Patients with Protein-Truncating PKD1 Mutations and Mild ADPKD

Visual Abstract

Abstract

Background and objectives

Progression of autosomal dominant polycystic kidney disease (ADPKD) is highly variable. On average, protein-truncating PKD1 mutations are associated with the most severe kidney disease among all mutation classes. Here, we report that patients with protein-truncating PKD1 mutations may also have mild kidney disease, a finding not previously well recognized.

Design, setting, participants, & measurements

From the extended Toronto Genetic Epidemiologic Study of Polycystic Kidney Disease, 487 patients had PKD1 and PKD2 sequencing and typical ADPKD imaging patterns by magnetic resonance imaging or computed tomography. Mayo Clinic Imaging Classification on the basis of age- and height-adjusted total kidney volume was used to assess their cystic disease severity; classes 1A or 1B were used as a proxy to define mild disease. Multivariable linear regression was performed to test the effects of age, sex, and mutation classes on log-transformed height-adjusted total kidney volume and eGFR.

Results

Among 174 study patients with typical imaging patterns and protein-truncating PKD1 mutations, 32 (18%) were found to have mild disease on the basis of imaging results (i.e., Mayo Clinic Imaging class 1A–1B), with their mutations spanning the entire gene. By multivariable analyses of age, sex, and mutation class, they displayed mild disease similar to patients with PKD2 mutations and Mayo Clinic Imaging class 1A–1B. Most of these mildly affected patients with protein-truncating PKD1 mutations reported a positive family history of ADPKD in preceding generations and displayed significant intrafamilial disease variability.

Conclusions

Despite having the most severe mutation class, 18% of patients with protein-truncating PKD1 mutations had mild disease on the basis of clinical and imaging assessment.

Podcast

This article contains a podcast at https://www.asn-online.org/media/podcast/CJASN/2021_02_18_CJN11100720_final.mp3

Introduction

Autosomal dominant polycystic kidney disease (ADPKD) is the most common hereditary kidney disease worldwide, and accounts for 5%–10% of patients requiring KRT (1,2). It is typically characterized by age-dependent development of innumerable cysts with bilateral kidney enlargement, which eventually leads to advanced kidney failure in a majority of patients (2). Mutations of two genes, PKD1 and PKD2, account for 60%–78% and 15%–26% of patients where a pathogenic mutation is identified, respectively (3–5). In more recent clinical studies, mutations of additional cystic disease genes (GANAB and DNAJB11) were discovered in <1% of patients (6,7), while 10%–15% of patients had no mutation detected (3–5).

Progression of kidney disease in ADPKD is highly variable between families, in part due to a strong gene locus (PKD1 versus PKD2) effect (8). Patients with PKD1 mutations have more kidney cysts, larger kidney volume, and require KRT 16–20 years earlier than patients with PKD2 mutations (3–5,8,9). More recent studies have further delineated a strong genotype-phenotype relationship: on average, protein-truncating PKD1 mutations are associated with the most severe disease, nontruncating PKD1 mutations with intermediate disease, and PKD2 mutations with mild disease (3–5). Significant within-family kidney disease variability is well documented and suggests a deleterious modifier effect from genetic and/or environmental factors in ADPKD (10–12).

With the recent approval of tolvaptan as the first disease-modifier drug for ADPKD (13,14), identifying patients at high risk for progression who may benefit from this treatment is a clinical priority (15–19). The Consortium for Radiologic Imaging Studies of Polycystic Kidney Disease has shown that total kidney volume expands quasi-exponentially during adult life in ADPKD at approximately 5%/yr (17–19). Both the Food and Drug Administration and European Medicines Agency have accepted total kidney volume as a validated prognostic imaging biomarker for enrichment of high-risk patients for clinical trials. Using age- and height-adjusted total kidney volume, the Mayo Clinic Imaging Classification provides a useful clinical tool to predict the rate of eGFR decline (20,21). Since March 2016, we routinely performed both mutation- and imaging-based risk assessment in all patients seen at the Center for Innovative Management of Polycystic Kidney Disease (https://www.cimpkd.ca/) in Toronto. As we collated the results of their genetic and imaging results, we noticed that some patients with protein-truncating PKD1 mutations had very mild cystic disease, a finding not previously well recognized. We therefore performed and report here a systematic study to define the prevalence and clinical characteristics of patients with protein-truncating PKD1 mutations and mild disease.

