PKD1 Truncating Mutations Accelerate eGFR Decline in Autosomal Dominant Polycystic Kidney Disease Patients Open Access

Autosomal dominant polycystic kidney disease (ADPKD) is the most common monogenic disease that affects the kidneys, with a reported prevalence of around 1 in 1,000 individuals [1]. ADPKD is a systemic disease characterized by the formation and accumulation of bilateral renal fluid-filled cysts, resulting in a progressive increase in kidney volume accompanied by a gradual decrease in kidney function, ultimately leading to end-stage renal disease (ESRD) [2, 3]. ADPKD patients can also develop extrarenal manifestations including early-onset hypertension, liver and pancreatic cysts, and intracranial aneurysms [4]. ADPKD is caused by mutations in several genes, including PKD1, which is associated with approximately 78% of reported cases, and PKD2, linked to around 15% of reported cases [5]. Recently, mutations in IFT140 have emerged as the third most common cause of ADPKD as it was associated with 1–2% of reported cases [6]. Additionally, mutations in genes such as GANAB, DNAJB11, and ALG9 have been linked to other less frequent ADPKD cases [7‒9].

ADPKD exhibits significant disease severity where mutations in PKD1 were linked to an earlier onset of ESRD compared to PKD2 (54.3 years vs. 79.7 years, respectively) [1, 10]. The severity is not only affected by genic heterogeneity but also allelic heterogeneity where truncating PKD1 mutations are associated with a more severe form of the disease when compared to non-truncating (NT) PKD1 mutations [11, 12]. In addition, complex inheritance patterns such as biallelic PKD1 or in ADPKD-related genes may contribute to a more severe form of the disease [13, 14]. While it is complex, the availability of genetic information might represent a valuable prognostic insight that can help in identifying the subset of ADPKD patients with a more severe disease course [15]. This is important for initiating timely treatment, when available, and management strategies and supporting clinical trial participation. Early intervention and comprehensive care can contribute to improved outcomes and quality of life for individuals with ADPKD.

In this work, we analyzed the effect of PKD1 mutation type on kidney function with time. We followed up a cohort of ADPKD patients with confirmed genetic and clinical diagnosis for an average of 6.6 ± 3.8 years. Their kidney function was monitored and estimated glomerular filtration rates were taken during each clinic visit for each patient. We fitted generalized additive mixed model (GAMM) to analyze unbalanced eGFR longitudinal data to study the effect of PKD1 mutation types. Identifying patients who are susceptible to the severe form of the disease may aid in early intervention and better disease management to improve outcomes.

PKD1-ADPKD patients were recruited from Mubarak Al-Kabeer Hospital and Mubarak Al-Abdullah Al-Sabah Dialysis Center in Jabriya. Inclusion criteria required a confirmed clinical diagnosis of ADPKD supported by positive PKD1 genetic test results [16]. Ethical approval for the study was obtained from both the Ministry of Health (MOH) Research Ethics Committee (Reference: 1139/2019) and the Joint Committee for The Protection of Human Subjects in Research of the Health Sciences Center (HSC) at Kuwait University and the Kuwait Institute for Medical Specialization (KIMS) (Reference: VDR/JC/690). Prior to participation, written informed consent was obtained from all enrolled patients in accordance with the guidelines established by the MOH Research Committee. ADPKD patients with mutations other than PKD1 were excluded. For ADPKD patients with PKD1 truncating mutations, 9 singletons and 4 multiplex families were recruited while for patients with PKD1 NT mutations, 13 singletons and 3 multiplex families were recruited.

Clinical data for each enrolled patient was accessed and demographic information (e.g., age and gender) was obtained from medical files. Patients were followed up retrospectively and data on creatinine (μmol/L) per each visit was recorded. During each visit, we calculated eGFR (mL/min/1.73 m²) based on creatinine, sex and age at visit using the CKD-EPI formula, without race coefficient. The eGFR was calculated using the “CKDEpi.creat.rf” function from the “nephro” package in R [17]. Collected longitudinal clinical data started in April 2009 and ended in July 2022. Each patient had a different follow-up duration depending on data availability. Patients were divided into two groups depending on the PKD1 mutation type (Truncating vs. Non-Truncating). Genetic testing was performed as per our previously documented protocol [1]. ESRD is diagnosed when with a glomerular filtration rate of less than 15 mL/min. Renal imaging is not commonly employed in local clinical settings to monitor the progression of ADPKD. Instead, renal function tests are utilized for routine follow-ups due to their practicality and cost-effectiveness.

