Моногенные заболевания почек и моногенный диабет присутствуют у пациентов почечной клиники с негенетическим диагнозом.

Introduction

Genetic testing is useful in the management of people with chronic kidney disease 1 , 2 . A genetic cause is demonstrated in more than 50% of people with persistent glomerular haematuria 3 , up to 30% of those with proteinuria or Focal and Segmental Glomerulosclerosis (FSGS) 4 , as well as 10–30% of those with kidney failure 5 , 6 . In addition, genetic kidney disease is common in people with familial IgA nephropathy 7 .

Genetic kidney disease is suspected associated with particular phenotypes such as a structural anomaly (Congenital Anomalies of the Kidney and Urinary Tract, CAKUT) 8 , some glomerulopathies 9 , cystic kidney diseases 10 , and some tubulopathies 11 ; where there is a younger age at presentation; where there is another affected family member; or where there are syndromic or extrarenal manifestations 12 . Extra-renal features have been reported in most forms of genetic kidney disease and retinal abnormalities are often associated with the causative genes 13 – 16 but are infrequent with the commonest conditions, namely, autosomal dominant (AD) Alport syndrome, and the typical and atypical forms of AD Polycystic kidney disease.

However genetic kidney disease is often unrecognised. A likely genetic cause is overlooked where the phenotype is not obvious or the features are incompletely penetrant or expressed, where there is no other affected family member, and, sometimes, where Whole Exome Sequencing (WES) is used for testing 17 . A genetic cause is also often overlooked where the patient already has another non-genetic diagnosis.

Many patients in a general nephrology clinic have diabetic kidney disease, where some have monogenic diabetes, formerly known as Maturity Onset Diabetes of the Young (MODY). This is a heterogeneous disorder characterised by a young age at onset, with a positive family history but lacking autoantibodies 18 . Monogenic diabetes affects 1- 5% of all those with known diabetes 19 , 20 , but its insidious onset, overlap with other forms of diabetes, and infrequent genetic testing 21 mean that many of those affected are unrecognised or misdiagnosed with Type 1 or 2 Diabetes mellitus 21 . Monogenic diabetes results from pathogenic variants in the 8 main genes that control pancreatic beta cell development and function. Pathogenic variants in these genes affect glucose sensing and insulin secretion 22 . The diagnosis of monogenic diabetes and identification of the affected gene are important because of the different treatments required for glucose control, the association with extra-renal features and the identification of other affected family members.

The aim of this study was to determine how often patients from a general nephrology clinic, who were not suspected of having an underlying genetic kidney disease, were found to have a likely disease-causing variant. The aim was not to determine the proportion of clinic patients with a familial kidney disease. The number of patients with diabetic kidney disease together with a positive family history encouraged us to examine additionally for a monogenic cause of diabetes.

Participants and methods

Participants

Sixty-four patients attending a routine 3-monthly review at a weekly general Nephrology clinic at a metropolitan teaching hospital were recruited over 20 weeks in 2023. This clinic included about 15 people, attending face-to-face, who were reviewed by one of three clinicians for investigation and management of haematuria, proteinuria, renal impairment, diabetic kidney disease, follow-up of acute kidney injury and for the management of kidney failure, dialysis and post-transplant care.

The same renal physician (JS) examined each clinic patient list and reviewed the patients’ charts the day before they were due to attend. The approach was standardised. She noted their diagnoses, and excluded any people with a reported genetic kidney disease, features suggesting genetic kidney disease (from notes, renal biopsy or imaging), or known family history of kidney disease, noted in the last 3 visits, in order to identify potential study participants (Fig.  1 ). Patients were not asked specifically about a family history of genetic kidney disease.  Fig. 1 Figure demonstrating the number of patients attending the clinic and those recruited .

Figure demonstrating the number of patients attending the clinic and those recruited .

Genetic kidney diseases that excluded patients from the study included: any chromosomal abnormalities, tuberous sclerosis, Alport syndrome, any form of CAKUT on imaging (reflux nephropathy, single kidney, bifid collecting system, horseshoe kidney etc.) or bilateral enlarged multicystic kidneys suggesting AD polycystic kidney disease (ADPKD), steroid-resistant FSGS, HNF1B -nephropathy, or C3 glomerulopathy.

Patients were also excluded if they had another affected family member with kidney disease noted in their three most recent medical consultations.

The presence of diabetes, hypertension, renal impairment, and evidence of haematuria, proteinuria, kidney cysts or kidney stones were also noted.

All of those recruited provided written informed consent for their participation.

