesand14different ing the default settings of two well published algorithms: the ancestries).23,24 Populations were determined by visual inspection BRLMM-P algorithm designed by Affymetrix Inc. (DeKAF Ge- of the first two principal components (by J.v.S.). Descriptive sta- nomics, Gen03, and TransplantLines-Genetics) or PennCNV tistics were generated for the demographic and primary renal

1774 Journal of the American Society of Nephrology J Am Soc Nephrol 29: 1772–1779, 2018 mScNephrol Soc Am J Table 1. Phenotypical information about the 26 patients with NPHP1 homozygous gene deletion, including age at presentation, age at first RRT, and the presence of various renal and extrarenal symptoms of nephronophthisis Ophtamologic Neurologic Ancestry Age at Age at Positive Family Patient Hypertension, Polyuria, Proteinuria, Anomalies Anomalies Sex (Geographic Clinical Diagnosis Presentation, First History of Bone Disease No. Age in yr Age in yr Age in yr Associated Associated Origin) yr RRT, yr Renal Disease 29: with NPH with NPH

1772 1 Woman Caucasian (United Tubular and interstitial Not retrievable 18 Not retrievable Not retrievable Not retrievable Not retrievable Not retrievable Not retrievable Not retrievable States) disease –

79 2018 1779, 2 Woman Caucasian (Europe) NPH 19 19 NR ——+, Age NR ——— 3 Man Caucasian (United Chronic renal failure Not retrievable 21 Not retrievable Not retrievable Not retrievable Not retrievable Not retrievable Not retrievable Not retrievable States) with unknown etiology 4 Woman Caucasian (United Chronic renal failure Not retrievable 21 Not retrievable Not retrievable Not retrievable Not retrievable Not retrievable Not retrievable Not retrievable States) with unknown etiology 5 Man Caucasian (Europe) Glomerular disease 22 22 — 22 NR 22 ——— 6 Woman Caucasian (Europe) Chronic renal failure 21 23 — 21 ————— with unknown etiology 7 Woman Caucasian (United Chronic renal failure 23 23 — 23 NR NR NR —— States) with unknown etiology 8 Woman Caucasian (United Chronic renal failure 23 23 + 23 NR NR + —— States) with unknown etiology 9 Man African (United Hypertensive 25 25 — 25 NR NR NR —— States) nephrosclerosis 10 Woman Caucasian (Europe) Sporadic primary reflux 27 28 — 27 — NR ——— nephropathy 11 Man Caucasian (Europe) GN, histologically 28 28 + ——NR ——Severe examined developmental delay 12 Woman Caucasian (Europe) NPH 29 29 ———NR + —— NPHP1 13 Woman Caucasian (United Chronic renal failure Not retrievable 30 Not retrievable Not retrievable Not retrievable Not retrievable Not retrievable Not retrievable Not retrievable States) with unknown

eein n05 fAutESRD Adult of 0.5% in Deletions etiology 14 Man Caucasian (Europe) NPH 27 30 NR 27 ————— www.jasn.org 15 Man Caucasian (Europe) Vascular nephropathy NR 30 — +, Age NR NR NR NR Congenital Tremor nystagmus 16 Man Caucasian (United Medullary cystic disease NR 31 NR — NR — NR NR NR States) 17 Man Caucasian (Europe) Chronic renal failure 30 34 + 30 ————— with unknown RESEARCH CLINICAL etiology 18 Woman Caucasian (Europe) Chronic renal failure 940+ 9 ————— with unknown etiology 19 Man Caucasian (Europe) Urate nephropathy 28 40 ——————— 1775 CLINICAL RESEARCH www.jasn.org

diagnosis variables using SPSS (version 23 for Windows; IBM Corp.,Armonk,NY).Additionally, we applied a two-sided Fisher — exact test to assess these binomial variables. All samples were also with NPH Anomalies Associated Neurologic analyzed on identity by state patterns to identify patients who were related up to and including to the third degree (identity by state ,10%). le (information noted as fi blindness with NPH Anomalies Associated Ethical Constraints Ophtamologic The enrollment of participants for all iGeneTRAiN Consor- tium cohorts was approved by the institutional review boards e patient of the hospitals at which participants were included. All par- ——— NR NR NR ticipants signed informed consent for genomicdata analysis via

