A Combination Approach to Improving Prognostication in Autosomal Dominant Polycystic Kidney Disease

Autosomal dominant polycystic kidney disease (ADPKD) is the most prevalent monogenic cause of ESKD. The vasopressin receptor antagonist tolvaptan slows cyst growth and kidney function decline and was approved by the US Food and Drug Administration for use in patients considered to be at high risk of disease progression. There is a need for effective risk stratification methods to guide clinical management and prevent complications.^1,2 On average, total kidney volume (TKV) enlarges by 3.4%–12.6% annually, with larger kidneys being at a higher risk of disease progression.¹ Therefore, TKV was formally qualified by the US Food and Drug Administration as a prognostic enrichment biomarker for selecting patients at high risk of a progressive decline in kidney function for inclusion in interventional clinical trials. Subsequently, the Mayo Clinic Imaging Classification (MCIC) was developed using height-adjusted TKV (htTKV) and age as predictors and is the most used tool to identify high-risk progression.² Genetic mutation in the polycystin 1 (PKD1) and polycystin 2 (PKD2) genes account for nearly 75%–80% of all patients with ADPKD with PKD1 mutations leading to ESKD nearly 20 years earlier than PKD2 mutations. Further refining the genetic information on the basis of whether the PKD1 mutation is a truncating or nontruncating mutation and combining this with age, sex, and specific clinical events has led to the development of the predicting renal outcomes in ADPKD (PROPKD) risk score.³

Both PROPKD and MCIC scores have been widely used in predicting the risk of progression; however, it is not clear whether the predictive power would enhance with combining these scores. Considering having an available treatment option (tolvaptan), there is still a need for providing an optimal predictive scoring system for identifying high-risk patients. A recent paper by Wolff et al.⁴ presents a novel approach to identifying patients at risk of rapid disease progression by integrating both the MCIC and PROPKD scores. The authors conducted a comprehensive retrospective multicenter cohort study involving 468 patients diagnosed with ADPKD across four tertiary centers in Europe from 2016 to 2023, excluding patients with atypical PKD because they are not risk stratified using MCIC. Patients were categorized into two groups—rapid progression and slow progression—on the basis of eGFR decline per year and PROPKD and MCIC scores. Patients with an annual eGFR decline of ≥3 ml/min per 1.73 m² for at least 2 years and a PROPKD score of 7–9 in addition to MCIC class of 1D–1E were categorized as the rapid progression group. The authors proposed a refinement to the PROPKD scoring system by incorporating more genetic information through the Rare Exome Variant Ensemble Learner (REVEL) analysis. REVEL is used to predict the pathogenicity of missense variants and can be used to prioritize the most likely clinically relevant variants related to the disease.⁵ It is not as reliable for other types of mutations, such as nonsense, frameshift, or splice site variants, which may often have more obvious effects on protein function. A REVEL score higher than 0.75 was categorized as pathogenic. The primary end points were defined as the development of advanced CKD (aCKD) stages 4 or 5, the need for dialysis, or an eGFR of <15. Linear and logistic regression models were used to evaluate the significance of covariates in predicting eGFR decline and htTKV. Kaplan–Meier survival analysis was used to compare the primary end points.

The study demonstrated several interesting findings that advance our understanding of risk prediction in ADPKD. There was a low concordance between the MCIC and PROPKD scores, particularly among patients classified as intermediate risk which was driven by MCIC 1C and PKD1 nontruncating variants. The combined use of MCIC and PROPKD scores showed increased specificity but lower sensitivity for detecting rapid progressors. The area under the curve for MCIC and PROPKD for predicting aCKD was modest, and combining the two did not significantly improve the area under the curve compared with either alone. The use of REVEL did not enhance the ability of PROPKD in predicting rapid progression.

The use of combination risk scores has limited value at the extremes of each risk score (low or high risk by MCIC or PROPKD) where management decisions are somewhat straightforward. However, there remains a large segment of the population with intermediate risk on the basis of either risk score alone or with discordant risk scores, with one being intermediate to high risk. For instance, in the current study, more than 40% of patients identified as rapid progressors by either MCIC or PROPKD score were not classified as such by the other. Combining scores was able to demonstrate a 7-year difference in onset of aCKD between intermediate–low and intermediate–high-risk groups, and when using a combination of both scores, approximately 38% of intermediate-risk (MCIC or PROPKD) patients were classified as intermediate–high risk, potentially making them candidates for disease-modifying therapy with tolvaptan. Of note, only 60% of the intermediate–high-risk group were receiving tolvaptan at the time of the study indicating that reclassification can change the current practice profoundly because it can lead to more personalized and potentially aggressive treatment strategies and delaying ESKD in a new subgroup of patients.

Two prior studies have combined genotype and imaging data to predict the outcomes in ADPKD. Using the data from the Consortium for Radiologic Imaging Studies of Polycystic Kidney Disease study and the Halt Progression of Polycystic Kidney Disease trial, Lavu et al. analyzed the effects of eGFR and htTKV in patients with PKD1 and PKD2 mutations and found that combining the genotype and imaging data enhanced predictive power and highlighted the strong effect of PKD1 truncating mutations on the onset of ESKD.⁶ The second study evaluated mean kidney length and genotype in combination and demonstrated increased the sensitivity and specificity for identifying patients with rapid disease progression defined as a loss of eGFR >2.5 ml/yr for 5 years.⁷ While these studies focused on the ability of genetic and imaging factors to predict ESKD, they do not highlight discordance in risk with imaging and genetic information and how this group can benefit from combining risk variables, something the current study by Wolf et al. does. Although MCIC and genotype remain the most used risk predictors, other emerging biomarkers such as imaging texture analysis, total cyst number/volume, cyst–parenchyma surface,⁸ and plasma and urinary biomarkers⁹ could further refine our ability to predict disease trajectory alone or in combination with MCIC and genotype.¹⁰

The paper by Wolff et al.⁴ has several strengths, including its multicenter design, a relatively large cohort of patients, and the ability to subclassify an adequate number of patients on the basis of the different risk scores to draw meaningful conclusions. There remain several limitations though that must be considered when interpreting the results, because the study was performed across four centers in Europe and the study's findings may not be generalizable to a more diverse population. Moreover, the lack of external validation and limited diversity in the patient population present significant limitations to this risk stratification model. Future studies with larger cohorts are essential to confirm the discrepancies observed between the MCIC and PROPKD scoring systems for intermediate-risk patients because only 37 patients had a combination of a low MCIC and high PROPKD score in the current study.

In conclusion, combining the MCIC and PROPKD scores has the potential to refine risk stratification, particularly among patients with intermediate- and discordant-risk scores. This improvement could directly and positively affect patient care, leading to earlier interventions, and may also serve to enrich clinical trial populations with more high-risk patients. Further multicenter studies are needed to validate these findings across diverse populations, ensuring that risk stratification models can be effectively applied in various clinical settings. As the field advances toward personalized medicine, ongoing research and validation efforts will be fundamental to refine these tools and improve outcomes for patients with ADPKD, aiming to prevent ESKD.

Supplementary Material

cjasn-20-323-s001.pdf ^{(1.3MB, pdf)}

Disclosures

Disclosure forms, as provided by each author, are available with the online version of the article at http://links.lww.com/CJN/C158.

Funding

P.S. Garimella: Division of Diabetes, Endocrinology, and Metabolic Diseases (R01 DK139291).

Author Contributions

Conceptualization: Pranav S. Garimella, Sayna Norouzi.

Writing – original draft: Pranav S. Garimella, Sayna Norouzi.

Writing – review & editing: Pranav S. Garimella, Sayna Norouzi.