All methods were carried out in accordance with relevant guidelines and regulations. All experimental protocols were approved by the Austin Health Human Research Ethics Committee (HREC/16/Austin/538). Genomic variants considered pathogenic or likely pathogenic were submitted with clinical details to the LOVD database.

Whole exome sequencing and variant filtering

Study participants provided venous blood samples and DNA was extracted using the QIAamp DNA Blood Maxi Kit (Qiagen, The Netherlands). Fifty ng of purified DNA underwent Whole Exome Sequencing (WES) on an Illumina NovaSeq platform (California) at the Australian Genome Research Facility (AGRF, Melbourne).

Germline single nucleotide variants (SNV) and short insertion/deletions (indels) were called using standard workflows based on the Genome Analysis Toolkit (GATK v4.4.00 (Suppl Figure S1 ) 23 . Variants were annotated using the Ensembl Variant Effect Predictor tool ( www.ensembl.org/info/docs/tools/vep/ 24 .

Variants were filtered for pathogenic variants in 495 genes from the Genomics England panels for genetic kidney disease (Haematuria, Renal proteinuria, Congenital Anomalies of the Kidney and Urinary Tract, Cystic Kidney disease and Renal ciliopathies, Kidney failure with no obvious cause). Variants were also filtered for pathogenic variants in the 8 commonest genes affected in monogenic diabetes ( GCK, HNF1A, HNF1B, HNF4A, NEUROD1, PDX1, ABCC8, KCNJ11 ).

Genetic variants were filtered for rarity (generally in < 6 alleles in gnomAD v2.1 25 ), and for assessments of Pathogenic or Likely Pathogenic according to Alamut (Sophia Genetics, https://www.sophiagenetics.com/ ) using the ACMG/AMP criteria 26 and in ClinVar ( https://www.ncbi.nlm.nih.gov/clinvar/ ).

Copy number variants

Copy number variants were examined for a ratio < 1.5 or > 2.5 using GatK Germline CNV Caller ( https://gatk.broadinstitute.org/hc/en-us/articles/360047217671-GermlineCNVCaller ) (Suppl Figure S2 ), and then examined in ClinVar and Alamut ( https://www.sophiagenetics.com/ ) for pathogenicity.

Variants in genetic kidney disease genes were considered disease-causing if they were Pathogenic or Likely pathogenic in both ClinVar and in Alamut. Variants in the monogenic diabetes genes were considered possibly disease-causing if they had been reported previously associated with monogenic diabetes. Pathogenic variants were required to be consistent with the mode of inheritance, that is, monoallelic for AD disease and biallelic for AR disease.

Statistical methods

The confidence interval for the proportion of patients with a genetic kidney disease was calculated using the Wald method 27 .

Results

Recruitment

Of the 15 patients attending the clinic, typically 5 were assessed as without a genetic kidney disease and potential study participants (Fig.  1 ). Three of the 5 were recruited on the basis of interpreter availability, and the ability and time to walk to the ward research room for consent and venesection. The major reason for not recruiting all 5 was where two patients were booked for clinic review at overlapping times. Overall, about five people refused to participate in the study, usually because they did not want to ‘be late for their appointment’ and, less often, because they ‘did not want to know’.

Clinical features

Sixty-four patients were recruited including 44 males and 20 females with a median age of 66 years (range 26 to 87). Their underlying kidney diseases included diabetes (n = 36, 56%) with probable diabetic kidney disease (DKD, n = 28,44%), glomerulonephritis (IgA, minimal change, antineutrophil cytoplasmic antibody (ANCA)-mediated vasculitis, systemic lupus erythematosus (SLE), FSGS or membranous glomerulonephritis, n = 20, 31%) or other disease (hypertension, renovascular disease, myeloma, post nephrectomy, sarcoid, obstruction, unknown, n = 12, 19%) (Table 1 ). In general, the diagnosis of autoimmune glomerulonephritis was confirmed by renal biopsy but the diagnosis of diabetic kidney disease was often assumed on the basis of a long history of diabetes, proteinuria and diabetic retinopathy. Six (9%) people volunteered a family history of kidney disease or kidney failure at recruitment. Two of these had diabetes. Table 1 Clinical features in a cohort from a general nephrology clinic. Patient characteristics N = 64 Age (median, range, years) 66 (26 to 87) Sex 44 M, 20F Types of kidney disease (including disease plus diabetes) • Probable diabetic kidney disease (n = 28) or diabetes (n = 8) 36 (56%) • Glomerulonephritis (IgA, minimal change, ANCA vasculitis, SLE, FSGS or membranous glomerulonephritis) 20 (31%) • Incidental –hypertension, myeloma, hypertension, postnephrectomy sarcoid, obstruction, etc 12 (19%) • Unknown 4 (6%) Family history of kidney disease 6 (9%) Haematuria >10 RBC/ul 17 (27%) Urinary albumin: creatinine > 3 mg/mmol 49 (77%) Kidney cysts (≥ one on ultrasound or CT-IVP) n = 59 20 (34%) Stones or renal calcification on imaging (n = 60) 8 (14%) eGFR > 60 ml/min/1.73m 2 6 (9%) eGFR 20—60 ml/min/1.73m 2 44 (69%) eGFR < 20 ml/min/1.73m 2 , dialysis or transplant 14 (22%)