Bone Disease SNP array. For the GoCAR cohort, TransplantLines-Genetics,

, symptom not present; . +, symptom present. and Vienna cohorts, the principal consent form included re- — trieval of relevant information from their patient file. Partic- le; fi — NR ipants enrolled in the DeKAF Genomics and Gen03 cohorts Age in yr

Proteinuria, had to additionally consent for data retrieval from their patient file. Furthermore, participants were not asked to consent to

— ————— return of genomic study results on an individual level. Thus, NR Polyuria, Age in yr we could not notify NPHP1 homozygous or heterozygous gene deletion carriers of this finding. — 34 RESULTS Age in yr Hypertension, +, Age NR NR NR NR Congenital In total, 5606 renal transplant recipients with start of first RRT at any age $18 years old were included. Overall de-

— — scriptive characteristics of the five cohorts were reported 19 History of previously by the iGeneTRAiN Consortium. Of these pa- Renal Disease Positive Family tients with adult ESRD onset, 26 (0.5%) patients (Figure 1) displayed the same approximately 96,389- deletion First Age at RRT, yr (size on the basis of SNP array) (example in Figure 2) on both alleles, including all NPHP1 gene exons. The 26 patients were not related up to and including to the third yr 39 42 42 52 + 50 NR 61 degree. Age at None of the 3311 (donor) controls displayed this homo- Presentation, Not retrievable 43Not Not retrievable retrievable Not retrievable 44 Not retrievableNot Not retrievable Not retrievable retrievable Not Not retrievable retrievable 46 Not Not retrievable retrievable Not Not Not retrievable retrievable retrievable Not Not retrievable retrievable Not Not retrievable retrievable Not Not retrievable retrievable Not retrievable Not retrievable Not retrievable zygous deletion. None of the 5606 recipients showed a de- letion of any of the other 19 known NPH genes.6 Markedly, we detected a higher number of heterozygous NPHP1 dele- tions in transplant recipients (n=36) compared with the transplant (donor) controls (n=10; P,0.001). Although these patients were not additionally assessed with next gen- Clinical Diagnosis with unknown etiology with unknown etiology with unknown etiology disease polycystic kidney disease nephrosclerosis fi Chronic renal failure Chronic renal failure Tubular and interstitial eration sequencing, this nding points to the possibility of compound heterozygosity (a full gene deletion on one allele and a different pathogenic mutation on the other allele) in the recipients. rst RRT (youngest to oldest). For seven patients, not all data were retrievable due to the patients not consenting to additional data retrieval from th fi When addressing the phenotype of the patients with Origin) Ancestry

(Geographic NPHP1 deletions (Figure 1 and Table 1), the median age at States) States) States) start of RRTwas 30 years old (range, 18–61), with 14 patients ). Additional information is discussed in Ethical Constraints in the Methods section. NPH, nephronophthisis; NR, not reported in patient ” (54%) ages 30 years old or older. Interestingly, the preva- Continued

Sex lence of homozygous NPHP1 deletions was 0.9% in recipi- ents between 18 and 50 years old at the start of first RRT (n=24 of 2794) and even higher (2.1%) in recipients ages 18–29 years old No. not retrievable 20 Woman Caucasian21 (Europe)22 Chronic renal failure Woman Caucasian (Europe) Man23 Medullary cystic disease Caucasian (United Man24 NR Caucasian (United 25 Man 42 Caucasian (United Woman Caucasian26 (Europe) Autosomal dominant NR Woman Caucasian (Europe) Hypertensive Table 1. Patient Patients are sorted by age at “ (Figure 1).