Materials and Methods

Study Population

The extended Toronto Genetic Epidemiologic Study of Polycystic Kidney Disease cohort included probands with ADPKD (n=612) and their affected relatives (n=778) from a total of 612 unrelated families. Study patients were seen at the Toronto General Hospital, University Health Network (n=877) between December 1, 2006 and July 31, 2018, or St. Joseph’s Healthcare in Hamilton (n=21) between September 1, 2016 and July 31, 2018. Diagnosis of ADPKD was confirmed in all patients using ultrasound or magnetic resonance imaging–based criteria (22,23). The median follow-up of patients with kidney volume imaging was 2.1 years (Quartile 1 to Quartile 3 [Q1–Q3], 1.5 to 2.8 years). All participants provided informed consent according to a prespecified protocol approved by the Institutional Ethics Review Boards from both sites.

Exposure

All recruited probands and their available affected relatives underwent standardized clinical evaluation, which included a detailed family history of ADPKD, noting kidney disease severity, age at KRT (if applicable), and survival status, and were comprehensively screened for PKD1 and PKD2 mutations (Supplemental Material) (4,24,25). Since March 2016, 500 patients from 442 unrelated families also underwent kidney volume measurement. All patients had serum creatinine and eGFR (calculated using the CKD Epidemiology Collaboration equation) within 3 months of their kidney imaging (26).

Kidney volume was measured by magnetic resonance imaging in 489 (98%) patients and by computed tomography in 11 (2%) patients using the ellipsoid method read by a single radiologist (MP); Mayo Clinic Imaging Class was assigned using age- and height-adjusted total kidney volume (20); all patients with atypical imaging patterns (class 2) were excluded. Patients with Mayo Clinic Imaging Class 1A or 1B were classified as “low risk” and 1C, 1D, or 1E were classified as “high risk” for the need of KRT, respectively.

Outcomes

Using low-risk Mayo Clinic Imaging Class 1A or 1B as a proxy for mild disease, we defined the prevalence and clinical characteristics of mild disease in 489 patients with protein-truncating PKD1 mutations. For patients with protein-truncating PKD1 mutations who did not have kidney imaging available, we used age at KRT >65 years as a proxy for mild ADPKD (27). To identify affected relatives with concordant or discordant disease to define the prevalence of within-family disease variability, we further used Mayo Clinic Imaging Class 1C–1E or age at KRT <55 years as proxies for severe disease, as previously reported (27).

Statistical Analyses

Continuous variables were reported as mean and 95% confidence interval (95% CI) if normally distributed, and medians and Q1–Q3 if not. Discrete variables were reported as percentages. Groups were compared with a t test; Wilcoxon or chi-squared tests were used when appropriate. Multivariable linear regression was performed to test the effects of age, sex, and mutation classes on log-transformed height-adjusted total kidney volume and eGFR using R (version 3.3.3, R Foundation for Statistical Computing). The main relationships were first tested, followed by tests for interaction of age, sex, and mutation classes. The beta of log₁₀ transformed height-adjusted total kidney volume from the regression model was back transformed by exponentiation to obtain the percent difference per change in predictor variable.

Results

Characteristics of Study Cohort

From the extended Toronto Genetic Epidemiologic Study of Polycystic Kidney Disease (n=1390), 500 (36%) patients had both genetic testing and kidney volume measurement. After reviewing their imaging patterns, 13 patients with atypical ADPKD imaging patterns were excluded. Of the remaining 487 patients with typical imaging patterns, 190 (38%) had Mayo Clinic Imaging Class 1A–1B and 297 (59%) had Class 1C–1E; none of them required KRT at their last follow-up. Of 890 patients without kidney volume measurement, 602 (68%) had died or required KRT at their last follow-up.

A total of 570 (41%) patients in the study cohort had protein-truncating PKD1 mutations; 174 had kidney volume measurement and 34 (20%) had Mayo Clinic Imaging Class 1A–1B (Figure 1). Among 396 patients with protein-truncating PKD1 mutations but no kidney volume measurement, 290 required KRT at their last follow-up. The age distribution of patients with protein-truncating PKD1 mutations requiring (n=290) and not requiring (n=280) KRT at their last follow-up is shown in Figure 2. Of interest, 26 out of 290 (9%) patients with protein-truncating PKD1 mutations did not require KRT until age >65 years, and 15 out of 280 (5%) had CKD stages 2–4 at age ≥60 years (Figure 2B); all of these patients had mild disease similar to patients with PKD2 mutations.

