Several new effective therapies have emerged in the last 5 years for patients with CKD, including sodium-glucose cotransporter 2 inhibitors and the mineralocorticoid receptor antagonist finerenone. Furthermore, many new therapeutic strategies and interventions are in advanced stages of development to further reduce CKD progression. These new therapies were tested in dedicated randomized controlled trials (RCTs), which provide high-level evidence about the efficacy and safety of existing or novel interventions for the population at large. However, these RCTs did not take into account variability in underlying disease mechanism and response among individuals. However, advancements in the application of high-throughput data-intensive analytical and bioinformatic technologies have revealed a large variation in disease pathophysiology, molecular pathways involved in disease progression, and response to treatment. This is illustrated by a recent kidney-transcriptomic biopsy study demonstrating a large heterogeneity in tissue morphological features and molecular signatures that could not be captured with the current classification system for CKD.1
Because of the substantial heterogeneity in underlying disease pathophysiology and response to treatment, clinical trials recruit a large number of patients and consequently have become complex and increasingly challenging to manage. With the advent of multiple new effective therapies, the average CKD progression rate is decreasing, which will likely lead to even larger trials to maintain sufficient statistical power. Indeed, the aim of a recently announced clinical trial in nephrology was to recruit 11,000 patients.2 This appears to be unsustainable for the future.
New clinical trial designs should therefore be considered that take into account the variation in therapy response according the concept of personalized medicine. This has the potential of smaller and more efficient clinical trials and better patient outcomes. Some early studies in the nephrology area have explored innovative clinical trial designs and have provided important insights to apply personalized medicine approaches in future clinical trials.
Clinical Trial Designs for Personalized Medicine
Different clinical trial designs can apply personalized medicine strategies. They can be classified into four main categories. The categories, their definitions and methodological principles, and examples in CKD are described in Table 1.
Table 1.
Characteristics clinical trial designs applied to personalized medicine
| Category | Study Design Example | Study Design Definition | Study Design Methodology | Application in Nephrology |
|---|---|---|---|---|
| Single or multiple cross-over | N=1 study | N=1 studies evaluate individual patient responses to interventions by randomly allocating different time periods within an individual to repeated intervention(s) and control | An individual is randomized to a series of fixed treatments of fixed duration with crossover between the experimental intervention and control | ROTATE trial3PRECISION trial4 |
| Enrichment | Sample enrichment design | The enrichment design recruits patients more likely to reach the end point on the basis of a single or multiple biomarkers. Biomarker-negative patients are not randomized or followed | Before randomization, patients are stratified according to their risk profile. Biomarker-positive patients will be randomly allocated to the experimental or control intervention | PRIORITY trial5 |
| Biomarker strategy | Biomarker strategy with treatment randomization | The biomarker-strategy design with treatment randomization according to biomarker status enables identification of biomarker-positive subgroup(s) in whom the experimental intervention is superior compared with control in the targeted population. It can also assess drug efficacy in the biomarker-positive versus biomarker-negative group | Patients are stratified according to the biomarker threshold into biomarker-positive and biomarker-negative subgroups and subsequently randomly allocated to the experimental or control intervention | SONAR trial6 |
| Master protocol | • Basket trial • Umbrella trial • Platform trial |
Master protocol RCTs assess multiple experimental interventions or multiple (sub)populations in parallel under an overarching single protocol with trial-specific appendices | A basket trial is conducted across various CKD etiologies and characterizes the drug effect in multiple disorders, allowing for a wider range of potential indications for a given therapy An umbrella trial can be conducted in a conventionally defined disease (e.g., diabetic kidney disease). Within the disease, various biomarker-based subgroups are defined and different drugs are tested in these subgroups. This design supports individualizing therapies In a platform trial, randomization can start with equal distribution across different interventions but can be adjusted to favor assignment of patients (potentially defined by patient characteristics/biomarkers) to interventions with higher response rates |
• AFFINITY (basket) • NEPTUNE-Match (umbrella)7 • CAPTIVATE (platform) |
RCT, randomized controlled trial.
