Unmet needs and challenges for follow-up and treatment of autosomal dominant polycystic kidney disease: the paediatric perspective

Abstract

Awareness is growing that the clinical course of autosomal dominant polycystic kidney disease (ADPKD) already begins in childhood, with a broad range of both symptomatic and asymptomatic features. Knowing that parenchymal destruction with cyst formation and growth starts early in life, it seems reasonable to assume that early intervention may yield the best chances for preserving renal outcome. Interventions may involve lifestyle modifications, hypertension control and the use of disease-modifying treatments once these become available for the paediatric population with an acceptable risk and side-effect profile. Until then, screening of at-risk children is controversial and not generally recommended since this might cause psychosocial and financial harm. Also, the clinical and research communities are facing important questions as to the nature of potential interventions and their optimal indications and timing. Indeed, challenges include the identification and validation of indicators, both measuring and predicting disease progression from childhood, and the discrimination of slow from rapid progressors in the paediatric population. This discrimination will improve both the cost-effectiveness and benefit-to-risk ratio of therapies. Furthermore, we will need to define outcome measures, and to evaluate the possibility of a potential therapeutic window of opportunity in childhood. The recently established international register ADPedKD will help in elucidating these questions. In this review, we provide an overview of the current knowledge on paediatric ADPKD as a future therapeutic target population and its unmet challenges.

CHILDHOOD ADPKD

Historically, autosomal dominant polycystic kidney disease (ADPKD), the most common hereditary kidney disease, was named adult polycystic kidney disease. Nowadays, awareness is growing that the clinical course of ADPKD already begins in childhood in some patients, although in a variable way. Indeed, the disease comprises a wide phenotypic spectrum, from severely affected neonates to clinically asymptomatic affected adults [1] (Figure 1). Even within families, with patients suffering from the same PKD1 or PKD2 mutation, high phenotypic variability exists, for which the underlying cause(s) are uncertain [2]. Presumably, disease severity correlates with the level of the functional polycystin (PC) complex, which is formed by both PKD gene products, PC1 and PC2, and is modified by other factors [3] such as glucosidase II subunit and DNAJB11, encoded by GANAB and DNAJB11, respectively, and involved in the maturation and trafficking of PC1 [4, 5]. Regarding quantitative defects in PC functionality, two hypotheses have been put forward in the past: first, the somatic second hit model in which the disease, with a dominant inheritance, has the features of a recessive disease on a cellular level [6]; and secondly, the more generally accepted threshold hypothesis, which states that cyst growth can occur if the level of functional PC falls below the cystogenic threshold. These models do not necessarily exclude each other, as the second hit might be one of the factors influencing the functional PC level, together with other ‘third hits’. The latter include low birth weight, urinary tract infections and acute renal injury. Moreover, given the cellular and molecular similarities between renal repair and ADPKD, cystogenesis might be seen as a futile attempt to repair a non-existant renal injury [7].

ADPKD may lead to both renal and extrarenal manifestations, which are also reported in children: urinary concentration defects, hypertension, left ventricular hypertrophy (LVH), microalbuminuria, proteinuria and haematuria [8–13]. The literature on the clinical presentation of childhood ADPKD has recently been summarized [14]. Some of the features may cause symptoms, while others remain silent and undetected unless actively screened for. Related to this, ADPKD can be diagnosed in children following the investigation of symptoms, but the diagnosis may also be the consequence of coincidental findings or active family screening in asymptomatic children (Figure 2). Apart from the phenotypic variability in ADPKD patients per se, this broad diagnostic spectrum is a second factor contributing to the heterogeneity of the data on the paediatric ADPKD population. The diversity of recommendations regarding childhood screening between different expert centres contributes to this diagnostic spectrum, together with the informed opinion of the children’s parents. Also, for individuals <15 years of age, there are no validated diagnostic imaging criteria available [15, 16]. Given the rarity of childhood simple renal cysts [17], children with a family history of ADPKD, having a single cyst, preferably over 1 cm, are assumed to have ADPKD, although this criterion was never fully evaluated [14].