We fitted a GAMM to analyze unbalanced longitudinal data where there were multiple visits for each individual at different time intervals. The relationship between kidney functions and mutation type was assessed in a linear mixed effects model with random intercepts for each individual and adjusted for sex, age at visit, and birth year. The rate of decline (per year) in eGFR across mutation types was examined in three ways: (1) stratified models based on mutation, (2) a model with interaction term between age at visit and mutation type, and (3) a nonlinear estimation through a smooth interaction between age at visit and mutation type. In the nonlinear estimation, we used penalized splines in GAMM models that can fit smooth curves to data in non-parametric regression with a penalty term to control the smoothness. In other words, the model is penalized for wiggliness allowing maximum possible flexibility without overfitting. All models were estimated using the restricted maximum likelihood method, which produces unbiased estimation of variance components in the presence of fixed effects. The models were fitted using the “gamm” function from the “mgcv” package in R. We reported coefficients with 95% confidence intervals and a 0.05 level of significance. We used the p value from the Wald test that corresponds to the interaction term to formally test the interaction of mutation type on the yearly decline of eGFR.

To examine the relationship between PKD1 mutation type (Truncating and Non-Truncating) and the time to ESRD, a renal survival analysis was conducted. Survival times were constructed from the day of diagnosis to the onset of ESRD. Kaplan-Meier product-limit renal survival curves were generated for patients with PKD1 truncating and NT mutations using MATLAB and Statistics Toolbox, release 2012b (MathWorks, Natick, MA, USA). The mean survival time was calculated as the shortest duration in which the estimated probability of renal survival was equal to or less than 0.5. If the largest survival time was censored, the estimation was based on that value. Statistical significance of the differences between the two groups was assessed using the log-rank (Mantel-Cox) test, with a significance threshold set at p values <0.05.

A total of 42 PKD1-ADPKD patients were recruited in this study. Twenty patients had PKD1 truncating mutations (T) while 22 had PKD1 non-truncating mutations (NT). Patients were followed for an average of 6.1 ± 3.3 years. All basic demographics and clinical characteristics of the ADPKD cohort are presented in (Table 1). About 57.1% of the cohort were males, with a slightly higher percentage in the truncating group 63.6% than the non-truncating 50% (p = 0.37). For the whole cohort, at baseline, the mean age was 38.3 years, creatinine was 133 μmol/L and eGFR was 79.1 mL/min/1.73 m². At the last visit, age increased to 44.8 years, creatinine to 381.9 μmol/L and eGFR decreased to 50.2 mL/min/1.73 m². The annual changes in creatinine and eGFR are visualized in (Fig. 1). Seventy-five percent of patients with PKD1 truncating mutation had hypertension with an average age of diagnosis of 33.35 years while 70% of patients with PKD1 non-truncating mutation had hypertension with an average age of diagnosis of 35.69 years (online suppl. Table 1).

Using mixed-effects models, patients in the PKD1 truncating group exhibited higher creatinine levels (effect estimate: 184.6 μmol/L, 95% CI: 57.1, 312.1, p = 0.007) and lower eGFR levels (−19.2 mL/min/1.73 m², 95% CI: −38.9, 0.42, p = 0.063), compared to the PKD1 non-truncating group (Table 2). On an annual basis, the PKD1 truncating group experienced a significantly faster rate of creatinine increase (33.4 μmol/L per year, 95% CI: 27.3, 39.5). Those with PKD1 truncating mutations had a more rapid rate of eGFR decline (−4.7 mL/min/1.73 m² per year; 95% CI: −5.0, −4.4) compared to patients with PKD1 non-truncating mutations (−3.5 mL/min/1.73 m² per year; 95% CI: −4.0, −3.1) (p for interaction <0.001).

