Renal Prognostic Identification in Patients With Autosomal Dominant Polycystic Kidney Disease by Whole Genome Sequencing: A Prospective, Observational Study

Abstract

Rationale & Objective

The association between autosomal dominant polycystic kidney disease (ADPKD) genetic variants and renal prognosis remains unclear. We conducted whole genome sequencing to identify the factors contributing to disease severity.

Study Design

Prospective, observational study.

Setting & Population

Using data collected from a 2-year prospective cohort of 200 patients with ADPKD, gene mutations were identified using whole genome sequencing.

Exposure

None.

Outcomes

The primary endpoint was the rate of increase in total kidney volume. The secondary endpoints were composite renal endpoints (induction of dialysis, kidney transplantation, or a decrease in the estimated glomerular filtration rate of ≥25%).

Analytical Approach

Logistic regression analyses were performed to determine the factors associated with the outcomes.

Results

Among 169 patients for whom genetic diagnosis was performed, genetic mutations were identified in 144 cases, with 109 (75.7%) PKD1, 34 (23.6%) PKD2, and 1 (0.7%) GANAB variants identified. The median annual increase in total kidney volume was 5.9%. Among the patients who were followed, 60 patients (33.5%) achieved the composite renal endpoint. The independent risk factors for reaching the renal composite endpoint were estimated glomerular filtration rate at enrollment (OR, 0.93; 95% CI, 0.91-0.96) and PKD1 truncation (OR, 3.05; 95% CI, 1.11-8.40). Hypertension and overweight exacerbated disease severity, particularly in patients with PKD1 truncation. The annual rate of kidney function deterioration was higher in the order of PKD1 truncating, PKD1 non-truncating, PKD2 truncating, and PKD2 non-truncating variants. The rate of Mayo imaging classification 1C-1E was highest in the same order.

Limitations

Owing to the various PKD variants, the sample size for each variant was insufficient for comprehensive evaluation of kidney function.

Conclusions

PKD1 truncation is a sensitive severity marker in patients with ADPKD, and PKD2 non-truncation is the least severe. Genetic diagnosis is useful for predicting renal prognosis.

Plain-Language Summary

This prospective study included 200 patients with autosomal dominant polycystic kidney disease (ADPKD), with the objectives of the following: (1) ascertaining the feasibility of conducting comprehensive ADPKD genetic diagnosis using whole genome sequencing with next-generation sequencing, and (2) identifying factors associated with renal prognosis and disease severity. This study showed that differences in genotype influence the rate of kidney function deterioration and disease severity. Furthermore, we found that hypertension and being overweight are particularly important factors that worsen the severity of ADPKD. Consequently, blood pressure and weight management are crucial in patients with ADPKD.

Autosomal dominant polycystic kidney disease (ADPKD) is a progressive disease in which cysts increase with age and kidney function gradually declines. Approximately 50% of patients develop end-stage kidney failure by the age of 70 years.¹ However, the rate of kidney function decline in ADPKD varies widely among individuals.

Previous studies investigating the relationship between ADPKD and renal prognosis have elucidated the effects of genetic mutations, hypertension, increased total kidney volume (TKV), and urinary abnormalities.2, 3, 4, 5, 6 PKD1 (16p13.3) and PKD2 (4q21) have been identified as causative genes of ADPKD.⁷ PKD1 mutations occur more frequently than PKD2 mutations and have been associated with a poorer renal prognosis.^8,9 Recent studies have identified additional genes involved in ADPKD, including GANAB, DNAJB11, HNF1B, IFT140, SEC61B, ALG8, and ALG9.^7,10 These recent findings highlight the importance of conducting comprehensive genetic diagnosis in ADPKD cases.

However, the identification rate of genetic mutations in ADPKD remains insufficient owing to the wide gene region (47.2 kb), large number of exons (46), inclusion of 6 pseudogenes, and high GC content. Whether the detected mutations are pathological needs to be confirmed using various databases and publications, and the diagnosis often remains unclear, resulting in the detected mutations being regarded as of unknown significance. Recently, the Mayo classification was proposed and used to classify the severity of ADPKD; however, the relationship between genetic variants and disease severity has not yet been addressed. In addition, the relationship between different genetic variants and factors that worsen kidney function is not fully understood.

Therefore, the objectives of this 2-year prospective study of Japanese patients with ADPKD were as follows: (1) to determine whether comprehensive ADPKD genetic diagnosis can be conducted by whole genome sequencing (WGS) using next-generation sequencing (NGS), and (2) to identify specific genomic mutations associated with renal prognosis and severity.