Crossover Designs
Single or multiple crossover RCT designs using one or multiple interventions determine treatment efficacy and safety at the individual patient level. These trials, also referred to as n=1 trials, randomize the same patient to multiple treatment blocks of fixed duration with wash-out periods in between. The treatment blocks can consist of the same intervention or multiple interventions. This trial design requires four conditions: the time course of the effect is fast, the outcome can be objectively measured, the disease is stable during the trial period, and the treatment response disappears during the wash-out period. An example of a multiple crossover trial in nephrology is the ROTATE trial, which assessed the albuminuria-lowering effect of four different interventions in patients with diabetes and albuminuric CKD.3 This study showed a large variation to four different interventions within and between individuals. Another example of a multiple N=1 clinical trial is the PRECISION trial in hypertension, which showed a similar heterogeneity in BP responses to four different antihypertensive drug classes.4 These trials support the potential of personalized medicine to address patient heterogeneity in complex multifactorial diseases.
Although N=1 trials provide comprehensive insight if the intervention works and for whom, they can only assess treatment effects on surrogate outcomes. Long-term clinical outcomes cannot be determined as it violates the first condition of a stable disease. Other trial designs are available.
Enrichment Designs
Enrichment clinical trials are RCTs that use single or multiple biomarkers to enrich the population based on the risk of the outcome or likelihood to respond. These trials only compare the intervention with the control in the biomarker-positive subgroup. The biomarker-negative subgroup is excluded from the trial. In essence, past RCTs in nephrology that recruited patients with low GFR or high albuminuria fulfill the definition of an enrichment design. A more sophisticated enrichment trial starting to address the mechanistic heterogeneity of the targeted patient population and connecting this to a drug's mechanism of action was the PRIORITY trial.5 In the PRIORITY trial, high-risk patients were selected on the basis of a proteomic risk score mainly consisting of fibrosis markers. Only biomarker-positive patients were randomized to spironolactone or placebo based on evidence that spironolactone may have antifibrotic properties. However, the results of the study were inconclusive due to lower than anticipated number of randomized patients and a low event rate.
Randomized Biomarker-Stratified Designs
In contrast to the enrichment design where only the biomarker-positive subgroup is selected, in the biomarker-stratified design, all individuals are stratified into biomarker-positive and biomarker-negative subgroups on the basis of the results of the biomarker assessment. This allows assessment of drug efficacy and safety in the overall population, in the biomarker-positive and biomarker-negative subgroups separately, and assessment if the biomarker-stratified approach is better than treating the overall population. The biomarker assessment for stratification can be based on a single measurement at baseline (predictive biomarker) or the change in the biomarker after a short-term exposure to the active intervention, for example 6 weeks (dynamic biomarker). The SONAR trial is an example of a randomized dynamic biomarker-stratified design in the area of diabetes and nephropathy.6 The trial used an open-label active run-in period to assess the change in albuminuria during 6-week treatment with the selective endothelin receptor antagonist atrasentan. Patients who responded to atrasentan, defined as a urinary albumin-to-creatinine ratio reduction ≥30%, were randomized to continue atrasentan or to switch to placebo. This subgroup comprised the primary intention-to-treat population for assessing the safety and efficacy of atrasentan. The trial also enrolled a selection of patients with a <30% urinary albumin-to-creatinine ratio reduction to atrasentan to determine the long-term effect of atrasentan in these nonresponders. The result of the trial showed that atrasentan reduced the clinical kidney outcomes both in responders and nonresponders (hazard ratio, 0.65 [95% confidence interval, 0.49 to 0.88] and 0.75 [95% confidence interval, 0.55 to 1.03] P-interaction = 0.41). This suggests that the benefit of atrasentan would be independent of the degree of albuminuria lowering during the 6-week enrichment period. This unexpected finding led to further analyses that suggested that the within-individual random variation in albuminuria made it difficult to separate responders and nonresponders. A regression to the mean type of effect, which hampered correct stratification of responders and nonresponders, and a legacy effect may have contributed as well.8 The SONAR trial highlights the need for careful considerations in the design stage of a biomarker-stratified trial, particularly regarding the analytical and biological aspects of the stratification approach.
