Clinical and genetic analyses of a Dutch cohort of 40 patients with a nephronophthisis-related ciliopathy

Patients were recruited at all eight university medical centers in The Netherlands, via the Dutch Kidney Patient Association and the Dutch Joubert Syndrome Patient and Parents’ Network, and via the webpage www.​kouncil.​nl. Patients were included in the AGORA (Aetiologic research into Genetic and Occupational/environmental Risk factors for Anomalies in children) data- and biobank from March 2014 until May 2017 at the Department of Genetics and the Department of Pediatric Nephrology of the University Medical Center Utrecht, the Department of Human Genetics of the Radboudumc Nijmegen, and the Department of Nephrology of the VU University Medical Center Amsterdam. Approval by the regional Committee on Research involving human subjects (CMO Arnhem-Nijmegen 2006/048) was obtained. Informed consent was available for all subjects for access to medical records, DNA testing, and/or collection of biomaterial including urine, fibroblasts, and/or deciduous teeth.

Inclusion criteria were a molecularly confirmed diagnosis of NPH-RC or suspected NPH-RC. NPH-RC was defined as isolated NPH, NPH with extrarenal features that do not constitute a recognizable syndrome, and syndromic NPH. Suspected NPH-RC was based on the presence of CKD stages 2–5 in all patients [15], a family history compatible with autosomal recessive inheritance, and additional extrarenal features associated with a ciliopathy and/or clinical characteristics of NPH including signs of a renal concentration defect (e.g., polyuria), renal ultrasound showing parenchymal damage (e.g., increased echogenicity) but absence of congenital anomalies of the kidneys and urinary tract that could explain the phenotype, urinalysis showing absence of hematuria, erythrocyte casts and proteinuria in the nephrotic range, absence of recurrent urinary tract infections, and exclusion of other causes of early-onset renal failure. Cases in which the clinical diagnosis was ambiguous were discussed with a pediatric nephrologist (M. L.) and a clinical geneticist (N. K.) until consensus was reached. We included five patients with Bardet-Biedl syndrome and a nephronophthisis phenotype; however, the classification of Bardet-Biedl syndrome as NPH-RC is not consistent across literature [6, 11]. Siblings of patients with molecularly confirmed NPH-RC who had the same biallelic mutations in an NPH-RC-associated gene and who showed some but not all clinical features were included for the analysis of initial phenotypic features of NPH. In addition to 40 patients with NPH-RC, we included 12 patients with a ciliopathy and renal features that did not fulfill clinical criteria of NPH to assess the phenotypic spectrum of renal ciliopathies.

To record detailed clinical and genetic data from patients with NPH-RC, we developed a multicenter Nephronophthisis Registry using Castor EDC (Castor Electronic Data Capture, Ciwit BV, Amsterdam, The Netherlands). Access to electronic medical records or paper copies of medical records was requested for all patients. For standardized data collection, a set of clinical parameters and definitions was established by a team of clinical experts from relevant fields including pediatric nephrology, clinical genetics, and clinical molecular genetics. Consensus was reached on all parameters. Human Phenotype Ontology (HPO) terms were used for most of the clinical terms recorded to phenotype the patients, thereby also providing a basis for future integration with clinical data from other nephronophthisis registries and for clinical matching of patients in HPO-centered registries of rare diseases [16]. Case report forms were structured in categories including medical history, physical examination, ophthalmological examination, and genetic analysis. For longitudinal data, we developed report forms for laboratory results, medication, physical re-examination, histological examination, and imaging. Syndrome diagnoses were based on clinical data on renal and extrarenal phenotypes (e.g., presence of molar tooth sign on brain imaging in Joubert syndrome) according to internationally recognized diagnostic criteria [17‐20], in combination with the assessment of the referring physician.

Of 17 patients without a molecular diagnosis upon inclusion, 13 patients from 12 families gave consent for genetic testing. Targeted deletion/duplication analysis was performed to assess the occurrence of NPHP1 deletions in patients with isolated NPH, Senior-Løken syndrome, and Joubert syndrome, and to assess copy number variations in ciliary genes in which we identified single heterozygous point mutations (Supplementary Methods S1). Targeted next-generation sequencing using a gene panel containing 15 genes associated with NPH was performed in one patient (10:44). Whole-exome sequencing (Supplementary Methods S2) was subsequently performed in this patient and 10 additional patients with no deletion or a heterozygous deletion in NPHP1, including a patient with Bardet-Biedl-like syndrome in whom biallelic NPHP1 deletions were unlikely to be causal.

Clinical and genetic data from the Nephronophthisis Registry were exported to Excel and analyzed using GraphPad Prism (version 7.02).

Forty patients from 36 families with NPH-RC were recruited, including patients with isolated NPH (n = 18), Joubert syndrome (n = 8), Bardet-Biedl syndrome (n = 5), Senior-Løken syndrome (n = 4), cranioectodermal dysplasia (n = 2), and NPH-RC phenotypes that did not constitute a recognizable syndrome (n = 3) (Fig. 1). Twenty-two out of 40 patients (55%) had extrarenal ciliopathy-associated features including impaired vision (16/40; 40%), oculomotor abnormalities (10/33; 30%), and developmental delay (13/36; 36%), indicating enrichment of syndromic forms of NPH-RC in our cohort (Supplementary Table S1). Table 1 depicts the frequency of occurrence of clinical features per causative gene for 40 patients.