1776 Journal of the American Society of Nephrology J Am Soc Nephrol 29: 1772–1779, 2018 www.jasn.org CLINICAL RESEARCH

DISCUSSION

Our data indicate that, with a 0.5% preva- lence of homozygous NPHP1 full gene deletions, the frequency of NPH in adult- onset ESRD is considerably higher than previously reported.6,16 When interpreting our results, one should note that, inherent to the method applied in our study (which only analyzes larger deletions and duplica- tions), the overall prevalence of NPH in adult-onsetESRDthatwereporthereis very likely to be an underestimation (Fig- ure 4). First, with this method, any combi- Figure 3. Only 12% of homozygous NPHP1 full gene deletion cases were correctly nation of two smaller NPHP1 mutations, fi identi ed as having nephronophthisis, while 88% were clinically diagnosed as having such as (homozygous) pathogenic intra- something other than nephronophthisis. genic deletions up to approximately 700 base pairs and single-nucleotide variants, Regarding the clinical primary renal disease diagnosis, only could not be analyzed.22 Because this is outside the scope of three patients (12%), of whom two were younger than 30 years this study, compound heterozygosity and homozygosity for old at the start of firstRRT,wereclassified as having NPH other mutations were not assessed. The assumption that (Figure 3). The other patients (88%) were diagnosed with NPHP1-related disease is probably more frequent in adult- CKD with unknown etiology (n=11), cystic disease (n=2), onset ESRD is supported by our recipients showing hypertensive nephrosclerosis (n=2), tubular and interstitial asignificantly higher frequency of heterozygous deletion car- disease (n=2), glomerular disease (n=1), GN (histologically riers than donors, suggesting that a subset of these dele- examined; n=1), sporadic primary reflux nephropathy (n=1), vas- tion carriers likely also carries a second mutation on the other cular nephropathy (n=1), urate nephropathy (n=1), and autoso- allele. Second, the causative mutations in other NPH genes mal dominant polycystic kidney disease (n=1; no mutation in generally are not recurrent full gene deletions.5,6 This is un- PKD1 or PKD2). derscored by the fact that our analysis showed no large dele- Twopatients displayed extrarenal anomalies associated with tions in these genes.6 We have not performed any (additional) NPH, namely congenital blindness (possibly Senior–Løken next generation sequencing to assess any other types of mu- syndrome type 125) and severe neurodevelopmental delay tation. In the light of these two considerations, our findings (possibly Joubert syndrome type 426). Both of these patients point to an underestimation of the number of causative NPH had not been clinically diagnosed with NPH but with hyper- mutations in the overall cohort. tensive nephrosclerosis and GN (histologically examined), With regard to age at onset of ESRD, over one half of the respectively (Table 1). patients with NPHP1 in our study were 30 years old or older at

Figure 4. At least one in 200 (0.5%) patients has all adult-onset ESRD due to nephronophthisis, but the overall prevalence is likely higher. The late presentation of nephronophthisis might be due to genetic modifier effects. Accurately diagnosing a monogenic disease such as nephronophthisis can have wide-ranging clinical implications.

J Am Soc Nephrol 29: 1772–1779, 2018 NPHP1 Deletions in 0.5% of Adult ESRD 1777 CLINICAL RESEARCH www.jasn.org