Methods

Study Design

This was a multicenter prospective cohort study. We enrolled 200 patients with ADPKD from 5 facilities in which nephrologists were practicing in the Ibaraki Prefecture, Japan. The enrollment period was from October 2014 to December 2020, with a 2-year follow-up period after enrollment. The basis for calculating the number of cases was cyst and TKV growth with worsening kidney function in patients with ADPKD. Assuming that the rate of increase is approximately 2%-5% per year,^11,12 the difference in the rate of increase between the group with fast and slow increase in renal volume is 3%, the standard deviation is 6.5,^12,13 in which α is 5% bilaterally, the power is 80%, and the dropout rate is 20%, the required sample size is 195 cases; thus, the target number of cases was set at 200 cases. Blood counts, biochemistry, and urinalysis were performed every 6 months for 2 years. The following clinical data were collected: age, sex, family history, medical history, height, weight, body mass index, systolic and diastolic blood pressure, TKV, echocardiography, brain aneurysm, and gene mutation.

Analyzed Population

Eligibility Criteria

Patients with ADPKD who were currently receiving treatment at the participating centers, including newly diagnosed patients, were included in the study. ADPKD was diagnosed on the basis of the Pei classification.¹⁴

Exclusion Criteria

Patients who could not provide informed consent, those who had already undergone maintenance dialysis or kidney transplantation, and those deemed inappropriate for inclusion by the responsible physician or study investigator were excluded.

Predictors and Outcomes

Primary Endpoint

The primary endpoint was the annual TKV growth rate (percentage per year).

Secondary Endpoint

The secondary endpoints included the rate of deterioration of kidney function by estimated glomerular filtration rate (eGFR) (mL/min/1.73 m²/y); eGFR (mL/min/1.73 m²) = 194 × Cr^−1.094 × age^−0.287 [× 0.739 (female)] (Japanese GFR estimation formula¹⁵); renal composite endpoint (induction of dialysis, kidney transplant, decrease in eGFR of ≥25%), death; renal replacement therapy; subarachnoid hemorrhage; cerebral infarction or hemorrhage; cardiac disease [myocardial infarction, angina pectoris, heart failure, peripheral arterial stenosis]; severe liver injury [total bilirubin elevation >2 times the upper reference limit or serum aspartate aminotransferase or alanine aminotransferase elevation >3 times the upper reference limit]; infection requiring hospitalization; diverticulitis; and renal artery embolization.

Evaluation of TKV

TKV was measured at the time of enrollment and measured again 2 years later to calculate the annual growth rate. For patients who reached endpoints, such as death or dialysis induction, and for whom 2-year kidney volume measurements were unavailable, we estimated the annual growth rate using a formula based on age at enrollment and height-corrected kidney volume. The Ht-TKV–estimated annual growth rate α (percentage per year) was derived from the equation Ht-TKVt = 150(1 + α/100)^t, in which Ht-TKVt is the Ht-TKV at age t.¹⁶ The kidney volume was measured using VINCENT software (FUJIFILM Co, Japan).

Mutation Analysis

Genomic DNA was extracted from peripheral blood using the QIAamp DNA blood maxi kit (Qiagen). Mutational analyses of PKD1 and PKD2 were performed using the NGS method, WGS analysis. Sequence data were analyzed based on germline short variant discovery (single nucleotide polymorphisms + indels) of the GATK Best Practice. That is, the sequence reads were mapped to a human reference genome hg38 by Burrows-Wheeler Aligner with Maximal Exact Matches algorithm.¹⁷ Using GATK Tools,^18,19 duplications of reads were removed and base quality scores were recalibrated. Single nucleotide polymorphisms and short indels of each sample were called, and the mutations of all samples were gathered into a vcf file by joint call. Mutations were filtered using the GATK standard filter and variant quality score recalibration. Mutations were annotated using snpEff,²⁰ and those in PKD1 and PKD2 were identified and evaluated.

Variant Nomenclature

All variants were numbered according to the guidelines of the Human Genome Variation Society (http://www.hgvs.org/mutnomen) using the NCBI reference sequences NM_001009944.3 for PKD1, NM_000297.4 for PKD2, and NM_001278192.2 for GANAB.