A more conservative approach, which does not solely rely on the biomarker-positive subgroup, is the fall-back randomized biomarker-stratified design. Under this design, the biomarker-positive and biomarker-negative subgroups are both randomized to active or control intervention. Subsequently, the effect of the intervention is first tested in the overall population. If the test is statistically significant, the intervention is effective in the overall population; however, if the test is NS, the effect of the intervention is subsequently assessed in the biomarker-positive patients at an reduced significance level. This clinical trial design has not been used in nephrology yet.
Master Protocols: Basket, Umbrella, and Platform Trials
Basket, umbrella, and platform trials are new clinical trial designs that originated from the oncology area and can be applied to personalized medicine of chronic diseases. These designs use master protocols that define key design features, including adaptive elements and Bayesian approaches, which can be used to assess treatment effects in targeted patient subgroups. The basket trial evaluates targeted therapy on multiple disease types that share the mechanism targeted by the intervention. Umbrella trials evaluate multiple targeted therapies for the same disease entity. A platform trial with a multiarm multistage design compares several intervention groups with a common control.
NEPTUNE-Match is an example of an umbrella trial in nephrotic syndrome. Within this public–private collaboration, noninvasive biomarker profiles have been developed to identify patients with distinct mechanisms of kidney injury. Patients with their biomarker profiles are subsequently matched with molecular targeted therapies to align the right therapy to the right patient.7 The NEPTUNE-Match platform allows for optimal contribution of the patients with rare disease to the best matching clinical trials. A critical prerequisite is the availability of a trial network recruiting a broad patient population willing to participate in the trial-match process. An example of a platform trial in CKD is the Global Kidney Patient Trial Network and CAPTIVATE trial (ClinicalTrials.gov, NCT06058585). The trial is designed to test multiple treatments within a common research platform (the Global Kidney Patient Trials Network registry) and includes novel design elements to tailor optimal therapy.
Future Developments and Research Recommendations
While incorporation of personalized medicine approaches in clinical trial designs has found its way in nephrology, there is still much to learn and to improve.
A better understanding of the cellular processes defining the molecular mechanism that contribute to CKD progression in individual patients is needed. Kidney biopsies will be instrumental in this respect.9
Noninvasive biomarkers are needed that capture the CKD pathophysiology and can be linked to a drug's mechanism of action. Significant progress has been made with data and samples from large clinical trials now becoming available.10
Knowledge about analytical chemistry aspects of the identified noninvasive biomarkers for patient stratification, such as day-to-day variability, is a key element to avoid some of the issues experienced in the SONAR trial. The optimal threshold to define biomarker-positive and biomarker-negative subgroups is another important but under-recognized area.11
Information about event rates in biomarker subgroups is vital, since the PRIORITY and SONAR trials inform us that these may be different from what can be expected.5,6
Effective communication and implementation strategies have to be developed. Involving patients in the trial design stage is critical to achieve this. Implementation aspects of introducing clinical trial designs for personalized medicine are studied in the Personalized Response Implementation and Evaluation in CKD program (www.prime-ckd.com).
With the introduction of multiple new therapies for the management of CKD, clinicians are offered the possibility to tailor optimal therapy for each patient. Several clinical trial designs are available to generate high-level evidence on how to best tailor these new therapies to further improve patients' prognosis. It is the responsibility of the nephrology community to initiate and execute these trials to move from a one-size-fits-all to one-fit-for-everyone approach.
Acknowledgments
The views and opinions expressed are those of the author(s) only and do not necessarily reflect those of the European Union. Neither the European Union nor the granting authority can be held responsible for them.
Disclosures
Disclosure forms, as provided by each author, are available with the online version of the article at http://links.lww.com/JSN/E702.