RATIONALE FOR EARLY TREATMENT

Fetal cyst formation is an important ADPKD characteristic. Indeed, sonographically detected renal cysts in newborns have had a very high growth rate in utero, under the influence of stochastic, intrinsic and humoral factors. Thereafter, cystic expansion proceeds at much lower growth rates [18]. Also, mathematical models have demonstrated that those cysts that started early in life are the main contributors to the total cyst volume and renal size [19]. Moreover, it is known that ADPKD cysts derive from only 1–3% of the 2 million nephrons in both kidneys [20]. The probability that an epithelial kidney cell will form a cyst is not only dependent on its PC level, but also on its biological context, as taken into account in the so-called ‘cystic probability landscape’ [21]. This small fraction of cyst-forming cells will nonetheless be able to destroy most of the renal parenchyma, leading to end-stage kidney disease (ESKD) in 50% of ADPKD patients in their sixth decade of life, since cysts can block glomerular filtration by upstream nephrons [22]. Additionally, initial cysts are the principal trigger for a snowball effect driving the formation of new cysts, leading to disease progression, including inflammation and fibrosis [23], towards ESKD [24]. Taking together the evidence for early parenchymal destruction and the substantial prevalence of symptoms in children [14, 25], the best chance for preserving long-term renal function may be given by an early therapeutic start [26].

Whereas early intervention is increasingly considered in order to attenuate both long-term renal and cardiovascular complications [27], only two interventional trials have been performed to date in children with ADPKD [28, 29], and another one is currently ongoing [30, 31] (discussed below).

EARLY IDENTIFICATION OF PATIENTS FOR FUTURE THERAPY: WHOM AND WHEN TO TREAT?

In the absence of an effective established treatment for ADPKD children, screening of at-risk children is controversial [32] and not generally recommended [33–35], since this might cause psychosocial and financial harm. Once targeted and safe therapies become available for children, this may clearly change clinician’s attitudes. However, depending on both the efficacy and side-effect profiles of the treatment options, there will be a need for validated methods to stratify patients based on their (predicted rate of) progression. Indeed, correct identification of rapid progressors at childhood age will improve both the cost-effectiveness and benefit-to-risk ratio of therapies. These children will need a more intensive follow-up programme, with an early introduction of supportive treatment. It will be important to identify these patients while avoiding the potential disadvantages of early diagnosis for those children that would not benefit. Indeed, slow progressors should not be over-medicalized and could start lifelong therapy at a later age, but still within the ideal window of opportunity to slow down or halt disease progression maximally. Unfortunately, a validated classification discriminating slow from rapid progressors does not exist in the paediatric population. This is one of the most important challenges in childhood ADPKD for the near future. Ideally, such a classification system is based on a (combination of) biomarker(s), which can be easily determined in all at-risk children at a young age, potentially even without setting the diagnosis. In the adult ADPKD population, different prognostic indicators have been identified from observational studies, and have been shown to correlate with change in renal volume and/or function [2, 36, 37]. Without any doubt, the PKD genotype and total kidney volume (TKV) corrected for height (HtTKV) as measured by magnetic resonance imaging (MRI) are the most extensively validated predictors for disease progression in adults. First, in the Genkyst cohort, renal survival in PKD2 mutation carriers was circa 20 years longer than in PKD1 subjects. Moreover, patients carrying a truncating PKD1 mutation had a worse renal prognosis compared with those with non-truncating PKD1 mutations [38]. Secondly, in the CRISP (The Consortium for Radiologic Imaging Studies in Polycystic Kidney Disease) study, which enrolled relatively young patients (15–46 years of age) with a well-preserved renal function, TKV was identified as a key indicator to select patients with ADPKD at high risk for a progressive decline in renal function, particularly in the early stages of ADPKD where significant renal enlargement occurs prior to any renal function decline [39]. This finding led to the approval of TKV as a prognostic enrichment biomarker by the Food and Drug Administration (FDA) in 2016 [40].