Bioinformatic Analysis and Interpretation of Variant Pathogenicity

To assess the pathogenicity of the missense variants, we used a comprehensive database of transcript-specific functional predictions (dbNSFP; https://sites.google.com/site/jpopgen/dbNSFP). In dbNSFP, we referred to SIFT (https://sift.bii.astar.edu.sg/), PolyPhen-2 (http://genetics.bwh.harvard.edu/pph2/), MutationTaster (https://www.mutationtaster.org/), PROVEAN (https://www.jcvi.org/research/provean), M_CAP (http://bejerano.stanford.edu/mcap/), and CADD (https://cadd.gs.washington.edu/snv) for in silico prediction tools, as well as the 1000 Genomes Project (https://www.internationalgenome.org/) and Genome Aggregation Database (gnomAD, https://gnomad.broadinstitute.org/) for healthy databases. We also referred to the ADPKD variant database (https://pkdb.mayo.edu/welcome), LOVD (https://databases.lovd.nl/shared/genes), and ClinVar (https://www.ncbi.nlm.nih.gov/clinvar/) as disease databases. The variants were classified as likely pathogenic with uncertain significance according to the guidelines of the American College of Medical Genetics and Genomics (Tables S1-S3).²¹ In this study, all variants of uncertain significance were considered diagnostic if they were associated with a typical ADPKD phenotype. Nonsense, frameshift, and canonical splice site variants were grouped as truncating variants, whereas nonsynonymous missense and in-frame insertion/deletion variants were grouped as non-truncating variants (Tables S4 and S5). All variants were verified by direct sequencing²² or long-range polymerase chain reaction (PCR)-based NGS (primer sequences for long-range PCR are listed in Table S6,²³ from the Kazusa DNA Laboratory catalog, Chiba, Japan).

Statistical Analysis

Patient demographics are summarized as the mean and standard deviation or median and interquartile range. Group differences were evaluated using unpaired t tests, Mann–Whitney U tests, Fisher exact tests, or χ² tests, and one-way analysis of variance, as appropriate. Logistic regression analyses were performed to determine the factors associated with the outcomes. Linear regression analysis was performed to calculate the annual rate of kidney function deterioration. For the multivariable analysis of the renal composite endpoint, a logistic regression analysis was performed using each variable. P values < 0.05 were considered statistically significant. Statistical power was calculated using R statistical software (version 4.1.2) and EZR, which is a modified version of the R commander designed to add statistical functions frequently used in biostatistics.²⁴

Ethics Statements

The study protocol was approved by the University of Tsukuba Hospital institutional review board (H26-059) and the institutional review boards of each participating facility. The study procedures adhered to the principles of the Declaration of Helsinki and were registered with the University Hospital Medical Information Network (UMIN000014674).

Results

Patient Background of All Patients in the Cohort

The study patients are shown in Fig S1. A total of 200 patients were enrolled, after excluding 19 who relocated, discontinued their visits, or dropped out for other reasons. During the 2-year observation period, 1 patient died, and another patient was transferred to another hospital owing to subarachnoid hemorrhage. The patient backgrounds are presented in Table 1. A family history was identified in 150 patients (82.9%), whereas 31 patients (17.1%), including unknown patients, had no confirmed family history. In this study, 20 families had multiple patients registered. Genetic mutations were found to be identical within the same family. Hypertension was a common complication, and several patients required treatment at a young age, even in the absence of kidney dysfunction. Furthermore, with a mean eGFR of 53.5 mL/min/1.73 m² and a mean Ht-TKV of 1,035 mL/m, many patients already had severe kidney dysfunction. Cerebral aneurysms were present in approximately 15% of the patients, and the proportion of patients with valvular heart disease was high (approximately 33%). Regarding medical history, subarachnoid hemorrhage was present in 6 (3.3%), cerebral hemorrhage/stroke in 8 (4.4%), and cardiac disease in 7 (3.9%) patients.

Table 1.

Baseline characteristics of patients with ADPKD in this study

Characteristic
N	181
Age, y	49.9 (12.7)
Sex (male)	88 (48.6%)
BMI, kg/m2	24.0 (4.1)
Systolic arterial pressure, mm hg	133.8 (16.6)
Diastolic arterial pressure, mm hg	81.7 (11.8)
Serum creatinine, mg/dL	1.55 (1.4)
eGFR, mL/min/1.73m2
Mean (SD)	53.5 (28.6)
Median (IQR)	52.6 (30.6-73.6)
TKV, mL/m
Mean (SD)	1,702 (1,297)
Median (IQR)	1,339 (838-2,226)
Ht-TKV, mL/m
Mean (SD),	1035 (793)
Median (IQR)	836 (526–1,337)
Family history	150 (82.9%)
Hypertension	136 (75.1%)
Brain aneurysm	27/179 (15.1%)
Valvular heart disease	38/115 (33.0%)
History of subarachnoid hemorrhage	6 (3.3%)
History of cerebral infarction or hemorrhage	8 (4.4%)
History of cardiac disease	7 (3.9%)
History of severe hepatic injury	1 (0.6%)
History of cyst infection	23 (12.7%)
History of diverticulitis	2 (1.1%)
PKD1 genotype	109/144 (75.7%)
PKD2 genotype	34/144 (23.6%)
PKD1 truncating	70/144 (48.6%)
GANAB genotype	1/144 (0.7%)

Note: Values for categorical variables are given as n (%) or n/N (%); values for continuous variables are given as means (SD) or median (IQR).