Figure 1.

Study flow diagram. Patients with protein-truncating PKD1 mutations in the extended Toronto Genetic Epidemiology Study of Polycystic Kidney Disease with and without Mayo Clinic Imaging Class. PT, protein truncating; TKV, total kidney volume.

Figure 2.

Frequency plots of age distribution of patients with protein-truncating PKD1 mutations in the Toronto Genetic Epidemiology of Polycystic Kidney Disease Cohort (n=570). (A) Patients who required KRT (n=290). (B) Patients who did not require KRT at the time of death (n=25) or last follow-up (n=255).

Patients with Protein-Truncating PKD1 Mutations and Mild Disease

Of 34 patients from 29 unrelated families with protein-truncating PKD1 mutations and Mayo Clinic Imaging Class 1A–1B, two had advanced kidney failure (eGFR <45 ml/min per 1.73 m²) that was discordant from their mild cystic disease burden by imaging: one was a 59-year-old male with longstanding history of type 2 diabetes mellitus and the other, a 44-year-old female with a history of poorly controlled hypertension; both were excluded from further analysis. The clinical characteristics of the remaining 32 patients with confirmed mild disease are shown in Table 1; 23 (72%) were >30 years of age. Representative magnetic resonance images comparing age-matched cystic disease severity in patients with different mutation types and Mayo Clinic Imaging Class are shown in Supplemental Figure 1. The familial mutations of these patients include 16 frameshift insertions or deletions, 11 nonsense mutations, and two canonical splice site (c.11538–2A>C, p.R3846fs, reported in Mayo PKD database; c.1385+1G>C, p.R462fs10×, novel) mutations; they spanned the entire gene with no preferential location or accumulation in the 3′ exons (Supplemental Figure 2). The relative frequency of mutations in the duplicated versus nonduplicated regions of PKD1 was not different between patients with Mayo Clinic Imaging Class 1A–1B and 1C–1E (P>0.5).

Table 1.

Clinical characteristics of patients by mutation type and Mayo Clinic Imaging Class

Patient Characteristics	Protein-Truncating PKD1 Mutation (Mild Imaging)a (n=32)	Protein-Truncating PKD1 Mutation (Severe Imaging)b (n=140)	PKD2 Mutation (Mild Imaging)c (n=66)
Number of families (%)	29	132	61
Age at time of MRI, yr, mean (95% CI)	36 (32 to 41)	38 (36 to 40)	47 (43 to 50)d
Male, n (%)	10 (31)	57 (41)	20 (30)
Body mass index, kg/m2, median (Q1–Q3)	23.1 (21.5 to 24.9)	25.4 (22.6 to 28.7)d	24.5 (22.0 to 29.5)
Ethnicity (%)
European	66	78	76
Asian	25	15	18
Black	6	2	4.5
Other	3	5	1.5
Ht-TKV, ml/m, mean (95% CI)	343 (273 to 405)	923 (662 to 1331)d	342 (249 to 556)
eGFR, ml/min per 1.73 m2, mean (95% CI)	95 (85 to 110)	75 (48 to 103)d	90 (68 to 105)
CKD stage (%)
1–2	94	65	79
3	6	25	18
4–5	0	10	3

MRI, magnetic resonance imaging; 95% CI, 95% confidence interval; Ht-TKV, height-adjusted total kidney volume.

^aProtein-truncating PKD1 mutations with Mayo Clinic Imaging Class 1A–1B (reference group).

^bProtein-truncating PKD1 mutations with Mayo Clinic Imaging Class 1C–1E.

^cPKD2 mutations with Mayo Clinic Imaging Class 1A–1B.

^dP<0.05 for pairwise comparisons to the reference.