As these parameters enable caregivers to measure ADPKD progression, the next step to take is the prediction of it. Scoring systems, using different combinations of the above-mentioned variables associated with ADPKD progression, have been shown to predict adult patients’ renal outcome [41–43] (Figure 3). The first algorithm, the PRO-PKD score, is based on sex, PKD genotype, the presence of hypertension and/or urologic events such as haemorrhagic events involving gross haematuria or cyst haemorrhages, cyst infections, or flank pain related to cysts, before the age of 35 years. Three risk categories are defined: low risk (0–3 points), intermediate risk (4–6 points) and high risk (7–9 points). These predict corresponding median ages at onset of ESKD of 70.6, 56.9 and 49 years, respectively [41]. As both hypertension and urologic events are well-known symptoms in childhood ADPKD [14], it would be interesting to evaluate the applicability of the PRO-PKD score in children. Another validated tool is the Mayo Imaging Classification, in which patients with typical ADPKD are stratified into subclasses 1A to 1E depending on their HtTKV range for age. Internal validation showed an increased frequency of ESKD onset after 10 years from subclass 1A (2.4%) to 1E (66.9%) while external validation, using the younger CRISP participants, demonstrated increased frequencies from 1C (2.2%) to 1E (22.3%) [42]. Another prediction instrument is the ADPKD Outcomes Model (ADPKD-OM), a fixed-time increment stochastic simulation model based on disease progression equations for HtTKV and estimated glomerular filtration rate (eGFR). The model was internally validated with data from the placebo arm of the TEMPO (Tolvaptan Efficacy and Safety in Management of Autosomal Dominant Polycystic Kidney Disease and Its Outcomes) 3:4 tolvaptan study [43]. Finally, a neural network for eGFR prediction was suggested. This network predicted eGFR within 3 years, based on the input of five consecutive eGFR values, obtained over a 2-year observation period [62]. This method is not validated yet and the current results are not precise enough for clinical trial design as only 33% of the predicted eGFR values were within a 10% range of the of actual eGFR values [63].

As mentioned before, in children, well-validated measures of and predictors for progression are still missing today. Nevertheless, this is required in order to optimize clinical trial design, treatment and patient care. Different potential paediatric progression markers have been suggested, namely the presence of very-early onset (VEO) ADPKD (this is ADPKD diagnosed before the age of 18 months) [64–67]; hypertension [8–10, 68, 69]; glomerular hyperfiltration [70]; genotype [71, 72]; proteinuria [13]; urine concentration capacity [73]; presentation at diagnosis (screening versus symptoms) [11, 67]; and left ventricular mass index (LVMI) [9] (Table 1). So far, none of them has been validated however. The renal volume is often used as an outcome parameter to evaluate potential predictors, but is also frequently studied as a predictor itself [74]. The comparison of genotype and the clinical modifiers between VEO and non-VEO ADPKD might be helpful in the understanding of both predictors and modifiers of the disease. Until now, only two reports comparing VEO and non-VEO subjects have been published from the same cohort, stating that VEO is a risk factor for disease progression [65, 67]. Important in this context as well is the reason for diagnosis, as one study states that the presence of symptoms at diagnosis is a risk factor for disease progression [67], while another demonstrated that children diagnosed due to symptoms versus those due to screening have similar proportions of nephromegaly, hypertension, microalbuminuria and decreased eGFR [11]. Another issue complicating the interpretation of study results is the fact that differences in definitions of the studied predictors are used. As an example, there are studies comparing hypertensive children (≥95th percentile for age, sex and height [75]) with normotensive children (<95th percentile), but also high blood pressure (≥75th percentile) versus normal blood pressure (<75th percentile). Also, a definition of early severe disease is lacking. One study arbitrarily chose to define severe childhood ADPKD as having ≥10 renal cysts by ultrasound before the age of 12 years [74].