Abbreviations: ADPKD, autosomal dominant polycystic kidney disease; BMI, body mass index; eGFR, estimated glomerular filtration rate; Ht-TKV: height-adjusted total kidney volume; IQR, interquartile range; SD, standard deviation; TKV, total kidney volume.

Genetic Diagnosis by WGS

Among the 181 patients enrolled in the cohort, WGS by NGS was performed for 169 patients. A total of 148 mutations were detected by NGS, of which 144 (97.3%) were confirmed by Sanger sequencing and long-range PCR. Thus, the ADPKD-responsible gene mutation was identified by WGS in 144 of the 169 patients (85.2%). WGS showed that 1 (0.7%) patient had a GANAB mutation, 109 (75.7%) had PKD1 mutations, and 34 (23.6%) had PKD2 mutations. Truncating and non-truncating forms of the PKD1 gene were observed in 70 (64.2%) and 39 (35.8%) patients, respectively, whereas truncating and non-truncating forms of the PKD2 gene were observed in 29 (85.3%) and 5 (14.7%) patients, respectively (Table 2). We were unable to identify any other minor genetic abnormalities involved in ADPKD pathogenesis. PKD2 showed a significantly higher proportion of truncating-type mutations 85.3% than PKD1 64.2% (P = 0.02). Among the truncating types, PKD1 had the highest proportion of frameshift variants, and PKD2 had the highest proportion of stop-gained variants, but there was no statistically significant difference between the 2.

Table 2.

Genetic Variant Differences Between PKD1 and PKD2

Mutation type	Entire	PKD1	PKD2	P
	n = 143	n = 109	n = 34
Truncating	99 (69.2%)	70 (64.2%)	29 (85.3%)	0.02
Splice variant	7 (4.9%)	3 (2.8%)	4 (11.8%)	0.06
Frameshift variant	50 (35.0%)	38 (34.9%)	12 (35.3%)	> 0.99
Stop-gained	42 (29.4%)	29 (26.6%)	13 (38.2%)	0.20
Nontruncating	44 (30.8%)	39 (35.8%)	5 (14.7%)	0.02
Missense variant	36 (25.2%)	31 (28.4%)	5 (14.7%)	0.12
Conservative in-frame ins/del, n (%)	8 (5.6%)	8 (7.3%)	0 (0%)	0.20

Note: Values are n (%).

Abbreviation: ins/del, insertion or deletion.

Primary and Secondary Endpoints

The mean and median annual renal volume gains, as primary endpoints, were 6.0% and 5.9%, respectively (Table 3). Regarding complications during the follow-up period, 8 (4.4%) patients received renal replacement therapy, 7 (3.9%) had cardiovascular disease, and 3 (1.7%) had subarachnoid hemorrhage because of a ruptured cerebral aneurysm. All 8 cases of serious liver injury were side effects due to tolvaptan administration, and all recovered normally after administration was discontinued. Cystic infections requiring hospitalization were a frequent complication in 12 (6.6%) patients, and 1 patient died of worsening hepatic cystic infection. Next, we examined the number of patients who reached the composite renal endpoint (dialysis, kidney transplantation, or ≥25% reduction in eGFR) and found that 60 (33.5%) of 179, or approximately one-third of all patients, reached the composite renal endpoint. We observed differences in hypertension, eGFR, Ht-TKV, rate of kidney function deterioration, proteinuria, PKD1 truncation, and tolvaptan in the group that reached the composite endpoint compared with the group that did not (Table 4). Multivariable analysis showed that baseline kidney function (odds ratio [OR], 0.93; 95% confidence interval [CI], 0.91-0.96; P < 0.001) and PKD1 truncation (OR, 3.05; 95% CI, 1.11-8.40; P = 0.03) were independent risk factors for the renal composite endpoint (Table 5). In addition, a higher proportion of patients who reached the renal composite endpoint received tolvaptan; however, this was not a risk factor in the multivariable analysis.

Table 3.

Primary and Secondary Endpoints During the Follow-up in This Study

Primary Endpoint
Total kidney volume growth rate, %/y
Mean (SD)	6.0 (5.9)
Median (IQR)	5.9 (1.8-9.0)
Secondary Endpoint
Kidney function deterioration rate, mL/min/1.73 m2/y	−4.11 (2.95)
Renal replacement therapy or ≥25% eGFR reduction	60 (33.5%)
Death	1 (0.6%)
Renal replacement therapy	8 (4.4%)
Subarachnoid hemorrhage	3 (1.7%)
Cerebral infarction or hemorrhage	2 (1.1%)
Cardiac disease	2 (1.1%)
Severe hepatic injury	8 (4.4%)
Cyst infection	12 (6.6%)
Diverticulitis	2 (1.1%)
Renal artery embolization	1 (0.6%)

Note: Values for categorical variables are given as n (%); values for continuous variables are given as mean (SD) or median (IQR).