Table 1 Clinical features in a cohort from a general nephrology clinic. Patient characteristics N = 64 Age (median, range, years) 66 (26 to 87) Sex 44 M, 20F Types of kidney disease (including disease plus diabetes) • Probable diabetic kidney disease (n = 28) or diabetes (n = 8) 36 (56%) • Glomerulonephritis (IgA, minimal change, ANCA vasculitis, SLE, FSGS or membranous glomerulonephritis) 20 (31%) • Incidental –hypertension, myeloma, hypertension, postnephrectomy sarcoid, obstruction, etc 12 (19%) • Unknown 4 (6%) Family history of kidney disease 6 (9%) Haematuria >10 RBC/ul 17 (27%) Urinary albumin: creatinine > 3 mg/mmol 49 (77%) Kidney cysts (≥ one on ultrasound or CT-IVP) n = 59 20 (34%) Stones or renal calcification on imaging (n = 60) 8 (14%) eGFR > 60 ml/min/1.73m 2 6 (9%) eGFR 20—60 ml/min/1.73m 2 44 (69%) eGFR < 20 ml/min/1.73m 2 , dialysis or transplant 14 (22%)

Clinical features in a cohort from a general nephrology clinic.

Seventeen people (27%) had haematuria (> 10 RBC/ul), 49 (77%) had abnormal urinary albumin: creatinine (UAC > 3 mg/mmol). Six (9%) had an estimated glomerular filtration rate (eGFR) ≥ 60 ml/min/1.73m 2 , 44 (69%) had an eGFR between 20 and 60 ml/min/1.73m 2 , and 14 had an eGFR   < 20 ml/min/1.73 m 2 , or were undergoing renal replacement therapy. Twenty had at least one kidney cyst on imaging (34%) and 8 had kidney calcification or stones (13%). Five had multiple cysts in both kidneys but not the increased kidney size consistent with AD polycystic kidney disease (ADPKD).

QC of sequencing

All DNA samples and sequencing passed the AGRF QC standards. This included the standards for DNA concentration, purity and integrity; mapping rate (> 95%) and duplicates (< 15%), average coverage depth (at least × 30) and 95% target coverage.