Finally, there is a definite need to uniformly define which outcome parameters are of interest in the paediatric population. This is necessary to avoid the use of different primary endpoints in different studies, which dramatically compromises their comparability. In the two paediatric interventional trials, outcome parameters were renal volume, LVMI or microalbuminuria in the first [29], whereas it was ≥20% change of the same parameters in the second [28]. In adult clinical trials assessing treatment to slow down or halt progression, outcome parameters most frequently are measured by (over time change in) TKV and renal function, evaluated by eGFR or serum creatinine, as primary endpoints [37]. However, the highly variable kidney growth curves [39] and the fact that GFR evolution does not always follow a linear decline but can comprise prolonged stable intervals [76], complicates this approach. It is obvious that these biases may even be more relevant in a paediatric population. Other endpoints used are change in liver volume, change in total cyst volume (renal), blood pressure and change in urinary fatty acid-binding protein [37].

A clear definition of parameters of progression and outcome might also help to define the optimal therapeutic window of opportunity, assuming its existence.

(FUTURE) TREATMENT OPTIONS IN CHILDHOOD ADPKD

Disease-specific therapeutic options

Over the last decades, the PC complex, its aberrant downstream signalling pathways in ADPKD and possible approaches to target them have been studied extensively. Still, cystogenesis is not completely understood [77]. A curative treatment is not available so far.

In adults, the only currently available targeted disease-modifying option, the selective vasopressin V2 receptor antagonist tolvaptan, is not curative but only slows down disease progression [78]. The product’s adverse effects are mostly related to increased aquaresis: thirst, polyuria, nocturia and polydipsia, as a result of the excretion of electrolyte-free water. Furthermore, although rare, significant elevations of liver enzyme levels have been described in ADPKD patients on tolvaptan. Based on the results of the TEMPO 3:4 trial [78] and REPRISE (Replicating Evidence of Preserved Renal Function: an Investigation of Tolvaptan Safety and Efficacy in ADPKD) study [79], the European Medicines Agency (EMA) and the FDA approved tolvaptan in May 2015 and April 2018, respectively. The drug is recommended for adult patients who have documented rapid disease progression or are likely to have rapid disease progression [36]. Currently, a phase IIIb double-blind, placebo-controlled clinical trial, focusing on the safety, tolerability, pharmacokinetics and dynamics and efficacy of tolvaptan in ADPKD children, aged 4–17 years, is ongoing [31].

Another therapy tested in paediatric ADPKD is the use of HMG-CoA reductase inhibitors. In a 3-year randomized, double-blind, placebo-controlled phase III clinical trial, including patients with ADPKD aged 8–22 years, pravastatin was shown to slow down the progression of structural kidney disease, as fewer patients in the intervention group [treated with angiotensin-converting enzyme inhibitor (ACEi) and pravastatin] achieved ≥20% increase in HtTKV compared with the placebo group (only treated with ACEi) [28]. Importantly, therapy was well tolerated with no adverse effects on serum liver or muscle enzymes. Currently, pravastatin is approved for use in children from the age of 8 years with heterozygous familial hypercholesterolaemia by the FDA [80], but not by the EMA. The underlying mechanism of how pravastatin leads to potential positive effects in ADPKD remains to be fully elucidated, but statins have been shown to enhance renal blood flow and therefore glomerular filtration, to attenuate inflammation through vascular and glomerular nitric oxide production and may lower cAMP through downregulation of Gαs protein, leading to decreased cell proliferation [27]. Also, in in vitro experiments on proliferating murine tubular cells, simvastatin induced apoptosis in a time- and dose-dependent manner via down-regulation of Bcl-xL expression [81]. Moreover, in the paediatric pravastatin trial, the intervention group had reduced plasma concentrations of cyclooxygenase- and lipoxygenase-derived plasma lipid mediators as compared with the placebo group. These metabolites are known to enhance the pro-fibrotic effects of angiotensin II [82]. In a post hoc analysis of the HALT-PKD (HALT Progression of Polycystic Kidney Disease) trials, statin therapy was not shown to be beneficial in adult ADPKD patients, although the study had several limitations such as a small number of statin users and the use of different statin drugs and doses [83].