Comparison of Disease Severity by Mutation Types and Imaging Class

Comparing patients with mild disease by imaging (Mayo Clinic Imaging Class 1A–1B), we did not find any difference in eGFR (P=0.18) between those with protein-truncating PKD1 versus PKD2 mutations; however, patients with PKD2 mutations were older than those with protein-truncating PKD1 mutations (47; 95% CI, 43 to 50 years versus 36; 95% CI, 32 to 41 years, P=0.002) (Table 1). The slopes of log-transformed height-adjusted total kidney volume (β_{PKD1&MCIC1A-B}=1.0% larger kidney per year; 95% CI, 0.8 to 1.3 versus β_{PKD2&MCIC1A-B}=1.3% larger per year; 95% CI, 1.1 to 1.5, P=0.3) and eGFR (β_{PKD1&MCIC1A-B} =−1.3 ml/min per 1.73 m² per year; 95% CI, −1.7 to −0.9 versus β_{PKD2&MCIC1A-B} =−1.5 ml/min per 1.73 m² per year; 95% CI, −1.8 to −1.2, P=0.6) overlap in patients with Mayo Clinic Imaging Class 1A–1B regardless of whether they had protein-truncating PKD1 or PKD2 mutations (Figure 3). As expected, patients with both protein-truncating PKD1 mutations and Mayo Clinic Imaging Class 1C–1E had distinctly more severe disease compared with either of the former groups.

Figure 3.

Disease severity is similar in patients with Mayo Clinic Imaging Class (MCIC) 1A–1B regardless of whether they had protein-truncating PKD1 (n=32) or PKD2 (n=66) mutations. Plots of log-transformed height-adjusted total kidney volume (A) and eGFR (B) versus age. Patients with protein-truncating PKD1 and MCIC 1C–1E (n=140) had more severe disease than either of the former groups.

By multivariable analysis, age, male sex, and Mayo Clinic Imaging Class 1C–1E were independently associated with larger kidney volume and reduced eGFR. Comparing patients with Mayo Clinic Imaging Class 1A–1B after adjusting for age and sex, patients with PKD2 mutations had marginally smaller kidney volumes (β=−7.7%; 95% CI: −13.9% to −0.01%; P=0.02) and their eGFR was not significantly different (β=9.0 ml/min per 1.73 m²; 95% CI: −0.4 to 18.3 ml/min per 1.73 m², P=0.06) compared with those with protein-truncating PKD1 mutations (Table 2). There was no evidence of interaction between age and sex, age and mutation class, or sex and mutation class for either kidney size or eGFR (P>0.1, data not shown).

Table 2.

Predictors of kidney size and eGFR from multivariable regression analysis

				Interpretation of Results
Predictors	B	95% Confidence Interval	P	Percentage Difference	ml/min/1.73m2
log10-transformed height-adjusted total kidney volume
Age (per yr)	0.013	0.011 to 0.015	<0.001
Back transformeda	1.013	1.011 to 1.015		1.3 larger
Female (versus male)	–0.08	−0.12 to−0.04	<0.001
Back transformed	0.923	0.886 to 0.961		7.7 smaller
PT PKD1 mutations (severe) versus (mild)	0.41	0.35 to 0.47	<0.001
Back transformed	1.49	1.419 to 1.600		49 larger
PKD21A-1B versus PT PKD1 mutations (mild)	−0.08	−0.15 to−0.01	0.02
Back transformed	0.923	0.861 to 0.990		7.7 smaller
eGFR
Age (per yr)	−1.7	−1.9 to −1.5	<0.001		1.7 lower
Female (versus male)	6.4	0.7 to 12.1	0.02		6.4 higher
PT PKD1 mutations (severe) versus (mild)	−15.9	−24.1 to −7.6	<0.001		15.9 lower
PKD2 1A-1B versus PT PKD1 mutations (mild)	9.0	−0.4 to 18.3	<0.06		9.0 higher

PT PKD1 mutations (mild): protein-truncating PKD1 mutations with Mayo Clinic Imaging Class 1A–1B.

PT PKD1 mutations (severe): protein-truncating PKD1 mutations with Mayo Clinic Imaging Class 1C–1E.

^aPercent difference obtained by exponentiation of the beta of log₁₀ height-adjusted total kidney volume.