Funding
This editorial was written in the context of the European Union's HORIZON Research and Innovation Actions (HORIZON-HLTH-2022-TOOL-11-01 - Tools and technologies for a healthy society) undertaking under grant agreement No. 101095146. This project is funded by the European Union.
Author Contributions
Conceptualization: Hiddo J.L. Heerspink.
Writing – original draft: Hiddo J.L. Heerspink.
Writing – review & editing: Matthias Kretzler.
References
- 1.Reznichenko A Nair V Eddy S, et al. Unbiased kidney-centric molecular categorization of chronic kidney disease as a step towards precision medicine. Kidney Int. 2024;105(6):1263–1278. doi: 10.1016/j.kint.2024.01.012 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Boehringer Ingelheim. Aldosterone Synthase Inhibitor on Top of Empagliflozin Delivers Promising Results in Phase II Trial. 2023. Accessed May 26, 2024. https://www.boehringer-ingelheim.com/human-health/chronic-kidney-disease/promising-phase-ii-results-chronic-kidney-disease [Google Scholar]
- 3.Curovic VR Jongs N Kroonen MYAM, et al. Optimization of albuminuria-lowering treatment in diabetes by crossover rotation to four different drug classes: a randomized crossover trial. Diabetes Care. 2023;46(3):593–601. doi: 10.2337/dc22-1699 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Sundstrom J Lind L Nowrouzi S, et al. Heterogeneity in blood pressure response to 4 antihypertensive drugs: a randomized clinical trial. JAMA. 2023;329(14):1160–1169. doi: 10.1001/jama.2023.3322 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Tofte N Lindhardt M Adamova K, et al. Early detection of diabetic kidney disease by urinary proteomics and subsequent intervention with spironolactone to delay progression (PRIORITY): a prospective observational study and embedded randomised placebo-controlled trial. Lancet Diabetes Endocrinol. 2020;8(4):301–312. doi: 10.1016/S2213-8587(20)30026-7 [DOI] [PubMed] [Google Scholar]
- 6.Heerspink HJL Parving HH Andress DL, et al. Atrasentan and renal events in patients with type 2 diabetes and chronic kidney disease (SONAR): a double-blind, randomised, placebo-controlled trial. Lancet. 2019;393(10184):1937–1947. doi: 10.1016/S0140-6736(19)30772-X [DOI] [PubMed] [Google Scholar]
- 7.Trachtman H Desmond H Williams AL, et al. Rationale and design of the Nephrotic Syndrome Study Network (NEPTUNE) Match in glomerular diseases: designing the right trial for the right patient, today. Kidney Int. 2024;105(2):218–230. doi: 10.1016/j.kint.2023.11.018 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Heerspink HJL Xie D Bakris G, et al. Early response in albuminuria and long-term kidney protection during treatment with an endothelin receptor antagonist: a prespecified analysis from the SONAR trial. J Am Soc Nephrol. 2021;32(11):2900–2911. doi: 10.1681/ASN.2021030391 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Lake BB Menon R Winfree S, et al. An atlas of healthy and injured cell states and niches in the human kidney. Nature. 2023;619(7970):585–594. doi: 10.1038/s41586-023-05769-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Sen T Ju W Nair V, et al. Sodium glucose co-transporter 2 inhibition increases epidermal growth factor expression and improves outcomes in patients with type 2 diabetes. Kidney Int. 2023;104(4):828–839. doi: 10.1016/j.kint.2023.07.007 [DOI] [PubMed] [Google Scholar]
- 11.Bakker E, Starokozhko V, Kraaijvanger JWM, Heerspink HJL, Mol PGM. Precision medicine in regulatory decision making: biomarkers used for patient selection in European Public Assessment Reports from 2018 to 2020. Clin Transl Sci. 2023;16(11):2394–2412. doi: 10.1111/cts.13641 [DOI] [PMC free article] [PubMed] [Google Scholar]