Genetic results in monogenic kidney disease genes

Seven single nucleotide variants (7/64, 11%, 95%CI 0.05 to 0.21) assessed as Pathogenic or Likely pathogenic in ClinVar were found in 7 people in this cohort (Table 2 ) . No copy number variant that was Pathogenic or Likely pathogenic was identified. Corresponding diagnoses included diabetic kidney disease 2 (29%), diabetes in 3 (43%), autoimmune glomerulonephritis in 3 (43%) or another disease (vascular disease in 2, 29%). Variants that were excluded as Pathogenic or Likely pathogenic variants are shown in the Supplementary material. Table 2 Clinical features and disease-causing genetic variants in patients from a General Nephrology clinic. Patient Kidney disease Kidney phenotype Family history Gene, transcript, dbSNP Disease, MIM, MoI Variant and protein change , LOVD submission number MAF in gnomAD ClinVar assessment Alamut assessment Monogenic kidney disease genes F, 42 Minimal change glomerulo-nephritis  < 10RBC/ml, UAC 33.9, eGFR > 90 ml/min/1.73m 2 , no kidney cysts NO PKHD1, NM_170724.3 , Rs746838237 PKD4, with or without hepatic disease, MIM:63200, AR, AD c.5895_5896insA, p.Leu1966ThrfsTer4, LOVD:00472317 1.4 × 10 –4 Pathogenic Likely pathogenic (PM2,PM4) Score = 6 M, 57 FSGS 100 RBC, UAC 208.5, end-stage kidney failure, bilateral kidney cysts YES CFH, NM_001014975.3 Rs976333015 Susceptibility to atypical Haemolytic Uraemic disease, MIM:235400, AD, AR c.158G > A, p.Arg53His, LOVD:00472359 4.96 × 10 –6 Likely pathogenic Likely pathogenic (PM1,PM2,PP3,PP5) Score = 6 M, 47 Diabetic kidney disease 327 RBC, UAC 86.2, eGFR 13 ml/min/1.73m 2 , no cysts reported YES CHD7, NM_017780.4 Rs2487915295 CHARGE syndrome, MIM: 214800, AD c.4439_4440insCCCC, p.Lys1481ProfsTer2 LOVD:00472360 Not present Pathogenic Pathogenic (score = 11; PVS, PM2, PP5) M, 59 Hypertensive disease, diabetes 51 RBC, UPC 75, eGFR 64 ml/min/1.73m 2 , one cyst in one kidney and one cyst in spleen YES PKHD1, NM_170724.3 Rs746838237 PKD4, with or without hepatic disease, MIM: 263200, AR, AD c.5895dup, p.Leu1966ThrfsTer4; LOVD: 00472361 1.4 × 10 –4 Pathogenic Pathogenic (score = 11; PVS1, PM2, PP5) M, 63 Diabetic kidney disease and partial nephrectomy for renal cell carcinoma  < 10 RBC, UAC 323, eGFR15 ml/min/1.73m 2 , one cyst in one kidney NO SEC63, NM_007214.5 Rs752868449 Polycystic liver disease 2, MIM:617004, AD c.1605_1606insA, p.Pro536ThrfsTer24, LOVD: 00472362 7.0 × 10 –4 Pathogenic/ Likely pathogenic Pathogenic (score = 11; PVS1, PM2, PP5) M, 71 Membranous glomerulo-nephritis, diabetes  < 10 RBC, UAC 239, eGFR 44 ml/min/1.73m 2 , two cysts in one kidney NO LRP5, NM_002335.4 Rs371514699 Polycystic liver disease 4 with or without kidney cysts, MIM:617875, AD c.1057C > T, p.Arg353Trp, LOVD:00472363 1.05 × 10 –5 Pathogenic Likely Pathogenic (score = 6; PM1, PM2, PP3, PP5) F, 79 Renovascular disease, partial nephrectomy, Diabetes  < 10 RBC, UAC < 0.6, eGFR 21 ml/min/1.73m 2 , no kidney cysts NO COL4A3, NM_000091.5 Rs1414411811 Alport syndrome, MIM: 104200, AD c.272G > A, p.Gly91Asp, LOVD: 00472358 3.1 × 10 –6 Pathogenic Likely Pathogenic (score = 6; PM1, PM2, PP3, PP5) Monogenic diabetes genes M, 81 Diabetic kidney disease, kidney stones UAC 4.4, eGFR27 ml/min/1.73m 2 , one kidney cyst NO HNF1A, NM_178850.3 Rs762703502 Monogenic diabetes c.864del or p.Pro291GlnfsTer51, LOVD:00472364 8.7 × 10 –6 Likely Pathogenic (ClinGen monogenic diabetes VCEP)*** Likely pathogenic (PVS1, PP5, score = 9) F, 63 Myeloma, diabetic kidney disease UAC 301, eGFR < 10 ml/min/1.73m 2 , one kidney cyst NO NEUROD1, NM_002500.5 Rs387906384 Monogenic diabetes c.616_617insC, p.His206ProfsTer38, LOVD:00472365 1.49 × 10 –5 Conflicting (VUS/Pathogenic) VUS (PM2, score = 2) F, 46 Diabetic kidney disease UAC 75.6, eGFR 23 ml/min/1.73m 2 , one kidney cyst NO ABCC8, NM_000352.6 Rs745948796 Monogenic diabetes c.3775G>A; p.Val1259Met, LOVD:00472366 6.2 × 10 –5 Variant of Uncertain Significance VUS (PM1,PM2,PP3, score = 5) All genetic variants described here were heterozygous.