Apart from tolvaptan, largely studied therapy classes in adult patients are the mammalian target of rapamycin (mTOR) inhibitors and somatostatin analogues. First, the mTOR inhibitors most probably have no place in paediatric treatment due to the many side effects and the disappointing effects on outcome in adults [84]. The same might apply to somatostatin analogues, which seem to have a role in managing ADPKD-associated polycystic liver disease [85, 86] but until now, data are not suggestive that these agents have the potential to preserve renal function. In the recently presented results of the DIPAK (Developing intervention strategies to halt progression of autosomal dominant polycystic kidney disease) trial, the use of lanreotide in adult patients with advanced disease (eGFR 30–60 mL/min/1.73 m²), was associated with less renal volume growth, but not with an improved rate of eGFR loss. Moreover, the lanreotide group had an increased incidence of adverse events compared with the control group receiving standard care only [87]. In the ALADIN (A Long-Acting somatostatin on DIsease progression in Nephropathy due to autosomal dominant polycystic kidney disease) study, patients with eGFR >40 mL/min/1.73 m² were randomized to a 3-year treatment with octreotide or placebo. After the first year, the average increase in TKV was significantly lower in the intervention group compared to placebo; however, in the third year, this outcome measure did not reach statistical significance. Over the 3 years, the annual reduction in GFR was not significantly different between both groups [88]. However, the full results of the DIPAK trial, ALADIN2 trial and LIPS (Lanreotide In Polycystic kidney disease Study) should be awaited.

The following drugs, currently tested in adult patients, might be tested in paediatric trials in the future, depending on their efficacy and safety in adults: the selective vasopressin V2 receptor antagonist lixivaptan (NCT03487913); AMPK activator metformin (NCT02656017); aldosterone antagonist spironolactone (NCT01853553); and tesevatinib, an inhibitor of several tyrosine receptor kinases. The latter compound is also studied in paediatric autosomal recessive polycystic kidney disease (ARPKD) (NCT03203642).

Promising curative approaches are cell-based and gene therapies, as demonstrated in animal models using bone marrow-derived mesenchymal stromal cells or induced pluripotent stem cells [89, 90]. The use of autologous bone marrow mesenchymal stromal cell infusion in ADPKD patients was shown to be safe and well tolerated in a single-arm phase I clinical trial with a 12-month follow-up. However, as the study included only six patients [91], no definite conclusions on outcome parameters can be drawn [92].

In general, starting treatment in ADPKD children, implying a life-long compliance for a slowly progressing disease that does not necessarily lead to apparent clinical symptoms over a long time, will only be feasible with agents inducing an absolute minimum of adverse effects.

Supportive measures

Counselling and promoting a healthy life style with regular exercise, abstinence from smoking, a healthy diet with appropriate amounts of calories, salt, protein and fat, increased fluid intake throughout the day and treatment of hypertension should be offered from an early age onwards both to patients with clinical disease manifestations and to those at risk for developing ADPKD [26, 93]. As overweight and, particularly, obesity were recently shown to strongly and independently associate with the rate of progression in early-stage ADPKD, a normal body mass index should be pursued from childhood [48].

Hypertension is highly prevalent in children with ADPKD, occurring in one-fifth to one-third of the population [10, 94]. From the HALT-PKD trial, it is known that in young adult ADPKD patients, strict blood pressure control compared with standard control was associated with 14.2% slower annual increase in TKV, reduced urinary albumin excretion and a greater reduction in LVMI, but no overall change in eGFR [95]. Therefore, it was suggested to prescribe ACEi to children with (borderline) hypertension (>75th percentile for age, sex and height [75]), and to achieve a goal blood pressure ≤50th percentile [25]. However, this is not generally implemented in clinical practice, as reflected by the absence of a corresponding recommendation in the currently available position statements [33–35]. Importantly, from a large population-based study, the superiority of 24-h ambulatory blood pressure measurement (ABPM) was shown over clinic or home measurements as a predictor of all-cause and cardiovascular mortality, which was highest in case of masked hypertension [96]. A recent study on ABPM in ADPKD children not only confirmed the prevalence of hypertension, but also showed high rates of nocturnal non-dipping and isolated nocturnal hypertension [10]. However, it is still unclear whether these findings should prompt antihypertensive treatment or not.

