Background

Polycystic kidney disease (PKD) is the most common genetic disorder leading to end stage renal disease (ESRD). Autosomal dominant polycystic kidney disease (ADPKD) is caused by germline mutations in PKD1 (85 % of cases) and PKD2 (15 % of cases), and is typically diagnosed later in life than autosomal recessive polycystic kidney disease [1–4].

ADPKD is characterised by the progressive development of multiple bilateral renal cysts and has a prevalence of less than five patients per 10,000 of the population in the EU [5–7]. ADPKD is a systemic disease, with extra-renal manifestations including liver cysts, intracranial aneurysms and hypertension [3, 8]. Approximately 70 % of patients with ADPKD will progress to ESRD at a median age of 56 years [9, 10]. Patients with ADPKD may experience chronic pain, which may be debilitating, in addition to increased morbidity due to enlarged kidneys, which can lead to poor health-related quality of life (HRQoL) and reduced social interaction [3, 11].

ADPKD is associated with a high degree of clinical variability between patients, both within and between families, especially in terms of the likelihood and rate of progression towards ESRD. In the early stages of ADPKD, before renal function is significantly compromised, progression can be difficult to detect as patients are often asymptomatic. Currently, there is a lack of consensus in the published literature regarding the optimal factors for prediction of renal outcomes and the ideal variables that should be measured in order to monitor disease progression, especially in the early stages of the disease. This has led to difficulties in identifying patients at high risk of ESRD and in defining appropriate primary endpoints to consistently measure and compare the efficacy of therapies in development.

There is a requirement for greater understanding of the predictors of disease progression in ADPKD in order to optimise clinical trial design, treatment and patient care. The objective of this systematic review was to identify publications that detail the natural history of ADPKD, considering the indicators of early- and late-stage progression. It was envisaged that this review would aid clinical practice by informing future research and the development of a predictive model to estimate the likely rate of progression and ultimate long-term outcomes in patients with ADPKD.

Methods

Systematic literature search

A systematic review of the literature was conducted to identify publications that detail the natural history of ADPKD, including diagnosis, prognosis and progression. In the absence of a definitive measure of disease progression, all measures linked to the progression of ADPKD were considered, although it was anticipated that those linked to cyst development/renal enlargement and renal function would be the most relevant.

The systematic review was conducted and reported according to the principles in the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement [12] (Additional file 1). A search was conducted in the following electronic databases:

  • MEDLINE and MEDLINE In-Process

  • Embase

  • Cochrane Central Register of Controlled Trials (CENTRAL)

  • The Cochrane Database of Systematic Reviews (CDSR)

  • NHS Economic Evaluation Database (NHS EED)

  • Health Economic Evaluations Database (HEED)

  • BIOSIS.

Electronic databases were searched on 7th April 2014 using a structured search string, including terms for ADPKD and study type (Additional file 2). The search was restricted to publications from the previous 10 years in all databases except BIOSIS, which was limited to the past 3 years. Publications in the field of ADPKD have become significantly more numerous in recent years; therefore, it was considered that a sufficient number of relevant articles would be identified and that any earlier studies containing key information would be identified through reference explosion. In addition, diagnostic and prognostic techniques have progressed with the advancement of technology, with earlier publications often relying on subjective measures of progression, including kidney palpation.

‘Grey literature’ searches of relevant congresses, limited to the past 3 years, were also conducted (World Congress of Nephrology, American Society of Nephrology, American Academy of Nephrology, British Renal Society, European Renal Association and International Society for Pharmacoeconomics and Outcomes Research). Reference lists of systematic reviews and meta-analyses identified by the search were reviewed to identify additional relevant articles that were not identified in the systematic search due to study type restrictions.

Studies meeting the following criteria were considered for inclusion in the review:

  • Study types: clinical trials, observational studies, systematic reviews, meta-analyses and non-systematic reviews

  • Population: human, adults (≥18 years of age) with PKD, or ADPKD specifically

  • Outcomes: diagnosis or prognosis

  • Language: English.

Study selection

Identified articles were screened by one reviewer, according to the criteria specified above. Articles were filtered using a positive exclusion method, whereby articles with insufficient information to warrant their exclusion remained in the review [13].