Figure 2 presents two theoretical scenarios of a male ADPKD patient born in 1972 (the average birth year in our sample). We use a post-estimation nonlinear prediction from the mixed-effects models to illustrate how the type of PKD1 mutation might affect his average eGFR decline over the years. In the first scenario, if the patient had a truncating mutation, the model predicts they would lose 100 mL/min/1.73 m² of kidney function by age 37. However, if the patient had a non-truncating mutation, the same loss of kidney function would occur at a slower pace, not reaching this level until age 50.

A Kaplan-Meier analysis was conducted to determine the average and median age at which ESRD occurs in patients with ADPKD and PKD1 mutations, specifically comparing patients with truncating and non-truncating PKD1 mutations (see Fig. 3). Patients with PKD1 truncating mutations exhibited a shorter duration of renal survival (median 51 years) compared to patients with non-truncating PKD1 mutations (median 56 years), and this difference was statistically significant as determined by the log-rank test (p = 0.008). The overall median age of ESRD, determined through Kaplan-Meier renal survival analysis, for patients with PKD1 mutations was 54 years (SE 0.725).

Here, we present compelling evidence demonstrating that PKD1 truncating mutations are strongly associated with a more rapid deterioration in kidney function and a shorter time to ESRD compared to ADPKD patients with non-truncating PKD1 mutations. We conducted an analysis using longitudinal clinical data obtained from ADPKD patients who were genetically diagnosed, employing generalized additive mixed-effects models. These models included random intercepts for each individual to account for repeated measurements of kidney function across different mutation types. By utilizing this approach, we were able to estimate the annual rate of decline in eGFR, as well as the rate of increase in creatinine levels in ADPKD patients, depending on the specific PKD1 mutation type. Our findings were further supported by Kaplan-Meier analysis of renal survival. Identifying ADPKD patients at risk of severe form of the disease at an early stage can enable early clinical interventions or, at the very least, prompt more aggressive monitoring strategies to improve overall patient outcomes.

ADPKD exhibits considerable variability in severity across patients, encompassing variations in the type and severity of manifestations as well as the onset of ESRD. This wide phenotypic spectrum underscores the complex molecular pathophysiology underlying the disease. Allelic heterogeneity is one factor contributing to this variability. In a study by Cornec-Le Gall et al. [11], the impact of PKD1 mutation type on ADPKD phenotype was strongly demonstrated. They reported that ADPKD patients with a truncating mutation of PKD1 had a 2.74-fold higher risk of developing ESRD compared to those with a non-truncating mutation. In our study, we employed generalized additive mixed-effects models to be able to analyze the unbalanced longitudinal data where there were multiple visits for each individual at different time intervals. This approach allowed us to estimate a significantly higher rate of decline in eGFR among patients with truncating PKD1 mutations, in comparison to patients with non-truncating mutations. Simultaneously, we estimated significantly higher rates of increase for creatinine in the PKD1 truncating mutation group (Table 2). The association of PKD1 truncating mutations with more severe form of the disease was also indicated by the earlier median age to ESRD compared to patients with PKD1 non-truncating mutations (51 years vs. 56 years) which agrees with findings from Ireland (49 years vs. 56 years) [18]. On the other hand, Cornec-Le Gall et al. [11] reported that the patients with PKD1 truncating mutations reach ESRD 12 years earlier, on average, than patients with PKD1 non-truncating mutations. While these results highlight the more severe form of the disease associated with PKD1 truncating mutations, it is essential to acknowledge that selection bias toward more severe cases cannot be entirely ruled out and results should be treated with caution.

Several previous investigations have examined the association between genotype and the phenotypic expression of ADPKD. These studies have indicated that PKD1 truncating mutations often correspond to a more severe manifestation of the disease, particularly in relation to the onset of ESRD, in comparison to PKD1 non-truncating mutations and PKD2 mutations [11, 15, 19]. However, a study conducted by Lanktree et al. [20] demonstrated that approximately 18% of ADPKD cases with PKD1 truncating mutations exhibited a milder form of the disease. This was determined through clinical and renal imaging assessments and such cases were reported to be similar to cases involving PKD2 mutations in term of severity. Taken together, these findings suggest that other modifying genetic and environmental factors may contribute to the disease phenotype.