Until now, one trial studying the effect of ACEi in patients with ADPKD aged 4–21 years is published. Over a 5-year follow-up period, the authors were unable to demonstrate a significant effect of ACEi on renal growth [29]. However, it is unclear whether this study was sufficiently powered, with 85 patients included in six different subgroups, depending on their baseline blood pressure and treatment allocation, and a dropout rate of 27%. Moreover, while ACEi monotherapy was effective in controlling blood pressure in children with borderline hypertension, multiple antihypertensive medications were necessary in hypertensive subjects.

FUTURE PERSPECTIVES

Given the limited data on clinical courses of paediatric ADPKD, more clinical research is needed to describe the development of patients during childhood and adolescence and the potential effects of, for example, early antihypertensive treatment. International collaborative research approaches have shown an enormous potential to identify and characterize specific aspects of paediatric kidney disease and to generate an observational evidence base for clinical recommendations. This includes studies on estimating times to ESKD in children with chronic kidney disease [97], cardiovascular consequences of chronic kidney disease in childhood [98, 99] and courses of paediatric dialysis [100]. A more recent study, ARegPKD, is dedicated to describe the details of clinical courses of patients suffering from ARPKD [101, 102]. This register study has collected the largest cohort of ARPKD subjects, as a basis to identify, for example, clinical risk markers of disease progression [103]. For ADPKD, the characterization of a similar cohort study, describing the courses of children diagnosed with ADPKD, is currently initiated as ADPedKD (https://www.adpedkd.org). In this international, web-based database, prospective and retrospective data on multiple clinical and genetic aspects are collected in basic data questionnaires and yearly follow-up visits to understand long-term clinical courses. In this way, ADPedKD aims to establish the first large and clinically deeply phenotyped paediatric ADPKD cohort. Obviously, such a work needs to be complemented by detailed information on genotype and family history to establish cohorts that fulfil unified disease criteria. Moreover, both paediatric and adult nephrologists involved in clinical ADPKD research should be stimulated to cooperate on the future scientific transition of childhood cohorts to the adult setting. This will allow investigators to study long-term follow-up, even after subjects have reached adulthood, which would be of great value to correlate long-term outcome parameters with early markers of progression.

In translational approaches, the collection of corresponding biosamples will be helpful and required to benefit from the power of modern high-throughput technologies with the potential to identify novel and complimentary disease-specific biomarkers. Non-invasive collection of biosamples such as patient cells from urine may help to study the effects of specific mutations.

In summary, the clinical and research communities on ADPKD are faced with important questions related to the nature of potential interventions, and their optimal indications and timing. From a paediatric perspective, challenges include the identification and validation of indicators for disease progression from childhood in order to discriminate slow from rapid progressors, and to define outcome measures. The ADPedKD initiative will help in elucidating these questions.

ACKNOWLEDGEMENTS

This paper is part of a Supplement on ADPKD supported by an educational grant from Otsuka Pharmaceutical Europe Ltd.

FUNDING

D.M. is supported by the Clinical Research Fund of UZ Leuven and by the Fund for Scientific Research, Flanders G0B1313N. S.D.R. is supported by the Fund for Scientific Research, Flanders 11M5214N. M.C.L. is supported by the Marga and Walter Boll-Foundation, and the German Federal Ministry of Research and Education (BMBF grant 01GM1515).

CONFLICT OF INTEREST STATEMENT

None declared.

REFERENCES