Initially, titles and abstracts of identified articles were reviewed (title/abstract screening) according to a checklist; full-text papers of the publications remaining after title/abstract screening were reviewed according to the same checklist (full text screening). Papers assessing HRQoL were excluded at full text screening if they did not report data on the progression of ADPKD.

Publications that included assessment of the link between ADPKD progression and genetic markers were only included in the review if they assessed the link with genotype (PKD1/PKD2). Publications assessing genetic markers at a molecular level, i.e. studies assessing specific single nucleotide polymorphisms, were excluded because this level of detail is not readily available to nephrologists in clinical evaluations. An independent reviewer checked a random selection of articles (10 %) at both title/abstract and full text screening to ensure consistency and accuracy.

Data extraction

Data were extracted from studies that met the inclusion criteria. If more than one article was found to present data from the same study population, results were collated as appropriate.

Extracted data consisted of:

  • Study characteristics, such as study design, duration and location

  • Patient characteristics (see ‘Definitions’)

  • Clinical characteristics and outcomes linked to disease progression (see ‘Definitions’)

  • Information regarding ESRD in the patient populations, such as age at ESRD onset and length of time on dialysis.

Definitions

For the purposes of this review, ‘patient characteristics’ were demographic data and ‘clinical characteristics’ were defined as baseline clinical parameters, baseline measures of renal size and function, and characteristics that were not influenced by changing renal parameters, e.g. genotype (Table 1). ‘Outcomes’ were defined as any measures or markers of disease progression, such as parameters reported over a time period, or measurements taken at both baseline and end of the study. The reporting of symptoms of ADPKD, such as pain, may have been considered as patient characteristics if only reported at baseline, but may also have been considered as outcomes if reported over time.

Table 1 Patient characteristics and outcomes in ADPKD
Full size table

Glomerular filtration rate (GFR) was a recognised measure of renal function, and was either estimated based on serum creatinine concentration using equations such as the Modification of Diet in Renal Disease (MDRD) equation or the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation (estimated GFR [eGFR]), or measured by creatinine clearance (measured GFR [mGFR]). For simplicity, we used the term ‘GFR’ to refer to studies that used either eGFR or mGFR or both, or where methods of determining GFR were not defined. Change in GFR was used as a measure of renal function and change in total kidney volume (TKV) was used for renal volume in this review.

Results

Figure 1 shows the selection process for articles included in the systematic review. A total of 2799 articles were identified in the original literature search, with a further 444 relevant conference abstracts identified. A total of 254 articles proceeded to the data extraction stage. Of the 254 papers that progressed to data extraction, 160 were observational studies, 33 were clinical trials and 61 were reviews.

Fig. 1
figure 1

PRISMA diagram of the systematic review. PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-Analyses

Full size image

Observational studies

The majority of the identified studies were observational studies (63.0 %), including case studies, registry reviews and patient surveys with durations ranging from 6 months to more than 40 years. The studies employed a range of different objectives, with many aiming to increase understanding of ADPKD. Most studies reported baseline patient characteristics, including average age and gender (65 and 58.8 % of studies, respectively), while symptoms of ADPKD frequently reported at baseline included hypertension and proteinuria (21.3 and 10.6 % studies, respectively). The most commonly reported measure of renal function, described in 66 (41.3 %) observational studies, was GFR. The most frequent measure of renal size was TKV (34 studies, 21.3 %).

Of the 38 (23.8 %) publications that reported outcomes of ADPKD measured over time, disease progression was measured in 32 studies (84.2 %) by GFR, estimated most commonly by the MDRD equation (41.6 %). Change in TKV was reported as a measure of disease progression in 12 studies (31.6 %), with 11 publications using magnetic resonance imaging (MRI) and one publication using computed tomography. Renal enlargement over time as a surrogate marker of disease progression was also measured by total cyst volume (TCV) (five studies) [14–18], volume of each kidney (four studies) [19–22], and total renal volume (TRV) (three studies) [19, 22, 23].