first onset of ESRD. Therefore, we postulate that NPH is not algorithm.21 All images in Figure 4 were purchased from thenoun- merely a pediatric disease entity. In the literature so far, only project.com under a royalty-free license. For two images, the color six patients across four families have been described with was changed from black to gray, and one image was cropped to ensure NPHP1 mutations and ESRD onset after 30 years of age, adequate fit. with the oldest being 56 years old.13–15 Our study extends The DeKAF-Genomics cohort and the GEN03 cohort were sup- this age of onset to 61 years old. In the subpopulation with ported by National Institutes of Health/National Institute of Allergy onset between 18 and 50 years old, we observed an NPHP1 and Infectious Diseases grants 5U19-AI070119 and 5U01-AI058013. deletion prevalence of 0.9%. The mechanism leading to phe- R.S., N.V.A.M.K., and A.M.v.E. are supported by Dutch Kidney notypic variance in age of ESRD onset remains unclear. It has Foundation grants KSTP12.010, 15OP14, and Kouncil CP11.18, and been hypothesized previously that this nonpediatric onset of A.M.v.E. is supported by NutsOhra Foundation grant 070-1303. Z.Z. ESRD, especially after 30 years old, is due to the influence of is partially supported by the Translational Collaborative Research yet unknown modifier genes.14,15 Because of the ancestry Initiative Grant from the Icahn School of Medicine at Mount Sinai. composition of our cohorts (the majority being Caucasian; data not shown), we could not analyze the presence of distinct ethnicity-specificmodifier effects.6,8 Nevertheless, modifier DISCLOSURES effects, whether ethnicity specific or not, could play a role in None. the discrepancy in age at onset between the patients in our report and those in older literature. In assessing the clinical diagnoses of the patients with REFERENCES NPHP1 deletions, we show that there is likely an underdiag- nosis of NPHP1-related disease in adult-onset ESRD in clinical 1. Levy M, Feingold J: Estimating prevalence in single-gene kidney dis- – practice. The underdiagnosis, which was universally present eases progressing to renal failure. Kidney Int 58: 925 943, 2000 fi 2. Hildebrandt F: Genetic kidney diseases. Lancet 375: 1287–1295, 2010 among ve cohorts originating from several countries in Eu- 3. Wolf MTF, Hildebrandt F: Nephronophthisis. Pediatr Nephrol 26: 181– rope as well as the United States, might, at least in part, be due 194, 2011 to the nonspecific phenotype, because it can be especially dif- 4. Slaats GG, Lilien MR, Giles RH: Nephronophthisis: Should we target ficult to recognize in adults presenting with advanced renal cysts or fibrosis? Pediatr Nephrol 31: 545–554, 2016 failure.14 This is underscored by our finding that 88% of pa- 5. Halbritter J, Porath JD, Diaz KA, Braun DA, Kohl S, Chaki M, et al.; GPN Study Group: Identification of 99 novel mutations in a worldwide cohort tients with NPHP1 had received a clinical diagnosis other than of 1,056 patients with a nephronophthisis-related . Hum NPH, even in patients with NPH-associated extrarenal anom- Genet 132: 865–884, 2013 alies. Notwithstanding, it is important to accurately define the 6. Braun DA, Hildebrandt F: . Cold Spring Harb Perspect Biol etiology of patients with ESRD not only to prevent misdiag- 9: pii: a028191, 2017 nosis and useless treatment but also, because a monogenic 7. Saunier S, Calado J, Benessy F, Silbermann F, Heilig R, Weissenbach J, 27 et al.: Characterization of the NPHP1 locus: Mutational mechanism in- disease can have a variety of clinical implications. For in- volved in deletions in familial juvenile nephronophthisis. Am J Hum stance, it can affect decisions related to living related kidney Genet 66: 778–789, 2000 donation, which is contraindicated in siblings at risk.28 It 8. Yuan B, Liu P, Gupta A, Beck CR, Tejomurtula A, Campbell IM, et al.: might also influence choices regarding family planning, espe- Comparative genomic analyses of the human NPHP1 locus reveal cially in consanguineous relationships.29 complex genomic architecture and its regional evolution in primates. fi PLoS Genet 11: e1005686, 2015 Insummary,wearethe rsttoshowinalargecohortthatNPH 9. Konrad M, Saunier S, Heidet L, Silbermann F, Benessy F, Calado J, et al.: due to NPHP1 homozygous full gene deletions has a prevalence Large homozygous deletions of the 2q13 region are a major cause of of one in 200 patients (0.5%) in all adult-onset ESRD. Although juvenile nephronophthisis. Hum Mol Genet 5: 367–371, 1996 the incidence was clearly higher in patients with an ESRD onset 10. Ala-Mello S, Koskimies O, Rapola J, Kääriäinen H: Nephronophthisis in fi between 18 and 50 years old (prevalence of 0.9%), NPH can have Finland: Epidemiology and comparison of genetically classi ed sub- groups. Eur J Hum Genet 7: 205–211, 1999 an onset at up to 61 years of age. Because the method that we 11. Blowey DL, Querfeld U, Geary D, Warady BA, Alon U: Ultrasound used underestimates the total number of causal mutations, we findings in juvenile nephronophthisis. Pediatr Nephrol 10: 22–24, 1996 conclude that NPH is a relatively frequent monogenic cause of 12. Wolf MTF: Nephronophthisis and related syndromes. Curr Opin Pe- adult-onset ESRD that is likely underdiagnosed in current daily diatr 27: 201–211, 2015 practice. Given the potential clinical implications of having a 13. Georges B, Cosyns JP, Dahan K, Snyers B, Carlier B, Loute G, et al.: Late-onset renal failure in Senior-Loken syndrome. Am J Kidney Dis 36: genetic diagnosis, our data warrant wider application of genetic 1271–1275, 2000 testing in adult-onset ESRD. 14. Hoefele J, Nayir A, Chaki M, Imm A, Allen SJ, Otto EA, et al.: Pseu- dodominant inheritance of nephronophthisis caused by a homozygous NPHP1 deletion. Pediatr Nephrol 26: 967–971, 2011 ACKNOWLEDGMENTS 15. Haghighi A, Savaj S, Haghighi-Kakhki H, Benoit V, Grisart B, Dahan K: Identification of an NPHP1 deletion causing adult form of neph- ronophthisis. Ir J Med Sci 185: 589–595, 2016 The authors thank Affymetrix Inc. (Santa Clara, CA) for performing 16. Bollée G, Fakhouri F, Karras A, Noël L-H, Salomon R, Servais A, et al.: part of the copy number variant calling using their BRLMM-P Nephronophthisis related to homozygous NPHP1 gene deletion as a