Table 2 Clinical features and disease-causing genetic variants in patients from a General Nephrology clinic. Patient Kidney disease Kidney phenotype Family history Gene, transcript, dbSNP Disease, MIM, MoI Variant and protein change , LOVD submission number MAF in gnomAD ClinVar assessment Alamut assessment Monogenic kidney disease genes F, 42 Minimal change glomerulo-nephritis  < 10RBC/ml, UAC 33.9, eGFR > 90 ml/min/1.73m 2 , no kidney cysts NO PKHD1, NM_170724.3 , Rs746838237 PKD4, with or without hepatic disease, MIM:63200, AR, AD c.5895_5896insA, p.Leu1966ThrfsTer4, LOVD:00472317 1.4 × 10 –4 Pathogenic Likely pathogenic (PM2,PM4) Score = 6 M, 57 FSGS 100 RBC, UAC 208.5, end-stage kidney failure, bilateral kidney cysts YES CFH, NM_001014975.3 Rs976333015 Susceptibility to atypical Haemolytic Uraemic disease, MIM:235400, AD, AR c.158G > A, p.Arg53His, LOVD:00472359 4.96 × 10 –6 Likely pathogenic Likely pathogenic (PM1,PM2,PP3,PP5) Score = 6 M, 47 Diabetic kidney disease 327 RBC, UAC 86.2, eGFR 13 ml/min/1.73m 2 , no cysts reported YES CHD7, NM_017780.4 Rs2487915295 CHARGE syndrome, MIM: 214800, AD c.4439_4440insCCCC, p.Lys1481ProfsTer2 LOVD:00472360 Not present Pathogenic Pathogenic (score = 11; PVS, PM2, PP5) M, 59 Hypertensive disease, diabetes 51 RBC, UPC 75, eGFR 64 ml/min/1.73m 2 , one cyst in one kidney and one cyst in spleen YES PKHD1, NM_170724.3 Rs746838237 PKD4, with or without hepatic disease, MIM: 263200, AR, AD c.5895dup, p.Leu1966ThrfsTer4; LOVD: 00472361 1.4 × 10 –4 Pathogenic Pathogenic (score = 11; PVS1, PM2, PP5) M, 63 Diabetic kidney disease and partial nephrectomy for renal cell carcinoma  < 10 RBC, UAC 323, eGFR15 ml/min/1.73m 2 , one cyst in one kidney NO SEC63, NM_007214.5 Rs752868449 Polycystic liver disease 2, MIM:617004, AD c.1605_1606insA, p.Pro536ThrfsTer24, LOVD: 00472362 7.0 × 10 –4 Pathogenic/ Likely pathogenic Pathogenic (score = 11; PVS1, PM2, PP5) M, 71 Membranous glomerulo-nephritis, diabetes  < 10 RBC, UAC 239, eGFR 44 ml/min/1.73m 2 , two cysts in one kidney NO LRP5, NM_002335.4 Rs371514699 Polycystic liver disease 4 with or without kidney cysts, MIM:617875, AD c.1057C > T, p.Arg353Trp, LOVD:00472363 1.05 × 10 –5 Pathogenic Likely Pathogenic (score = 6; PM1, PM2, PP3, PP5) F, 79 Renovascular disease, partial nephrectomy, Diabetes  < 10 RBC, UAC < 0.6, eGFR 21 ml/min/1.73m 2 , no kidney cysts NO COL4A3, NM_000091.5 Rs1414411811 Alport syndrome, MIM: 104200, AD c.272G > A, p.Gly91Asp, LOVD: 00472358 3.1 × 10 –6 Pathogenic Likely Pathogenic (score = 6; PM1, PM2, PP3, PP5) Monogenic diabetes genes M, 81 Diabetic kidney disease, kidney stones UAC 4.4, eGFR27 ml/min/1.73m 2 , one kidney cyst NO HNF1A, NM_178850.3 Rs762703502 Monogenic diabetes c.864del or p.Pro291GlnfsTer51, LOVD:00472364 8.7 × 10 –6 Likely Pathogenic (ClinGen monogenic diabetes VCEP)*** Likely pathogenic (PVS1, PP5, score = 9) F, 63 Myeloma, diabetic kidney disease UAC 301, eGFR < 10 ml/min/1.73m 2 , one kidney cyst NO NEUROD1, NM_002500.5 Rs387906384 Monogenic diabetes c.616_617insC, p.His206ProfsTer38, LOVD:00472365 1.49 × 10 –5 Conflicting (VUS/Pathogenic) VUS (PM2, score = 2) F, 46 Diabetic kidney disease UAC 75.6, eGFR 23 ml/min/1.73m 2 , one kidney cyst NO ABCC8, NM_000352.6 Rs745948796 Monogenic diabetes c.3775G>A; p.Val1259Met, LOVD:00472366 6.2 × 10 –5 Variant of Uncertain Significance VUS (PM1,PM2,PP3, score = 5) All genetic variants described here were heterozygous.

Clinical features and disease-causing genetic variants in patients from a General Nephrology clinic.

