Improving clinical trial design for inquiries into the mechanisms of cyst growth in ADPKD

Accurate and reliable estimates of kidney and cyst volume, the hallmark of disease progression in ADPKD are now available. These powerful and exciting tools make it possible to consider both short and long term randomized clinical trials in ADPKD at various stages of disease. Highlights of the work by Kistler and colleagues are now provided.

Accuracy, precision, and reliability of renal and cyst volume measures in human ADPKD

Estimates of total renal and cyst volume based on magnetic resonance (MR) without gadolinium demonstrate high accuracy, precision and reliability throughout a broad range of renal volumes in autosomal dominant polycystic kidney disease (ADPKD). Kistler and colleagues (1) present a meticulous and elegantly executed study of MR-based renal and cyst volume estimates over time in human ADPKD. Manual measurements using trained readers demonstrated > 0.99 concordance coefficients for inter- and intra-reader variability for total renal volume and slightly less (0.944) for total cyst volume. These findings are similar to the Consortium for Radiologic Imaging Studies in Polycystic Kidney Disease (CRISP) study (2) (inter-reader variability: 2.1% and intra-reader variability: 2.4% for renal volume) where gadolinium was used. High accuracy, precision and reliability without gadolinium enhancement allows for the benefit of reduced cost of imaging and removes the potential risk of developing nephrogenic systemic fibrosis.

Renal volume measurements in the study by Kistler et al (1) study were completed with the use of manual segmentation by trained personnel supported with software. The investigators acknowledge that manual segmentation requires significant effort and is therefore restricted from widespread use. Manual segmentation precludes clinical implementation for patients with ADPKD, and a high priority to develop sophisticated image analysis software to support more automated accurate and reliable renal and cyst volume measurement is needed in ADPKD. Development of such technology would allow clinicians to focus on patients at high risk for progression to renal failure and provide valuable predictive information to their patients.

Cyst volume measurements in this study were performed using stereological methods in T2-weighted images, similar to renal volume measures in the CRISP study (2). Intra- and interobserver variability was similar to that found in CRISP, without the use of gadolinium. However, estimations of non-cystic parenchyma without gadolinium are not feasible and others have suggested this may be an important consideration when evaluating disease progression in ADPKD (3). In addition, estimates of renal function, using contrast-enhanced MRI, a novel functional application of renal MR imaging (4) are also not possible in this setting.

Potentially the largest source of variability of renal volume estimates in ADPKD was not determined in this study. The CRISP study preliminarily addressed day-to-day variability of renal and cyst volume within ADPKD individuals by studying 4 patients on 4 occasions. Relatively low day-to-day variability (2.4%) was found using gadolinium enhanced images. These individuals were repeatedly studied over relatively short intervals (approximately 2 weeks between studies) (5) indicating that the day-to-day variability of renal volume in ADPKD is significantly less than other biologically relevant variables such as blood pressure and renal function. However, given the safety issues and cost savings related to gadolinium use, day-today variability of renal volume without gadolinium needs to be addressed more fully in larger numbers of patients over shorter periods of time at varying stages of hydration and dehydration.

Utilizing renal volume as an outcome measure in randomized clinical trials in ADPKD - study design considerations

Kistler et al (1) provide additional information regarding the frequency and cadence of medical complications related to ADPKD and renal volume. A significant association between clinical complications of ADPKD and renal volume was found. Increased renal volume was associated with greater frequency of hypertension, macrohematuria, urinary tract infections and recurrent flank pain. Importantly, 10% of participants demonstrated a significant reduction in renal volume over 6 months. These individuals demonstrated greater baseline renal volumes and more complications related to ADPKD than those who did not demonstrate a dramatic reduction in renal volume. Understanding the frequency of these events, as well as potential contributors (for example, a rapidly increasing rate of cyst or renal growth) provides critical information for establishing inclusion criteria for enrollment into randomized clinical trials in ADPKD as well as considerations for analytical approaches that account for these events during trial participation. Given the dramatic volumetric effect of spontaneous cyst involution or rupture, subjects at high risk for disease progression may be inadvertently excluded depending on the inclusion criteria used.

Although Kistler and colleagues (1) demonstrate that MR measurements every six months are sufficient to demonstrate a detectable change in renal volume in ADPKD, caution is warranted in considering this duration of participation in a randomized clinical trial. If one anticipates that a therapeutic agent has a significant impact (for example 25% reduction in the rate of cyst or renal expansion), assessing change in renal and cyst volume over a six month duration will be insufficient to detect real therapeutic effects. This magnitude of change is potentially less than the level of detection with current MR-based imaging protocols. On the basis of volumetric measures in the study by Kistler et al (1) and the CRISP study, only half of the subjects could be included in a 6 month study where this magnitude of effect is anticipated, and many individuals with progressive renal disease would be excluded.

Given the complex nature of cyst growth in ADPKD, such as alterations in epithelial proliferation, rates of apoptosis, extracellular matrix responses and the presence of cyst fluid secretion, multiple agents targeting different aspects of cyst growth and expansion will most likely be used in ADPKD. The duration of drug exposure to determine whether therapeutic impact is possible will vary on the basis of the aspect of cyst growth and formation targeted. Investigators will need to take into account the pace at which these alterations contribute to cyst growth in ADPKD. For example, inhibition of the renin-angiotensin system in ADPKD, may provide therapeutic benefit for renal and cyst growth in ADPKD over a relatively long period of time where tissue fibrosis and vascular remodeling are chronically affected. Alternatively, with agents that acutely alter cyst fluid secretion, such as secretin or somatostatin analogues, shorter and intermittent exposures may be sufficient to demonstrate a beneficial effect. For those interventions that affect epithelial proliferation or apoptosis (such as mTOR inhibitors) exposure time may need to be intermediate to determine efficacy.

Risk stratification for disease progression using renal volume estimates in ADPKD

This study by Kistler et al (1), as well as the CRISP study, highlights the important fact that ADPKD individuals at high risk for progression can now be identified by MR based measures. Renal volume can provide a stratification to high or low risk for disease progression in ADPKD. In the study by Kistler et al (1), those with the greatest renal volume demonstrated the fastest rate of increase in renal size. In addition, the greatest renal volume was associated with the greatest frequency of complications in ADPKD, including the frequency of hypertension, the presence of albuminuria, frequency of episodes of gross hematuria and asymptomatic cyst rupture. The lessons of Kister et al (1) show that it is reasonable to justify renal volume as a marker of disease severity and progression in ADPKD. Young ADPKD patients with large renal volumes demonstrate the most rapid progression of disease and should be identified as high risk ADPKD patients. Importantly, the magnitude of renal volume change correlates with initial renal volume throughout the study population but most strongly in the younger than 30 year age group.

Mechanisms responsible for cyst and renal growth in ADPKD

In both in this and the CRISP study, an extremely high level of agreement between left and right kidney volume was found. In addition, both studies demonstrated a strong association between change in left and change in right kidney volume over time. This level of agreement between kidneys within individuals where random somatic mutations or second hits are required for cyst development is difficult to reconcile. Only a small proportion of individuals demonstrated significantly discordant renal volumes and many of these were the individuals who developed asymptomatic cyst rupture. Heritability indices of renal volume in healthy controls have not been established with MR, but in ADPKD, heritability indices are extremely high, greater than 85% (6). The high correlation between left and right renal volumes in CRISP and the study by Kistler et al suggests that heritability of renal size in ADPKD is extremely high. Although somatic mutations or modifier influences play a significant role in disease progression in ADPKD, studies addressing the allelic contribution to disease severity (PKD1 or PKD2 mutations) deserve further attention.