In total, 19 studies reported patient and clinical characteristics of patients with ESRD. Of these studies, 18 reported age at ESRD and 11 reported the number of ADPKD patients reaching ESRD. Baseline characteristics of patients in each of the study groups were not reported in eight studies [24–31], but others reported a number of parameters including age (seven studies) [32–37], serum creatinine (four studies) [37–40], GFR (three studies) [33, 35, 38], TKV (three studies) [33, 36, 40], hypertension (one study) [41] and PKD mutation (one study) [42].

Publications were identified that measured GFR decline or increase in TKV (disease progression) and assessed the associated patient and clinical characteristics, defined in Table 1. If the association with patient and clinical characteristics was significant, these were then defined as prognostic indicators in the prediction of disease progression (Table 2).

Table 2 Summary of observational studies that identified prognostic indicators for ADPKD
Full size table

TKV and age at baseline were the two factors most commonly cited as significantly associated with a faster rate of ADPKD progression, each cited in ten studies (Table 3). A wide range of prognostic indicators (a total of 26) were reported, many of which were only reported by one study, indicating a lack of consensus in the literature on the parameters implicated in ADPKD progression.

Table 3 Publications that reported a correlation between prognostic indicators and change in renal function or volume
Full size table

Seventeen publications from the Consortium for Radiologic Imaging Studies in Polycystic Kidney Disease (CRISP) study were identified by the systematic review [15–17, 43–56]. CRISP is a prospective, long-term observational study, including 241 ADPKD patients with normal renal function who were considered at high risk of renal insufficiency. The study was initiated to increase the knowledge and understanding of ADPKD prognosis by establishing reliable measures of disease progression [2]. Patients were diagnosed due to a positive family history (asymptomatic family screening) or based on signs and symptoms related to the disease. The study includes patients with defined hypertension diagnosed before the age of 35 years; ADPKD diagnosed in utero or in the first year of life; 24-h urinary protein excretion greater than 300 mg/day; or an episode of gross haematuria in men before the age of 30 years (CRISP I population) [57]. A range of measurements have thus far been reported, and patients have been stratified by several different criteria across the CRISP publications, demonstrating the relevance of a number of prognostic factors, such as gender and age at diagnosis (Table 2). CRISP has identified the importance of TKV as a key prognostic indicator, particularly in the early stages of ADPKD where significant renal enlargement can occur prior to the loss of renal function [2, 46]. The CRISP study is ongoing, with 203 patients having been re-enrolled in CRISP II from July 2007 [15].

Patients were stratified by rate of ADPKD progression, and groups defined as fast or slow progressors (large or small annual decline in GFR, respectively), in four observational studies (Additional file 3) [56, 58–60]. Although this concept is important in the clinical setting, there is no definitive conclusion in the literature as to how the rate of progression should be measured or stratified. Across the four studies, a range of prognostic indicators were reported, including initial GFR, high blood pressure, age at diagnosis and TKV at a given time point.

Disease severity was correlated with HRQoL in a single study [11]. In general, for patients with a lower GFR, Short Form-36 health survey scores were lower, indicating reduced HRQoL [11].

A research group in Brest, France, conducted an observational study assessing 26 clinical, biological and genetic variables (not specified) for use in a prognostic model. The model was based on data from 1017 patients with ADPKD, and was tested in a sub-population of 255 patients who had either reached ESRD or were aged over 60 years [41]. The presence of hypertension along with at least one urologic complication, or the presence of a truncating PKD1 mutation was reported to be associated with poorest renal outcome [41].

Clinical trials

The 33 clinical trial publications were reviewed to ensure the same data were not extracted from multiple reports of the same study. Of these 33, 24 unique reports were identified. The trials assessed the efficacy of three main drug types to slow the progression of ADPKD:

  • Mammalian target of rapamycin inhibitors, e.g. sirolimus, everolimus and tacrolimus [61–67].

  • Somatostatin analogues, e.g. somatostatin, lanreotide and octreotide [68–73].

  • The vasopressin v2 receptor antagonist, tolvaptan [74–80].

Further drugs were:

  • Eicosapentaenoic acid, an anti-inflammatory omega-3 fatty acid [81].

  • Pravastatin, a statin [82].

  • Anti-hypertensives, e.g. angiotensin-converting enzyme inhibitors (ACEIs) (enalapril or ramipril) [83, 84] or the beta-blocker metoprolol [84].