PKHD1,

NM_170724.3 ,

Rs746838237

c.5895_5896insA,

p.Leu1966ThrfsTer4,

LOVD:00472317

CFH,

NM_001014975.3

Rs976333015

c.158G > A,

p.Arg53His, LOVD:00472359

CHD7, NM_017780.4

Rs2487915295

c.4439_4440insCCCC, p.Lys1481ProfsTer2

LOVD:00472360

PKHD1,

NM_170724.3

Rs746838237

SEC63, NM_007214.5

Rs752868449

Pathogenic/

Likely

pathogenic

LRP5,

NM_002335.4

Rs371514699

c.1057C > T,

p.Arg353Trp,

LOVD:00472363

COL4A3,

NM_000091.5

Rs1414411811

c.272G > A,

p.Gly91Asp,

LOVD: 00472358

HNF1A, NM_178850.3

Rs762703502

c.864del or p.Pro291GlnfsTer51,

LOVD:00472364

NEUROD1, NM_002500.5

Rs387906384

ABCC8,

NM_000352.6

Rs745948796

All genetic variants described here were heterozygous.

Family history

Three people with a pathogenic variant (43%) had another affected family member with kidney disease and four (57%) did not. Those with a pathogenic variant and another affected family member represented three of the six people with a positive family history. Two of the other three with an affected family member and no pathogenic variant had diabetic kidney disease. Three of the overall six with an affected family member had end-stage kidney failure.

Clinical features including kidney failure

Genetic kidney disease was found associated with all kidney phenotypes: diabetic kidney disease, haematuria, kidney stones and renal calcification, FSGS, kidney cysts and kidney failure. A pathogenic variant was found in two of the 6 people with an eGFR > 60 ml/min/1.73 m 2 , 2 of the 44 with an eGFR 20 – 60 ml/min/1.73 m 2 and 3 of the 14 (%) with an eGFR < 20 ml/min/1.73 m 2 .

Genetic kidney disease

Seven variants were rare and assessed as Pathogenic or Likely pathogenic in ClinVar. All were heterozygous and associated with monoallelic or AD genetic disease. All were disease-causing except CFH that was a susceptibility gene for atypical Haemolytic Uraemic Syndrome. Four were truncating and three were missense changes.

Overall these 7 people with a genetic kidney disease had a known diagnosis of glomerulonephritis in 3 people (43%), hypertensive or renovascular disease in two (29%) and diabetes or diabetic kidney disease in 5 (71%). The clinical associations were haematuria in 3 people (43%), increased UAC in 6 (86%), and an eGFR  ≤ 60 ml/min/1.73m 2 in 5 (71%). Three patients (43%) had a positive family history of kidney disease, and three (43%) were found in people with kidney failure.

Only one variant, in CHD7, was associated with both kidney failure and a positive family history. Four variants were associated with cystic kidney disease but the number of cysts varied from none to one cyst in one kidney to multiple cysts in one kidney together with a splenic cyst. Only one person with a genetic variant associated with kidney cysts had kidney failure. None of the people with a pathogenic variant in the genetic kidney disease genes appeared to have syndromic features, and, in particular, they had no reported cardiac or retinal anomalies.

An Alport variant ( NM_000091.5 : COL4A3 , c.272G > A, p.Gly91Asp) affected a position 1 Gly in the intermediate collagenous domain consistent with the diagnosis of AD Alport syndrome (MIM:104200). It was present in a 79 year old woman with a diagnosis of renovascular disease, and a partial nephrectomy for cancer. She had no haematuria, proteinuria or kidney cysts, but an eGFR of 21 ml/min/1.73 m 2 .

The CFH variant is associated with an increased susceptibility to atypical HUS in the heterozygous form (MIM:235400). The affected person had FSGS, a family history of kidney disease, but no known genetic cause. This variant has been confirmed to be pathogenic experimentally 28 .

The CHD7 variant was consistent with the diagnosis of Hypogonadotrophic hypogonadism 5 with or without anosmia (MIM:612370), or hearing loss /CHARGE syndrome (MIM:214800) and the affected man had low testosterone levels and diabetic kidney disease. His hearing and sense of smell were normal. His mother also had kidney failure but no further clinical details were available. This variant was a frameshift change resulting in a stop codon. CHD7- associated disease has been reported with stop codons distributed throughout the gene 29 , as well as missense and splicing variants.