1778 Journal of the American Society of Nephrology J Am Soc Nephrol 29: 1772–1779, 2018 www.jasn.org CLINICAL RESEARCH

cause of chronic renal failure in adults. Nephrol Dial Transplant 21: 24. 1000 Genomes Project Consortium; Auton A, Brooks LD, Durbin RM, 2660–2663, 2006 Garrison EP, Kang HM, Korbel JO, et al.: A global reference for human 17. Chaari I, Trabelsi M, Goucha R, Elaribi Y, Kharrat M, Guarguah T, et al.: genetic variation. Nature 526: 68–74, 2015 Prevalence and incidence estimation of large NPHP1 homozygous 25. Loken AC, Hanssen O, Halvorsen S, Jolster NJ: Hereditary renal dys- deletion in Tunisian population. Pathol Biol (Paris) 60: e84–e86, 2012 plasia and blindness. Acta Paediatr 50: 177–184, 1961 18. Omran H, Fernandez C, Jung M, Häffner K, Fargier B, Villaquiran A, 26. Parisi MA, Bennett CL, Eckert ML, Dobyns WB, Gleeson JG, Shaw et al.: Identification of a new gene locus for adolescent nephronoph- DWW, et al.: The NPHP1 gene deletion associated with juvenile thisis,onchromosome3q22inalargeVenezuelanpedigree.Am J Hum nephronophthisis is present in a subset of individuals with Joubert Genet 66: 118–127, 2000 syndrome. Am J Hum Genet 75: 82–91, 2004 19. International Genetics & Translational Research in Transplantation 27. Prakash S, Gharavi AG: Diagnosing kidney disease in the genetic era. Network (iGeneTRAiN): Design and implementation of the in- Curr Opin Nephrol Hypertens 24: 380–387, 2015 ternational genetics and translational research in transplantation net- 28. Huang E, Samaniego-Picota M, McCune T, Melancon JK, Montgomery work. Transplantation 99: 2401–2412, 2015 RA, Ugarte R, et al.: DNA testing for live kidney donors at risk for au- 20. Li YR, van Setten J, Verma SS, Lu Y, Holmes MV, Gao H, et al.: Concept tosomal dominant polycystic kidney disease. Transplantation 87: 133– and design of a genome-wide association genotyping array tailored for 137, 2009 transplantation-specific studies. Genome Med 7: 90, 2015 29. van Eerde AM, Krediet CTP, Rookmaaker MB, van Reekum FE, Knoers 21. The P, just P, finally P: BRLMM-P: A genotype calling method for the NVAM, Lely AT: Pre-pregnancy advice in chronic kidney disease: Do SNP 5.0 Array. In: ReVision, 2007, pp 1–16 not forget genetic counseling. Kidney Int 90: 905–906, 2016 22. Wang K, Li M, Hadley D, Liu R, Glessner J, Grant SFA, et al.: PennCNV: 30. Illumina: Interpreting infinium assay data for whole-genome structural An integrated hidden Markov model designed for high-resolution copy variation. In: Analysis, 2010, pp 1–9 number variation detection in whole-genome SNP genotyping data. Genome Res 17: 1665–1674, 2007 23. Jolliffe IT, Cadima J: Principal component analysis: A review and recent developments. Philos Trans A Math Phys Eng Sci 374: See related editorial, “The Underestimated Burden of Monogenic Diseases in 20150202, 2016 Adult-Onset ESRD,” on pages 1583–1584.