In the clinical trials assessing treatments to halt or slow the development of ADPKD, efficacy was assessed using measures of disease progression. Change in TKV or renal function, measured by GFR or serum creatinine concentration, were the most commonly used primary endpoints in the identified trials.

Other primary endpoints considered in studies of disease-modifying interventions in patients with ADPKD included change in liver volume, change in TCV and change in urinary fatty acid-binding protein. The number of different primary endpoints used in the identified trials demonstrated inconsistencies in the measurement of efficacy due to the paucity of data on the progression of ADPKD.

In the 21 publications that reported data at multiple time points, the most frequently reported outcomes were TKV (14 reports, 66.7 %), eGFR (12 reports, 57.1 %), serum creatinine (7 reports, 33.3 %) and systolic blood pressure (9 reports, 42.9 %). Of these, 12 studies reported data at two time points and eight studies reported data at three or four time points. Data were recorded at baseline and after 6 months in six studies [62, 65, 68, 71–73], at baseline and after 12 months in one study [70], at baseline and after 24 months in three studies [61, 66, 82], and at baseline and after 36 months in two studies [83, 84].

Due to the short study duration and lack of suitable sub-analyses or patient level data, disease progression data from these trials were not appropriate for the assessment of the prognostic potential of the variables measured at baseline.

Hypertension was targeted as a manifestation of ADPKD in the Halt Progression of Polycystic Kidney Disease (HALT-PKD) study, a clinical trial studying the intensive blockade of the renin-angiotensin system with combination ACEIs and angiotensin receptor blockers (ARBs). The study aimed to test the hypothesis that rigorous blood pressure control was more effective in slowing progression of renal disease in early ADPKD than moderate blood pressure control, by comparing the combination treatment to ACEI monotherapy alone. The primary endpoint was annual change in TKV, which was lower with rigorous blood pressure control than with moderate blood pressure control (P = 0.006). The rate of change in GFR was similar for the two treatment groups [85, 86].

Reviews

The 61 non-systematic and systematic reviews and meta-analyses identified were mainly focussed around new and potential treatment options, as well as reviews of the genetics, pathophysiology and manifestations of the disease. In general, the identified reviews acknowledged and discussed a range of different prognostic indicators, but did not provide conclusive agreement on these factors.

Discussion

Our review demonstrated that, as a result of the extensive clinical variability associated with ADPKD, a range of biological, genetic and clinical characteristics have been reported in the published literature for both the prediction and measurement of ADPKD progression. Currently, therapeutic options for ADPKD focus on management of the symptoms and complications of the disease since there is no available treatment to slow ADPKD progression. Stratification of patients based on their predicted rate of progression could improve symptom management. The uncertainty surrounding the clinical progression of the disease has also contributed to a paucity of clinical trials investigating potential interventions.

Due to the variability in the rate of ADPKD progression and the length of time taken for patients to reach ESRD, progression to ESRD is a poor candidate endpoint for use in clinical trials. Instead, predictors/indicators of time to disease progression have been used to assess the efficacy of treatments; however, a variety of different endpoints have been employed. This has led to problems in demonstrating conclusive efficacy and in comparing treatments across trials. As a result, no consensus has been reached regarding the optimal endpoints to assess the progression of ADPKD and the efficacy of potential therapies. A range of observational studies have attempted to address this, but there is inconsistency in the parameters reported and in the assessment of disease progression.

Studies evaluating ADPKD have generally been conducted in relatively small study populations, because the slow rate of progression and the lengthy period during which GFR is within the normal range limits the size of the population available for analysis. The CRISP observational study is a rich data source that has assessed several prognostic indicators in a relatively large cohort of patients with ADPKD, selected in order to study the clinical and radiological features of perceived fast progressors over a long follow-up period (6 years) [15, 46]. To assess the relationship between baseline TKV and renal growth over time, the cohort was stratified; a greater baseline TKV at a younger age was associated with a more rapid increase in TKV [46].