Finally, four people had pathogenic variants in genes associated with kidney cysts. Two affected PKHD1 (MIM:263200), and one each affected SEC63 (MIM:617004) and LRP5 (MIM:617875) which are typically associated with liver rather than kidney cysts. Only three of these four people had kidney cysts, and only one had kidney failure. One further individual had a NEK8 variant that was classified pathogenic, and while this gene is associated with AR nephronophthisis, a heterozygous NEK8 variant is only associated with kidney cysts when it affects the Ser-Thr kinase domain at the NEK8 amino terminus 30 . This patient had a renal cell carcinoma, for which he had undergone a right nephrectomy, and multiple cysts in the right kidney as well as a liver cyst. His eGFR was 72 ml/min/1.73 m 2 , and he was not included in the solved cases.

Genetic results in monogenic diabetes genes

Thirty-six participants had diabetes or diabetic kidney disease but only three (9%) had a possibly pathogenic variant in a gene associated with monogenic diabetes (Table 2 ). None of these three had a known family history of kidney disease.

Their clinical details were as follow. An 81 yo man with diabetic kidney disease and kidney stones, an eGFR of 27 ml/min/1.73m 2 , UAC of 4.4 mg/mmol, and one kidney cyst, had NM_178850.3 :c.864del or p.Pro291GlnfsTer51 in HNF1A that was classified as Likely pathogenic in ClinVar by the ClinGen Variant Curation Expert Panel and has experimental confirmation of pathogenicity 31 .

This man also had a variant in the ATP6V1C2, NM_001039362.1 :c.946_947del corresponding to p.Ser316Ter in a gene for distal renal tubular acidosis that predisposes to kidney stones. While this variant is considered pathogenic the associated disease has biallelic inheritance 32 so this variant was not included in the reported pathogenic variants in the cohort. While biallelic variants are often associated with severe classical forms of distal Renal Tubular Acidosis, monoallelic variants may also contribute to milder disease.

A 63 yo woman with myeloma and diabetes, an eGFR less than 10 ml/min/1.73m 2 , UAC of 301 mg/mmol, and one kidney cyst had a NEUROD1 variant, NM_002500.5 :c.616_617insC, p.His206ProfsTer38, that was classified as a VUS in Alamut and Conflicting (VUS/Pathogenic) in ClinVar 33 . This gene is associated with both Maturity -onset diabetes of the young 6 (MIM 606,394) and susceptibility to type 2 Diabetes mellitus (MIM125853).

A 46 yo woman with diabetic kidney disease, eGFR 23 ml/min/1.73m 2 , UAC 75.6 mg/mmol, and one kidney cyst had an ABCC8 variant, NM_000352.6 :c.3775G > A and p.Val1259Met in exon 31, that was considered a VUS in Alamut and ClinVar. Two other Val > Met substitutions in ABCC8 in exons 38 and 39 are associated in the heterozygous form with later-onset diabetes (MIM:621196 Maturity-onset diabetes of the young type 12) and Transient neonatal diabetes mellitus (MIM:610374) which is also associated with later onset diabetes) respectively. This person had an intellectual disability but no perinatal history of diabetes mellitus was available.

Discussion

This pilot study suggests that genetic kidney disease affects at least 11% of patients in the general nephrology clinic who already have another diagnosis and in whom a genetic cause was not suspected at three previous clinic visits. Pathogenic variants coding for atypical cystic kidney disease were commonest, but variants were also found in the CFH, CHD7 and COL4A3 genes, all of which demonstrated an AD mode of inheritance. Kidney function in the patients with an additional genetic kidney disease varied from normal to end-stage kidney failure so that any effect on kidney function was variable. More than half these patients had diabetes or probable diabetic kidney disease and three possibly pathogenic variants in monogenic diabetes genes were also identified.

However 11% is an underestimate for the proportion of patients in the general nephrology clinic with an undiagnosed genetic kidney disease because no pathogenic gene was identified in some people with a positive family history of kidney disease. In addition, people with clinically-obvious genetic disease such as CAKUT or ADPKD had already been excluded, and because WES, used here for variant detection, demonstrates only 80% of disease-causing variants 34 – 36 . Structural and Copy number changes that are common in MUC1, HNF1B and NPHP1, and deep intronic splice changes are commonly overlooked, and there are also further yet-to-be-discovered disease-causing genes.

People with a known or possible genetic kidney disease were deliberately excluded from this study. All patient records from the past three clinic visits were screened for a family history of genetic disease, but, surprisingly, six people told us at recruitment of another affected family member. Of these six with a positive family history, three had diabetes or diabetic kidney disease, and only two had a genetic cause identified by sequencing in this study ( CFH, CHD7 ). This observation indicates that asking the patient about a family history of kidney disease is useful even when recent physician records and the renal biopsy and imaging results do not suggest a genetic disease. Although patients already have a diabetic, autoimmune or other kidney diagnosis, a coincidental genetic cause may still be present.