Abbreviations: eGFR, estimated glomerular filtration rate; IQR, interquartile range; SD, standard deviation.

Table 4.

Renal Composite Endpoint

	None	Positive	P
Number	119	60
Age, y, mean (SD)	48.7 (13.3)	52.2 (11.4)	0.09
Sex (male), n (%)	54 (45.4%)	33 (55.0%)	0.27
Hypertension, n (%)	80 (67.2%)	55 (91.7%)	<0.001
BMI, kg/m2, mean (SD)	23.6 (3.7)	24.8 (4.8)	0.06
eGFR, mL/min/1.73 m2, mean (SD)	64.6 (25.0)	30.9 (20.4)	<0.001
Log Ht-TKV, log mL/m, mean (SD)	2.86 (0.27)	3.03 (0.30)	<0.001
Kidney function deterioration rate, mL/min/1.73 m2/y, mean (SD)	−3.43 (2.81)	−5.44 (2.79)	<0.001
Total kidney volume growth rate, %/y, mean (SD)	6.1 (5.5)	5.8 (6.9)	0.72
Proteinuria, n (%)	10 (8.4%)	22 (36.7%)	<0.001
History of CVD, n (%)	13 (10.9%)	9 (15.0%)	0.47
History of cyst infection, n (%)	16 (13.4%)	14 (23.3)	0.13
PKD1, n (%)	70/97 (72.2%)	38/45 (84.4%)	0.14
PKD1 truncating, n (%)	40/97 (41.2%)	30/45 (66.7%)	0.007
PKD2, n (%)	26/97 (26.8%)	7/45 (15.6%)	0.20
PKD2 truncating, n (%)	21/97 (21.6%)	7/45 (15.6%)	0.50
GANAB, n (%)	1/97 (1.0%)	0/45 (0)	>0.99
Tolvaptan, n (%)	18 (15.1%)	20 (33.3%)	0.007

Note: Renal composite endpoint includes induction of dialysis, kidney transplant, or eGFR decline of 25% or more. Age, body mass index, eGFR, Log Ht-TKV, and proteinuria are the values at the time of study enrollment. Proteinuria defined + or more by urine dipstick test.

Abbreviations: BMI, body mass index; CVD, cardiovascular disease; eGFR, estimated glomerular filtration rate; Log Ht-TKV, log height-adjusted total kidney volume.

Table 5.

Multivariable Logistic Models Predicting the Odds of Renal Composite Endpoint

	Model 1	P	Model 2	P
	Adjusted OR (95% CI)		Adjusted OR (95% CI)
Age (per 1-y increment)	0.96 (0.91-1.01)	0.08	0.96 (0.91-1.01)	0.15
Sex (male)	0.70 (0.24-2.10)	0.53	0.64 (0.21-1.96)	0.43
Hypertension	1.87 (0.39-9.06)	0.43	1.72 (0.35-8.52)	0.51
PKD1 truncating	3.05 (1.11-8.40)	0.03	2.78 (1.00-7.75)	0.05
Proteinuria	2.41 (0.63-9.15)	0.20	2.88 (0.74-11.20)	0.13
eGFR (per 1 mL/min/1.73m2 increment)	0.93 (0.91-0.96)	<0.001	0.94 (0.91-0.97)	<0.001
Log Ht-TKV (per 10 mL/m increment)	4.83 (0.43-54.72)	0.20	3.85 (0.34-43.50)	0.28
Tolvaptan	—	—	2.16 (0.71-6.62)	0.18

Note: Model 1, adjusted for age, sex, hypertension, PKD1 truncating, proteinuria, eGFR, and log Ht-TKV. Model 2, adjusted for model 1 variables and tolvaptan.

Abbreviations: CI, confidence interval; eGFR, estimated glomerular filtration rate; Log Ht-TKV, log height-adjusted total kidney volume; OR, odds ratio.

Genetic Variant and Kidney Function

To further examine the differences in clinical background by PKD1 truncating, PKD1 non-truncating, PKD2 truncating, and PKD2 non-truncating genetic variants, we examined kidney function and TKV at enrollment in each of these categories. Our results showed no significant differences in sex, Ht-TKV, or eGFR, but patients with the PKD1 truncating genotype were significantly younger than those with the other genotypes (Table S7). Furthermore, the annual rate of kidney function deterioration decreased in the order of PKD1 truncating, PKD1 non-truncating, PKD2 truncating, and PKD2 non-truncating variants (−4.86 ± 3.08, −4.24 ± 2.25, −3.19 ± 3.00, and −2.83 ± 1.91; P = 0.02) (Fig 1). Severity of decrease in annual rate of kidney function deterioration among the gene mutations was PKD1 splice, PKD1 stop-gained, PKD2 splice, PKD1 frameshift, PKD1 conservative insertion/deletion, PKD1 missense, PKD2 stop-gained, PKD2 frameshift, PKD2 missense, in that order, but the difference was not statistically significant (P = 0.27) (Fig S2). Patients with GANAB mutations did not show any deterioration of kidney function, with an eGFR of 76.3 mL/min/1.73 m² at study entry and 80.4 mL/min/1.73 m² at 2 years.