CRISP also illustrated that, although GFR does not typically change until the fourth or fifth decade of life, renal enlargement progresses significantly in the early stages of the disease, identifying large TKV as a key indicator of early disease progression [2, 46]. This provides evidence that early treatment may be important to slow or halt the progression of ADPKD before cyst development causes irreversible damage [46]. Other prognostic indicators of relevance included hypertension, gender and age at diagnosis, some of which correlate with those identified by other studies [2]. It is anticipated that these prognostic indicators will be used in the future evaluation of treatments that aim to slow the progression of ADPKD and will lead to improvement and standardisation of trial design [2]. However, the CRISP study also has limitations in that, compared with the general ADPKD population, CRISP enrolled a younger cohort with well-preserved renal function. Further analyses are required to validate these findings in the wider ADPKD population with a broader range of renal function and volumes.

Since the searches were conducted, three abstracts from the PKD Outcomes Consortium (PKDOC) have been presented [87–89]. The PKDOC dataset is the largest ADPKD dataset currently available and includes data from CRISP in addition to other study populations from the Mayo Clinic, Emory University and the University of Colorado [89]. These abstracts validate the use of TKV as an important prognostic indicator, identifying baseline TKV and baseline eGFR as prognostic biomarkers for both eGFR decline and progression to ESRD.

Our systematic review has identified a wide range of prognostic indicators that have been proposed or assessed in observational studies and clinical trials, but there is currently no agreement within the literature. The absence of long term patient level data from the identified clinical trials prevented correlations between variables measured at baseline and the rate of disease progression. In observational studies, age and TKV at baseline were the two factors most commonly cited as significantly associated with a faster rate of ADPKD progression. However, further research is required to validate these factors as true indicators of disease progression. Reported prognostic indicators may be complications of ADPKD that are exacerbated by disease progression, rather than factors driving the underlying progression of the disease, or may exhibit a cause-effect association, in which they both exacerbate and are worsened by the progression of the disease.

The specific genotype of patients with ADPKD has also generated interest as a potential indicator of disease progression rate. Although it is generally agreed in the literature that patients with PKD1 mutations have larger kidneys and worse renal function compared to patients with PKD2 mutations, few papers report the use of GFR and/or TKV to measure disease progression in patients stratified by genotype. Therefore, despite the large number of potential prognostic indicators identified, there may be an interaction between some, possibly enabling the collapse of several prognostic indicators into fewer key factors that are most informative. The two most commonly reported prognostic indicators identified by this review were age and TKV at baseline, which may encapsulate other factors such as PKD gene mutation, renal blood flow and baseline GFR.

Due to inter-patient heterogeneity in renal size and genetic profile, current diagnostic practices based on renal ultrasound and family histories are suboptimal. Recently, urinary proteomic and microRNA biomarkers have been identified that could be used as a non-invasive method of diagnosing patients with ADPKD [90, 91]. Long-term studies are ongoing to determine whether such markers also serve as prognostic indicators in ADPKD [90]. However, initial results are promising as a correlation was observed between the urinary peptide profile and height-adjusted TKV in patients with ADPKD [91].

The identification of prognostic indicators is of great importance in the drive to improve the management of ADPKD, as a better understanding of patients’ clinical prognosis may lead to improved symptom management that can be tailored to individuals, depending on their predicted rate of progression. This knowledge will also aid the stratification and selection of patients for disease-modifying treatments, as well as providing a basis for improved clinical trial design with standard endpoints to evaluate and compare treatments. A more thorough understanding of the natural history of ADPKD may be beneficial in terms of optimising treatment. The construction of a disease progression model to predict the progression of ADPKD, incorporating the identified prognostic indicators, could be informative for the future management of the disease.

Conclusions

There has previously been little clinical consensus regarding the prognostic indicators associated with disease progression in ADPKD, especially in the early stages of the disease. As a result, ADPKD progression has been assessed using a wide range of patient and clinical characteristics. A systematic review of the literature has identified age and TKV as the two factors most commonly cited as significantly associated with a faster rate of ADPKD progression, particularly in the early stages of the disease. By identifying these prognostic indicators it is hoped there is an opportunity to improve trial design and conduct further research, including a predictive model of ADPKD natural history, in order to provide clinicians with a clearer understanding of factors influencing disease progression, thereby helping to optimise patient care.