AFFILIATIONS

1Departments of Genetics and 3Cardiology and 2Center for Molecular Medicine, University Medical Center Utrecht, Utrecht University, Utrecht, The Netherlands; 4Department of Surgery, Penn Transplant Institute, Perelman School of Medicine, University of Pennsylvania, Philadelphia, Pennsylvania; 5Division of Genetics, The Children’s Hospital of Philadelphia, Philadelphia, Pennsylvania; 6Department of Medicine, Hennepin County Medical Center, 7College of Pharmacy, and 8Department of Surgery, University of Minnesota, Minneapolis, Minnesota; 9Department of Medicine, University of Alabama at Birmingham, Birmingham, Alabama; 10Department of Genetics and Genomic Sciences, 11Icahn Institute for Genomics and Multiscale Biology, and 12Division of Nephrology, Department of Medicine, Icahn School of Medicine at Mount Sinai, New York, New York; 13Division of Nephrology and Dialysis, Department of Internal Medicine III, Medical University of Vienna, Vienna, Austria; 14Department of Nephrology, Transplant Center, Institute for Clinical and Experimental Medicine, Prague, Czech Republic; 15Department of Molecular and Cellular Therapeutics, Royal College of Surgeons, Dublin, Ireland; 16Department of Nephrology, Beaumont Hospital, Dublin, Ireland; and 17Departments of Nephrology, 18Epidemiology, and 19Genetica, University Medical Center Groningen, University of Groningen, Groningen, The Netherlands

J Am Soc Nephrol 29: 1772–1779, 2018 NPHP1 Deletions in 0.5% of Adult ESRD 1779 SIGNIFICANCE STATEMENT

Nephronophthisis (NPH) is a classic pediatric Mendelian kidney disease, generally leading to ESRD around age 13 years old. With a frequency of 20%–25%, homozygous full gene deletions of the NPHP1 gene (encoding nephrocystin-1) are a prominent cause of NPH. However, little is known about the prevalence in the adult-onset ESRD population. This manuscript describes an analysis of genomic data from 5606 patients with adult- onset ESRD (unselected for etiology), making it the first to show that these deletions have a prevalence of 0.5% in adult-onset ESRD. The age at onset is also older than described before, ranging up 61 years old. Only 12% of the patients with homozy- gous NPHP1 full gene deletion were clinically di- agnosed as having NPH. Therefore, this paper concludes that NPH is a relatively frequent monogenic cause of adult-onset ESRD. This has many clinical implications: for instance, in patients with living related family transplantation. Consid- ering that this is only one type of mutation in only one gene known to cause monogenic kidney dis- ease, these results warrant wider application of genetic testing in adult-onset ESRD.