Within-Family Disease Variability in Patients with Protein-Truncating PKD1 Mutations and Mild Disease

A positive family history of ADPKD was confirmed in 23 of 29 (79%) probands with protein-truncating PKD1 mutations and mild disease by imaging. Of the remaining probands, three had no apparent family history and three had unknown or nonconfirmable family history. Within-family disease variability (on the basis of Mayo Clinic Imaging Class and/or age at KRT) of the affected relatives in these 23 families is shown in Figure 4 and Table 3. Of six families with Mayo Clinic Imaging Class available in at least two affected relatives, four (TOR094, TOR139, TOR173, and TOR198) demonstrated concordant disease, whereas two displayed discordant disease spanning at least two classes (TOR17 and TOR308). Patient TOR308.2 was found to have somatic mosaicism (PKD1: c.2605delC; p.R869FS28×; mutant allele fraction of approximately 10% in blood sample) and germline disease transmission to her daughter, as previously reported (24). Except for TOR308, all families shown in Table 3 had an older affected relative, which excludes younger generations as potential genetic mosaics.

Figure 4.

Within-family disease variability in families with PKD1-truncating mutations and mild disease by imaging. Each vertical line represents one family, and each dot represents an affected relative. Mild disease defined by Mayo Clinic Imaging Class 1A–1B are denoted by light blue open circles; severe disease defined by Mayo Clinic Imaging Class 1C–1E are denoted by dark blue open circles. Individuals requiring KRT with mild (i.e., >65 years) or severe disease (i.e., <55 years) are denoted by a light or dark blue solid circle, respectively. Individuals whose kidney imaging (black open circles) or age at KRT (solid black circle) did not allow assignment of mild or severe disease are deemed uninformative. A total of 16 families had at least one discordant (D) relative pair, and six families had at least one concordant (C) relative pair. The assessment of disease variability was indeterminate (I) in six families.

Table 3.

Within-family disease variability in affected relatives of patients with protein-truncating PKD1 mutations and mild kidney disease by imaging

Patient ID	Age (yr)	htTKV (ml/m)	Mayo Clinic Imaging Class	eGFR (ml/min per 1.73 m2)	No. of Affected Relatives Requiring KRT
					Before Age 55	After Age 65
Disease discordance by imaging in one or more relative pair
TOR308.2	57	927	1B	65	0	0
TOR308.1	25	637	1E	97
TOR17.1	51	623	1B	48	4	0
TOR17.13	28	655	1D	102
TOR17.12	23	712	1E	133
Disease concordance by imaging in one or more relative pair
TOR094.1	22	264	1B	93	2	0
TOR094.2	19	211	1B	91
TOR139.1	20	183	1A	127	3	0
TOR139.2	22	175	1A	125
TOR198.3	39	291	1B	105	1	0
TOR198.4	35	316	1B	93
TOR173.1	54	587	1B	86	0	2
TOR173.2	28	493	1C	115
Disease discordance by imaging and age at KRT in one or more relative pair
TOR094.1	22	264	1B	93	2	0
TOR139.1	22	183	1B	127	3	0
TOR198.3	39	291	1B	105	1	0
TOR420.1	39	275	1B	109	1	0
TOR431.1	25	254	1B	102	2	0
TOR438.1	25	296	1B	122	2	0
TOR48.9	35	352	1B	96	3	0
TOR595.1	32	344	1B	85	1	0
TOR663.1	39	324	1B	85	1	0
TOR833.1	36	221	1A	90	1	0
TOR841.1	18	237	1B	113	1	0
TOR845.1	34	378	1B	83	1	0
TOR968.1	59	371	1B	64	1	0
Disease concordance by imaging and age at KRT in one or more relative pair
TOR619.2	43	342	1B	79	0	1
Uninformative for assessment of within-family disease variability
TOR242.2	31	326	1B	119	0	0
TOR291.3	46	589	1B	105	0	0
TOR388.1	46	463	1B	103	0	0
TOR674.1	33	348	1B	115	0	0
TOR803.1	35	389	1B	93	0	0
TOR895.1	28	321	1B	120	0	0

htTKV, height-adjusted total kidney volume.

In older patients with more advanced disease and without kidney imaging, we used age at KRT <55 years to define severe ADPKD (27). Using this criterion, disease discordance was noted in 16 of 18 (89%) families with a protein-truncating PKD1 mutation and mild disease by imaging in the proband. Similarly, using age at KRT >65 years or Mayo Clinic Imaging Class 1A–1B as a proxy for mild disease (27), six of 18 (33%) families displayed disease concordance in at least one affected relative pair (Figure 4).