Figure 1.

Differences in the rate of deterioration of kidney function by PKD1/PKD2 genotype. Annual rate of kidney function deterioration in PKD1 truncation, PKD1 non-truncation, PKD2 truncation, and PKD2 non-truncation groups. PKD1 truncation is associated with the highest rate of kidney function deterioration. The horizontal axis represents the PKD1/PKD2 genotype, and the vertical axis represents the annual rate of kidney function deterioration.

Gene Mutation and Severity

When we examined the background differences between Mayo imaging classification 1A-1B and 1C-1E, we found that although the baseline kidney function was the same, patients with 1C-1E were younger (55.8 ± 12.6 vs 46.4 ± 11.5 years; P < 0.001), had a higher rate of kidney function deterioration (−3.14 ± 2.94 vs −4. 67 ± 2.82 mL/min/1.73 m²; P < 0.001), and a higher increase in kidney volume (4.8 ± 5.1 vs 6.7 ± 6.3%/y; P = 0.03) (Table S8). Next, we examined the association between ADPKD severity and type of gene displacement. The number and percentage of patients with 1C-1E, considered severe in the Mayo imaging classification at the beginning of the study, were 53 (75.7%), 24 (63.2%), 13 (46.4%), and 1 (20.0%), respectively, in the order of PKD1 truncating, PKD1 non-truncating, PKD2 truncating, and PKD2 non-truncating variants (P = 0.006), representing a statistically significant difference (Fig 2).

Figure 2.

Percentage of Mayo 1C-1E by genetic variant differences. Rate of Mayo 1C-1E expression in PKD1 truncation, PKD1 non-truncation, PKD2 truncation, and PKD2 non-truncation groups. PKD1 truncation had the highest percentage of Mayo 1C-1E. The horizontal axis represents the percentage of Mayo 1C-1E, and the vertical axis shows the genotype.

Influences of Hypertension and Being Overweight

We examined whether the presence or absence of hypertension was associated with kidney function and TKV (Table 6). The rate of reaching the renal composite endpoint was significantly higher in the hypertension group than that in the normotension group (40.8% vs 7.9%; P <0.001). In addition, the hypertensive group had a lower eGFR (46.7 ± 25.0 vs 78.1 ± 24.2 mL/min/1.73 m²; P <0.001) and a higher log Ht-TKV at the time of enrollment (2.98 ± 0.26 vs 2.71 ± 0.22 log mL/m; P <0.001). We also compared the presence or absence of hypertension among the 4 genotypes. We found that the proportion of patients with PKD1 truncation who reached the renal composite endpoint was significantly higher (52.8% vs 11.8%; P = 0.004). Additionally, log Ht-TKV at the time of enrollment was significantly higher (2.99 ± 0.23 vs 2.73 ± 0.20 log mL/m; P < 0.001). Furthermore, the rate of growth in TKV was higher (6.4 ± 5.9 vs 3.1 ± 5.3 %/y; P = 0.04).

Table 6.

Differences in Kidney Function and Total Kidney Volume Depending on the Presence or Absence of Hypertension

	Number		Differences in kidney function depending on the presence or absence of hypertension									Differences in total kidney volume depending on the presence or absence of hypertension
			Renal composite endpoint, n (%)			eGFR at enrollment, mL/min/1.73 m2 mean (SD)			Kidney function deterioration rate, mL/min/1.73 m2/y mean (SD)			Log Ht-TKV at enrollment, mean (SD), log mL/m			Total kidney volume growth rate, mean (SD), %/y
	HT	NT	HT	NT	P	HT	NT	P	HT	NT	P	HT	NT	P	HT	NT	P
Total	103	38	42 (40.8)	3 (7.9)	<0.001	46.7 (25.0)	78.1 (24.2)	<0.001	−4.48 (2.86)	−3.76 (2.91)	0.19	2.98 (0.26)	2.71 (0.22)	<0.001	6.7 (5.7)	4.5 (5.3)	0.04
PKD1 truncating	53	17	28 (52.8)	2 (11.8)	0.004	46.1 (27.6)	78.9 (27.1)	<0.001	−5.14 (3.11)	−3.97 (2.89)	0.17	2.99 (0.23)	2.73 (0.20)	<0.001	6.4 (5.9)	3.1 (5.3)	0.04
PKD1 non-truncating	28	10	7 (25.0)	1 (10.0)	0.65	46.4 (22.7)	75.0 (28.4)	0.003	−4.16 (2.01)	−4.46 (2.94)	0.73	2.98 (0.25)	2.72 (0.27)	0.009	8.6 (5.4)	5.6 (5.8)	0.15
PKD2 truncating	19	9	7 (36.8)	0 (0)	0.06	47.3 (22.7)	81.6 (17.4)	<0.001	−3.44 (2.94)	−2.67 (3.23)	0.54	2.98 (0.33)	2.67 (0.26)	0.021	5.2 (4.4)	5.6 (4.9)	0.81
PKD2 non-truncating	3	2	0 (0)	0 (0)	>0.99	54.4 (17.6)	70.8 (5.0)	0.31	−2.49 (2.38)	−3.35 (1.54)	0.69	2.83 (0.29)	2.67 (0.06)	0.55	5.4 (10.4)	5.9 (3.5)	0.95