Patients were a typical clinic cohort with a median age of 66 years, with the commonest diagnosis being diabetic kidney disease. A genetic variant was found in clinic patients with all renal phenotypes, including haematuria, proteinuria, FSGS, kidney failure, kidney stones and diabetic kidney disease, as well as a range of kidney function.

All people with a genetic kidney disease demonstrated here had a condition with monoallelic or AD inheritance. AD Alport syndrome is the commonest genetic kidney disease but was found in only one person and may not have affected kidney function. One variant was found in CFH which predisposes to atypical HUS and FSGS. Four people had a pathogenic or likely pathogenic variant in a gene for atypical AD polycystic kidney disease ( PKHD1, SEC63, LRP5). SEC63 and LRP5 usually result in AD polycystic liver disease with small kidney cysts. PKHD1 results in AR nephronophthisis in children but kidney cysts occur in adults in the monoallelic form. These four cystic kidney diseases do not typically cause kidney failure in the monoallelic form.

Three possibly pathogenic variants were demonstrated in the monogenic diabetes genes ( HNF1A, NEUROD1 and ABCC8) . All three individuals had diabetes and impaired kidney function, but none had a known family history of diabetes or diabetic kidney disease. These variants were assessed as Likely pathogenic, in one and a VUS in two cases according to ClinVar. While these variants’ pathogenicity is difficult to confirm and often requires a functional assay, in all cases they had been reported previously as disease-causing.

This pilot study was a research project and, as a condition of IRB approval, all participants in whom a pathogenic genetic variant was found were required to have the variant confirmed in an accredited testing laboratory. No clinical decisions could be made or segregation studies performed based only on our research findings. Confirmatory testing is currently underway but not all participants have returned for review.

It is still possible to consider the likely benefits and limitations of this study. Four patients had mutations in genes for atypical AD polycystic kidney disease that were unlikely to affect kidney function. Some of those with a variant did not even have kidney cysts possibly because of incomplete penetrance or variable expressivity. However their children who inherit the variant may still develop disease. One person had a variant in an Alport gene which is a risk factor for kidney failure. The significance of the CHD7 variant is still unclear. Nevertheless 5 of the 7 people with a variant in a genetic kidney disease gene had an eGFR less than 60 ml/min/1.75 m 2 .

The three variants in the monogenic diabetes genes all occurred in people with Diabetes type 2, and have been previously cited as disease-causing but their pathogenicity has not been confirmed in functional assays.

This study’s major strength was its novelty in identifying genetic kidney disease in a cohort without a suspected genetic diagnosis. Many of these people had diabetes but diabetes is common in a renal population and the diagnosis of diabetic kidney disease is often assumed rather than proven. The finding of a monogenic cause for diabetes in 3 people was unexpected.

The study’s limitations were that it was a single centre study with a relatively small and predominantly older adult cohort, with a possibly unintended bias in recruitment. Many participants had kidney failure, and previous series have suggested that at least 10% of people with kidney failure have a genetic basis. However these have often included people with obvious genetic disease such as ADPKD 5 . The approach to identifying people with likely non-genetic disease was standardised but did not include asking patients about another affected family member. We cannot be certain that more patients did not have a known family history. However the aim of this study was not to identify how many people in the renal clinic had a family history of kidney disease but rather to determine how many people without a recorded family history had a pathogenic variant detected. Finally, we did not find disease-causing variants in everyone in this study with a family history of genetic kidney disease or diabetes.

Future studies should include multiple sites and ask patients for a family history of kidney disease in addition to the screening undertaken here. Nevertheless incomplete penetrance and variable expressivity may mean that genetic kidney disease may not be recognised in other family members. Younger and more diverse cohorts could be tested too. Genetic kidney disease may be even more common in a younger cohort or where fewer people have diabetes.

These results suggest that genetic kidney disease is common in the nephrology clinic, and may coexist with other non-genetic kidney diseases. Genetic kidney disease may be so common that renal physicians should be suspicious when their patients have atypical clinical features or a positive family history, even where there is already a diagnosis of autoimmune, diabetic or other kidney disease. However some common genetic diseases such as atypical cystic kidney disease do not appear to significantly affect function. No kidney gene panels currently include monogenic diabetes genes, but the finding of a variant in one of these genes helps predict unsuspected extrarenal features, optimise diabetes treatment and diagnose other family members sooner.

Supplementary Information

Supplementary Information.

Supplementary Information.