From a statistical perspective, we deviated from methodologies previously employed to analyze the progression of eGFR decline in ADPKD [21, 22]. While Yu et al. (2019) imposed a quadratic polynomial shape on their data, we argue that this may impose a specific nonlinear curve, potentially masking the true association. Instead, we used the “penalized splines” method, which could be superior for estimating trajectories as it maximizes data fit without over- or under-fitting and does not enforce shape hypotheses. Similar to Brosnahan et al. [21], we avoided the imposition of a shape on trajectories and utilized a “Bayesian estimation method” for flexibility. However, our approach differs by presenting overall trends for all patients combined, in contrast to Brosnahan et al. [21] who focused on individual patient curves. Despite this difference, we both observed a linear decline in some individuals with truncating mutations.

A combined approach utilizing both genetic diagnosis along with renal imaging assessment could prove to be the better option to predict both functional and structural outcomes in ADPKD given the observed allelic and genetic heterogeneity and the corresponding wide phenotypic spectrum. Lavu et al. [23] demonstrated the value of such combined approach in identifying ADPKD patients at risk of rapid disease progression.

The current treatment approach for ADPKD focuses on slowing down the decline of kidney function and minimizing the associated morbidity and mortality from both renal and extrarenal complications. Therefore, it is crucial to identify ADPKD patients who are at risk of developing a more severe and progressive form of the disease at an earlier stage. This enables the timely initiation of available treatments and implementation of management strategies. Early intervention and comprehensive care play a significant role in enhancing outcomes and improving the quality of life for individuals affected by ADPKD [24, 25].

As our understanding of the genotype-phenotype relationship in ADPKD continues to improve, genetic testing has emerged as a valuable tool for identifying ADPKD patients who are at a higher risk of developing a severe form of the disease. This shall enable earlier intervention, leading to improved therapeutic outcomes [26]. However, genetic testing for PKD1 in particular, could be challenging mainly due to the existence of multiple pseudo-regions that share high sequence homology (97.8%) to exons 1–32 of PKD1 and are located proximal to the genuine gene which might complicate mutation identification [16].

This study is subject to several limitations. First, the sample size may be considered a limitation, although we enhanced statistical power through multiple visits and a lengthy follow-up period. Second, the absence of kidney volume data poses a limitation. Third, the lack of data on blood pressure control, medication use, and comorbidities is considered another constraint. Additionally, it is important to note that patient selection bias cannot be completely ruled out, and as such, the presented data should be interpreted with caution.

In conclusion, we have shown that PKD1 truncating mutations are associated with a more rapid decline in kidney functions compared to ADPKD patients with PKD1 non-truncating mutations. Genetic testing can help identify ADPKD patients at risk of severe disease at an earlier stage and hence can enable early clinical interventions and prompt more aggressive monitoring strategies to improve overall patient outcomes.

We would like to express our sincere gratitude to the Kuwait Transplant Society for their invaluable contribution to the scientific community and their unwavering commitment to enhancing patient’s outcomes.

The study was reviewed and received approval from two Ethics Committees: the MOH Research Ethics Committee (reference 1139/2019) and the Joint Committee for the Protection of Human Subjects in Research, which operates under the Health Sciences Center at Kuwait University and the Kuwait Institute for Medical Specialization (reference VDR/JC/690). All enrolled patients provided written informed consent, in adherence to the guidelines outlined by the MOH Research Ethics Committee and in compliance with the Helsinki Declaration.

The authors declare no competing interests.

The presented study was supported by funding from the Kuwait Foundation for the Advancement of Sciences (KFAS) research fund (PR17-13MM-07, awarded to H.A.) and the National Institute of Diabetes and Digestive and Kidney Diseases (grant DK058816, awarded to P.H.).

H.A. was responsible for conceptualization, data curation, funding acquisition, the investigation, project administration, and writing the manuscript. B.A. conducted the data analysis and modeling. S.R.S. performed the genotyping and data analysis. S.W. obtained the longitudinal clinical data and performed data input. Y.B. was responsible for patient recruitment and clinical analysis. M.A. and J.A. were responsible for resources and the investigation. M.Q. was involved in patient recruitment and data acquisition. F.A. was responsible for supervision and data interpretation. P.C.H. was involved in conceptualization, data analysis, the investigation, resources and review and editing of the manuscript.

The data that support the findings of this study are not publicly available due to privacy reasons but are available from the corresponding author upon reasonable request.