Patients developed ESRD at a median age of 13 years (range 5–47 years). The median age of symptom onset was 9 years in both isolated and syndromic forms of NPH (range 5–26 vs. 5–33 years respectively). Median time between reported symptom onset and ESRD was 1 year (range 0–8 years) in patients with isolated NPH (n = 11) and 5.5 years (range 0–14 years) in patients with syndromic forms of NPH (n = 4). Four out of 18 patients with isolated NPH (22%) did not yet have ESRD at the time of inclusion (median age 21 years; range 17–41 years), although two patients had at least CKD stage 3. In contrast, 10/20 patients with syndromic forms of NPH (50%) did not have ESRD at the time of inclusion (median age 15.5 years; range 10–25 years) and three patients had at least CKD stage 3. More longitudinal data will be collected to determine the rate of CKD progression in patients with isolated and syndromic forms of NPH.

To evaluate the range of renal ciliopathy-associated phenotypes, we included 12 patients with a renal ciliopathy without sufficient evidence for NPH. Phenotypes included perinatally lethal renal cystic hypodysplasia and renal insufficiency with proteinuria (Supplementary Results S1 and Supplementary Table S2).

To assess presenting symptoms of NPH, we reviewed medical records and conducted patient interviews. Presenting symptoms were documented for 24/40 patients. The most frequently reported initial signs and symptoms were fatigue (10/24; 42%), polydipsia and polyuria (8/24; 33%) and hypertension (5/24; 21%) (Fig. 2 and Supplementary Table S3). Anemia (2/24; 8%), growth retardation (2/24; 8%), enuresis (2/24; 8%), and uremic symptoms (1/24; 4%) occurred less frequently. Novel reported presenting symptoms included muscle cramps and injuries (n = 2), dizziness (n = 1), red patches on the skin (n = 1), and weight loss (n = 1). In addition, a patient with Joubert syndrome caused by NPHP1 mutations (34:73) reported muscle cramps after NPH was diagnosed. Based on these findings, persistent fatigue and muscle complaints in children should prompt evaluation of kidney function in addition to the classic presenting symptoms of NPH.

Overall occurring symptoms were analyzed in 40 patients with NPH-RC. The most frequently reported symptoms were polydipsia and polyuria (18/25; 72% and 15/25; 60%, respectively) and enuresis (4/19; 21%) (Table 2 and Supplementary Table S3). Fifteen out of 23 patients (65%) had normal-sized kidneys on renal ultrasound and eight patients (35%) had at least one small kidney (size < 5th percentile) (reference values: [21, 22]). Small- to normal-sized kidneys is a typical finding in NPH [6, 7, 23]. Additional ultrasound findings included increased echogenicity (17/26; 65%), renal cysts (13/30; 43%), and abnormal corticomedullary differentiation (7/22; 32%). Increased echogenicity was the most prevalent finding on renal ultrasound and could therefore be considered a more sensitive finding for nephronophthisis than renal cysts.

Renal biopsy was performed in 8/40 patients. Insufficient material was obtained from two patients, resulting in a second biopsy for one patient. Tubulointerstitial nephritis was a finding in 4/7 biopsies and tubular atrophy in 3/7. Renal cysts were reported in none of the patients. Instead, glomerulosclerosis was seen in 5/7 biopsies. Biopsies were taken in advanced stages of NPH (CKD stage ≥ 3 for all patients except patient 11:45), when differentiating findings such as thickening of tubular basement membranes and cysts on the corticomedullary border were expected to be no longer detectable. These findings suggest that the diagnostic value of a renal biopsy is limited in advanced stages of CKD, which is often already present when the diagnosis of NPH is first suspected.

Twenty-three out of 40 patients (58%) had a (likely) molecular diagnosis at the time of inclusion (Supplementary Table S4). We performed genetic testing for 13 patients from 12 families without a molecular diagnosis, including genotyping, bioinformatics analysis, and/or functional assays for the interpretation of genetic variants (Fig. 3a). A biallelic mutation in NPHP1 was detected in five out of 13 patients (38%), of whom two patients had a homozygous deletion and three patients had a compound heterozygous deletion and mutation (Table 3). Consequently, NPHP1 was the most frequently mutated gene in our cohort (16/36 patients (44%); Fig. 3b), which is consistent with literature [10].

Additional biallelic variants were detected in TMEM67 in one patient with Bardet-Biedl-like syndrome (23:58). Two unrelated patients with Joubert syndrome (1:34 and 5:38) harbored a compound heterozygous known pathogenic variant and the same missense variant of unknown significance (c.[628T>C] p.[Ser210Pro]) in TMEM67, suggesting a possible founder effect. This variant represents a novel candidate variant for NPH-RC. Interestingly, the latter two patients had progressive liver fibrosis, liver function abnormalities, and hepatomegaly, which is consistent with reports of liver involvement in patients with mutations in TMEM67 [6, 14].