Note: Values for categorical variables are given as n (%); values for continuous variables as mean (SD).

Abbreviations: eGFR, estimated glomerular filtration rate; HT, patients with hypertension and taking antihypertensive drugs; Ht-TKV, height-adjusted total kidney volume; NT, normotensive patients.

Next, we investigated whether overweight status was associated with kidney function and TKV (Table 7). Comparing the overweight group with the normal weight group at the time of enrollment, eGFR was significantly lower (47.0 ± 20.8 vs 58.9 ± 30.7 mL/min/1.73 m²; P = 0.02), log Ht-TKV was significantly higher (3.02 ± 0.27 vs 2.86 ± 0.26 log mL/m; P < 0.001), and the rate of growth in TKV was higher (7.8 ± 6.0 vs 5.3 ± 5.4 %/y; P = 0.02). Furthermore, we confirmed the impact of the presence or absence of being overweight on kidney function and TKV in 4 genotypes. At the time of enrollment, PKD1 truncation in overweight patients was significantly associated with lower eGFR (41.3 ± 23.2 vs 58.8 ± 32.0 mL/min/1.73 m²; P = 0.03), higher log Ht-TKV (3.04 ± 0.23 vs 2.89 ± 0.25 log mL/m; P = 0.02), and a higher rate of growth in TKV (8.4 ± 6.6 vs 4.5 ± 5.4 %/y; P = 0.01).

Table 7.

Differences in Kidney Function and Total Kidney Volume Depending on the Presence or Absence of Being Overweight

	Number		Differences in kidney function depending on the presence or absence of being overweight									Differences in total kidney volume depending on the presence or absence of being overweight
			Renal composite endpoint, n (%)			eGFR at enrollment, mL/min/1.73 m2 mean (SD)			Kidney function deterioration rate, mL/min/1.73m2/y mean (SD)			Log Ht-TKV at enrollment, log mL/m mean (SD)			Total kidney volume growth rate, %/y mean (SD)
	OW	NW	OW	NW	P	OW	NW	P	OW	NW	P	OW	NW	P	OW	NW	P
Total	45	96	18 (40.0)	27 (28.1)	0.18	47.0 (20.8)	58.9 (30.7)	0.02	−4.78 (2.60)	−4.06 (2.99)	0.17	3.02 (0.27)	2.86 (0.26)	<0.001	7.8 (6.0)	5.3 (5.4)	0.02
PKD1 truncating	19	51	12 (63.2)	18 (35.3)	0.06	41.3 (23.2)	58.8 (32.0)	0.03	−5.86 (2.29)	−4.49 (3.27)	0.10	3.04 (0.23)	2.89 (0.25)	0.02	8.4 (6.6)	4.5 (5.4)	0.01
PKD1 non-truncating	12	26	1 (8.3)	7 (26.9)	0.39	52.8 (18.6)	54.5 (30.6)	0.86	−4.57 (1.96)	−4.09 (2.40)	0.55	3.00 (0.30)	2.87 (0.27)	0.16	9.7 (4.8)	6.9 (5.8)	0.16
PKD2 truncating	12	16	5 (41.7)	2 (12.5)	0.10	48.8 (19.4)	65.5 (29.3)	0.10	−3.88 (3.06)	−2.68 (2.94)	0.30	3.04 (0.33)	2.77 (0.30)	0.03	6.1 (5.0)	4.7 (4.1)	0.43
PKD2 non-truncating	2	3	0 (0)	0 (0)	>0.99	55.4 (16.7)	64.7 (17.0)	0.59	−1.26 (1.41)	−3.88 (1.48)	0.14	2.90 (0.26)	2.67 (0.19)	0.34	1.1 (10.4)	8.6 (5.1)	0.34

Note: Overweight was defined as a body mass index ≥25 kg/m². Values for categorical variables are given as n (%); values for continuous variables as mean (SD).