Discussion

The Consortium for Radiologic Imaging Studies of Polycystic Kidney Disease has shown that kidney volume expands on average 5%/yr during adult life in ADPKD and is a sensitive imaging biomarker for predicting kidney-disease progression (17–19). Using age- and height-adjusted total kidney volume, the Mayo Clinic Imaging Classification provides a useful clinical tool for assessing disease severity in ADPKD (20,21). By integrating the results of both mutation- and imaging-based risk assessment, we defined the severity of our patients over a wide age range, including younger patients in whom eGFR poorly discriminates mild from severe disease. In a large and clinically well-characterized cohort, we found 18% (32/174) of patients with protein-truncating PKD1 mutations displayed mild disease by their kidney imaging pattern and eGFR; their mild disease was similar to patients with both PKD2 mutations and the same imaging classes. Among the older patients with protein-truncating PKD1 mutations without kidney volume measurements, we also found evidence for mild disease in 9% (26 out of 290) of patients who required KRT after age 65 years and in 5% (15 out of 280) of patients who did not require KRT at age ≥60 years. Taken together, these data indicate that mild disease described in the above patients is an extreme phenotype associated with protein-truncating PKD1 mutations.

Significant within-family kidney disease variability in ADPKD is well documented and implicates a modifier effect from genetic and environmental factors (10–12). Of the genetic modifiers identified to date, although uncommon, coinheritance of a second rare mutation in a cystic disease gene, such as PKD1, PKD2, or HNF1B, increased the severity of a PKD1 mutation (28–33), while coinheritance of a COL4A1 mutation increased the severity of a PKD2 mutation (34). Additionally, common genetic variants may increase the severity of ADPKD, as exemplified by the variant association observed with DKK3 (35). By comparison, concurrent medical conditions that are risk factors for CKD progression including diabetes, hypertension, and obesity; complications of ADPKD including kidney stone disease and cyst infections; and environmental and dietary factors including smoking and salt, carbohydrate, and protein intake may also increase the severity of ADPKD (36). Notably, all these modifiers are expected to be deleterious by increasing cystic disease severity.

In this study, we observed significant within-family disease variability in our patients with protein-truncating PKD1 mutations and mild disease. In this regard, our documentation of patient TOR308.2 as a somatic mosaic provides an explanation for her mild disease (24). Mosaicism refers to the occurrence of two genetically distinct cell populations within an individual resulting from a somatic mutation during embryogenesis (37). Due to dilution and variable involvement of the affected cells, a mosaic individual with ADPKD typically presents with de novo ADPKD with atypical imaging (i.e., focal, unilateral, or asymmetric) patterns. However, the presence of a positive parental family history of ADPKD in preceding generations in 84% of our patients and a typical ADPKD imaging pattern makes somatic mosaicism an unlikely explanation for the mild disease observed in most of our patients with protein-truncating PKD1 mutations. Overall, 72% of our patients with both protein-truncating PKD1 mutations and Mayo Clinic Imaging Class 1A–1B were older than 30 years of age when misclassification of mild disease is less likely; however, in the remaining younger patients, potential misclassification is a limitation of our study. Taken together, our findings strongly suggest the existence of a protective modifier effect in ADPKD from genetic and/or environmental factors.

Protective genetic factors have been identified in multiple pathologic conditions including HIV-1 infection, hypercholesterolemia, Alzheimer’s disease, nonalcoholic steatohepatitis, and various autoimmune disorders (38–42). For example, homozygous deletions (i.e., delta 32) of the HIV-1 coreceptor CCR5, present in approximately 1% of European population, protect against HIV infection (39); heterozygous loss of function (i.e., p.142×, p.679×) variants of proprotein convertase subtilisin/kexin type 9 (PCSK9), present in approximately 2% of Black individuals, lower LDL cholesterol and reduce the risk of coronary heart disease (40,41); and a heterozygous loss of function (i.e., rs72613567:TA) variant of the hepatic lipid droplet protein hydroxysteroid 17-beta dehydrogenase (HSD17B13), present in approximately 25% of European population, protects against nonalcoholic steatohepatitis (42). These studies provided important insight into disease pathobiology, advanced biomarker discovery, and identified novel targets for therapeutic development.