Abbreviations: eGFR, estimated glomerular filtration rate; Ht-TKV, height-adjusted total kidney volume; NW, patients with normal weight; OW, patients with overweight.

Discussion

Considering that genetic diagnosis of ADPKD is technically difficult in some cases, and because genetic abnormalities other than PKD1/PKD2 have been reported, a comprehensive genetic diagnosis is necessary. In this study, we performed whole genome analysis on patients with ADPKD. According to a previous study on whole genome analyses of ADPKD, diagnosis was possible in approximately 81% of typical cases,²⁵ and the detection rate was close to that of this study. In this study, approximately 15% of the genes were not diagnosed. Patients who cannot be diagnosed using WGS may experience the aforementioned diagnostic difficulties. ADPKD can be definitively diagnosed in numerous cases based on family history and imaging findings. In contrast, PKD1, particularly PKD1 truncation, is a poor prognostic factor for kidney disease. Therefore, a genetic diagnosis is crucial when considering tolvaptan therapy.

Several studies have examined differences in renal prognosis according to genotype, and patients with PKD1 mutations have been reported to have poorer renal prognosis than those with PKD2 mutations. This may be because of the difference in the number of cysts in patients with PKD1 and PKD2 mutations at the same age.²⁶ PKD1 is reported to have a higher frequency of somatic mutations. Therefore, patients with PKD1 mutations form more cysts, leading to a poorer renal prognosis. It has also been reported that patients with truncating mutations have a poorer renal prognosis than those with non-truncating mutations.⁸ In a mouse model, the non-truncating mutation only causes some loss of function of PKD1 but maintains the function of the product protein, polycystin.²⁷ This may be the reason why the noncleaved form is associated with a better renal prognosis than the truncated form. In this study, the annual rate of kidney function deterioration was as follows: PKD1 truncating, PKD1 non-truncating, PKD2 truncating, and PKD2 non-truncating variants. In particular, the rate of kidney function deterioration was less in the PKD2 non-truncating group, suggesting that renal prognosis may be favorable. Previous studies have reported no significant difference in the severity of ADPKD or renal prognosis owing to differences in PKD2 genotype.^9,28,29 In contrast, the PKD2 non-truncating mutation has been reported to be associated with a higher age-adjusted eGFR, which is a controversial finding.³⁰ To our knowledge, this is the first study to demonstrate that the PKD2 non-truncating variant is associated with less severe Mayo classification and better renal prognosis. Differences in renal prognosis according to the mutation type have not been fully elucidated. Kataoka et al⁹ reported that among truncating PKD1 variants, splicing and frameshift variants are associated with a poor renal prognosis, whereas nonsense variants correlate with a relatively good prognosis. The authors considered the possible involvement of nonsense mRNA-mediated decay, which degrades transcripts containing premature termination codons.^31,32

Kidney volume increases with age; however, the rate of increase in patients with ADPKD is known to exhibit significant inter-individual variability. Reports on the annual rate of increase in kidney volume in patients with ADPKD vary but have been reported to range from 2% to 5.5%.^6,11,13,33 This suggests that the study included a large number of patients with rapid growth rates or more severe ADPKD. Regarding renal decline, the progression of kidney function decline in patients with ADPKD has been reported to be particularly rapid in patients with advanced kidney function.^13,34

Previous studies have reported that hypertension and obesity are involved in the progression of ADPKD.^35,36 In this study, the proportion of patients who reached the renal composite endpoint was higher in those with hypertension than in those without hypertension, and the growth rate in TKV was higher in those who were overweight than in those with normal weight. These results suggest that hypertension and being overweight may contribute to the progression of ADPKD, such as deterioration of kidney function and an increase in TKV. In particular, hypertension and being overweight were found to have a stronger effect on disease progression than other genetic mutations in patients with PKD1 truncation. Consequently, we hypothesize that modifying these factors is essential for preventing disease progression.

This study had some limitations that warrant discussion. First, owing to the variety of PKD variants, the sample size for each variant was limited in terms of the rate of kidney function deterioration. A larger cohort for each variant may have showed differences between PKD1 truncating mutations and other variants.

Furthermore, because few patients with hypertension were treated solely with non-renin–angiotensin–aldosterone system inhibitors, the effect of these inhibitors could not be evaluated. In addition, not all patients were examined for genetic variants or clinical course. However, few studies have prospectively examined each variant and the rate of worsening kidney function, rendering this study valuable. Finally, tolvaptan was initiated after study enrollment in numerous cases, making it impossible to evaluate its effects.

Therefore, in this prospective study, we attempted a comprehensive genetic diagnosis using WGS in patients with ADPKD and found that PKD1 truncation is an important marker of renal prognosis and disease severity. Genetic diagnosis is important when considering ADPKD treatment.

Supplementary Material

Supplementary File (PDF)

Figure S1-S2; Table S1-S8