Mutations of multiple genes encoding for ciliary proteins cause cystic disease, suggesting a central role of the primary cilia in the pathogenesis of cystic kidney diseases (43). Of interest, a study of murine polycystic kidney disease models showed that ablation of primary cilia by inactivation of a ciliary gene (Kif3 or Ift20) resulted in milder disease, whereas inactivation of either Pkd1 or Pkd2 resulted in severe disease. Unexpectedly, ablation of the primary cilia together with inactivation of Pkd1 or Pkd2 in these models resulted in attenuated kidney disease compared with Pkd1 or Pkd2 inactivation, respectively (44). These data suggest that an unidentified cilia-based signaling pathway interacts with the polycystin complex to modulate cyst disease severity; loss of cilia function is protective in the context of ADPKD. Screening of our patients with protein-truncating PKD1 mutations and mild disease for loss-of-function ciliary gene mutations by next-generation sequencing may provide a promising approach to identify protective genetic modifiers.

Metabolic reprogramming and tissue inflammation are important pathogenic mechanisms that mediate the progression of experimental cystic kidney disease (45,46). By modulating the AMP-activated protein kinase–mammalian target of rapamycin signaling pathway, caloric restriction and intermittent fasting have been shown to attenuate the severity of experimental cystic kidney disease (47,48). Likewise, the gut microbiome has been shown to modulate the metabolism and tissue inflammation (49,50). The absence of systematic collection of metabolic comorbidities and dietary intake is a limitation of this study. Therefore, comparative studies of the dietary patterns and gut microbiomes in our discordant affected relative pairs with mild and severe cystic disease may provide another promising approach to unravel the potential environmental protective factors for ADPKD.

In conclusion, we found 18% of patients with protein-truncating PKD1 mutations displayed mild disease, on the basis of their kidney imaging pattern and eGFR. Similarly, among older patients with protein-truncating PKD1 mutations without kidney volume measurements, we also found mild disease in 9% who required KRT >65 years and 5% who did not require KRT at age ≥60 years. Taken together, these data indicate that mild kidney disease described here is not a rare phenotype associated with protein-truncating PKD1 mutation. On average, protein-truncating PKD1 mutations are associated with severe ADPKD; however, our data indicate that mutation class alone cannot be used to predict disease severity in individual patients with complete certainty. The presence of significant within-family disease variability in a majority of these mildly affected patients with protein-truncating PKD1 mutations strongly suggests a protective modifier effect from unidentified genetic and/or environmental factors.

Disclosures

A. Paterson reports employment with the Hospital for Sick Children. E. Guiard reports employment with Toronto General Hospital University Health Network. I.-A. Iliuta reports employment with Toronto General Hospital, University Health Network. M. Lanktree reports employment with St. Joseph Healthcare Hamilton and McMaster University; is a young investigator in the KRESCENT Program, a national kidney research training partnership of the Kidney Foundation of Canada, the Canadian Society of Nephrology, and the Canadian Institutes of Health Research; reports receiving grant funding from Canadian Institutes of Health Research, Hamilton Health Sciences, and Hamilton Academic Health Sciences Organization, and received compensation for participation as a speaker and an advisory board member with Otsuka Pharmaceuticals. S. Ahmed reports employment with Toronto General Hospital. Y. Pei reports employment with University Health Network and University of Toronto; consultancy agreements with Otsuka, Sanofi, and Vertex; receiving honoraria from Otsuka, Sanofi, and Vertex; receiving compensation for participation in advisory boards for Otsuka, Reata Pharmaceuticals, and Sanofi-Genzyme; and serving on the steering committee for a randomized controlled trial in ADPKD (Sanofi/Genzyme). All remaining authors have nothing to disclose.

Funding

This work was in part supported by Canadian Institutes of Health Research (CIHR) project grant PJT-376307 and CIHR Strategy for Patient Oriented Research program grant in CKD (CAN-Solve-CKD SCA-145103; to Y. Pei). M. Lanktree is a new investigator in the Kidney Research Scientist Core Education and National Training program funded by the CIHR Kidney Foundation of Canada and the Canadian Society of Nephrology.

Supplemental Material

This article contains the following supplemental material online at http://cjasn.asnjournals.org/lookup/suppl/doi:10.2215/CJN.11100720/-/DCSupplemental.

Supplemental Material. Mutation screening protocol for study cohort.

Supplemental Figure 1. Representative axial (left) and coronal (right) magnetic resonance images comparing age-matched cystic disease severity by Mayo class in patients with different mutation types.

Supplemental Figure 2. Location of PKD1-truncating mutations and mild Mayo Imaging Class.