University of Groningen Better prediction of drug response in diabetic kidney disease Idzerda, Nienke

University of Groningen

Better prediction of drug response in diabetic kidney disease

Idzerda, Nienke

DOI:

10.33612/diss.113117223

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.

Document Version

Publisher's PDF, also known as Version of record

Publication date: 2020

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):

Idzerda, N. (2020). Better prediction of drug response in diabetic kidney disease: a biomarker approach to personalize therapy. Rijksuniversiteit Groningen. https://doi.org/10.33612/diss.113117223

Copyright

Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policy

If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.

7

123

7

Summary

The overall premise of this thesis is that biomarker-based approaches can be utilized to tailor pharmacological therapy in diabetic kidney dis-ease. Diabetic kidney disease (DKD) is a heterogeneous disease with a complex underlying pathophysiology. Prior studies have shown that patients with DKD show a large variability in response to drug ther-apy. Analyses from completed large-scale clinical trials revealed that this individual drug response variability explains in part why several inter-ventions tested in these large trials did not result in additional renal or cardiovascular protection.[1–3] These findings emphasize the need to develop personalized strategies that take into account individual re-sponse variability in order to allocate drug treatment to individuals who will more likely benefit most from it. It has recently been shown that it may be feasible to use short-term changes in biomarkers to enrich clinical trials; the success of this still needs to be evaluated.[4] More re-search is also needed to identify biomarkers that can predict long-term efficacy of different drugs in patients with DKD.

The aim of this thesis was to explore whether (changes in) routinely measured biomarkers before drug exposure or during a short-term pe-riod of drug treatment predict individual drug response on long-term renal outcomes. Chapter 2 and 3 evaluated the use of single predictive and dynamic biomarkers to predict drug effects on long-term outcomes. In the subsequent chapters, we explored the utility of a previously devel-oped model which integrates short-term effects on multiple biomarkers to predict the long-term impact of different novel pharmacological in-terventions on renal disease progression.

In Chapter 2, a post hoc analysis was performed among 5081 pa-tients with type 2 diabetes mellitus to investigate whether NT-proBNP, a marker of vascular wall stress and fluid overload, can predict the effects of additional therapy with the direct renin inhibitor aliskiren on cardio-renal endpoints. The primary results of the trial showed that on a pop-ulation level, aliskiren did not confer renal or cardiovascular protection. The current study showed that the relative risks for cardiovascular and renal events achieved with additional aliskiren therapy were modified by baseline NT-proBNP levels. At higher NT-proBNP levels, aliskiren compared to placebo increased the risk of the cardiorenal endpoint,

124

Chapter 7

while in the lower NT-proBNP tertile, treatment with aliskiren tended to reduce cardiorenal risk. Effects of aliskiren compared to placebo on safety outcomes (hyperkalemia and hospitalization for acute kidney injury) were independent of NT-proBNP. Earlier studies showed that extracellular volume restriction, by means of moderating dietary so-dium intake or concomitant diuretic treatment improved the efficacy of RAAS blockade to decrease cardiorenal risk. The results of the present study suggest that NT-proBNP can help to identify individuals who may not respond to dual RAAS inhibition with aliskiren and who may benefit from diuretic treatment or dietary sodium lowering.

Chapter 3 presents a study in 471 patients with proteinuria who

participated in the PLANET I and II (Renal effects of Rosuvasta-tin and AtorvastaRosuvasta-tin in Patients Who Have Progressive Renal Dis-ease) trials that examined effects of atorvastatin and rosuvastatin on proteinuria and renal function. The PLANET trials showed that the proteinuria response to rosuvastatin and atorvastatin differ despite a similar response in cholesterol at a population level. In this post hoc analysis we found that proteinuria and cholesterol response from base-line to 14 weeks were not only variable between the statins, but also highly variable between patients for both statins. In addition to this between-patient variability, we observed that a reduction in proteinuria was not accompanied by a cholesterol reduction in a substantial num-ber of patients. The individual responses in proteinuria and cholesterol predicted eGFR decline during the subsequent 9 months of follow-up: a response in both proteinuria and cholesterol was associated with a stable renal function, whereas non-responders in both proteinuria and total cholesterol showed the fastest rate of eGFR decline, independent of the type of statin. These findings suggest that changes in both pro-teinuria and cholesterol should be individually monitored to identify who will benefit from statin therapy.

In Chapter 3, we demonstrated that statins exert multiple effects that do not run in parallel within an individual. These results extend on those from earlier studies, in which the same phenomenon was observed for angiotensin receptor blockers (ARBs).[5–7] Importantly, changes in separate biomarkers were independently associated with long-term renoprotection, implying that multiple short-term drug effects should be taken into account to adequately predict long-term drug response.

125

7

It has been shown that the PRE score, by integrating changes in risk markers after a short period of treatment, can be utilized to predict long-term drug response. Several existing risk scores were developed and validated to predict the risk for renal disease progression based on multiple baseline features. Up to now, it has not been determined whether these scores can also be used to predict response to drug therapy. In Chapter 4, we therefore compared two existing risk scores with the PRE score in their ability to predict long-term renal effects of ARB therapy in patients with type 2 diabetes and kidney disease. While all included scores showed a generally good performance for risk pre-diction at baseline, the PRE score provided more accurate prepre-dictions of long-term ARB effects on renal outcomes than the two existing risk scores. This implies that algorithms developed to predict absolute risk – which do not necessarily include biomarkers that reflect short-term drug effects – may not adequately predict relative benefit conferred by drug therapy. Alternatively, a multiple response score such as the PRE score may assist in better long-term drug efficacy prediction and identify patients who are most likely to benefit from pharmacological intervention. In the last part of this thesis, we assessed the utility of the PRE score for other drug classes that are currently used to reduce cardiovascular and/or renal risk in patients with type 2 diabetes.

In Chapter 5, we performed a secondary analysis on EXSCEL (Ex-enatide Study of Cardiovascular Event Lowering), a placebo-controlled trial investigating the cardiovascular safety of the glucagon-like pep-tide-1 analogue (GLP-1 RA) exenatide in a broad range of patients with type 2 diabetes with or without atherosclerotic cardiovascular dis-ease. Exenatide reduced glycated haemoglobin levels after 6 months treatment, but also significantly lowered several other cardio-renal risk markers, including systolic blood pressure and progression to micro- of macroalbuminuria. By integration of these short-term effects using the PRE score we calculated a substantial relative risk reduction in renal events which was of similar magnitude to the relative risk reduction ob-served in the trial. These results imply that the utility of the PRE score extends to GLP-1 RAs, suggesting that this algorithm may be used as a tool to predict treatment response for this class of drugs. Overall, the results of this study support further clinical trials to prospectively assess the renal efficacy of exenatide.

126

Chapter 7

In Chapter 6 we assessed whether the PRE score could predict the long-term effects of the sodium glucose co-transporter 2 (SGLT2) in-hibitor dapagliflozin on renal and heart failure outcomes. The short-term effects of dapagliflozin on multiple biomarkers were established in patients with type 2 diabetes and chronic kidney disease who partic-ipated in a Phase 3 program of dapagliflozin clinical trials. Integration of the observed short-term drug effects by the PRE score indicated that treatment with dapagliflozin would confer considerable improve-ments in kidney and heart failure outcomes in these patients. The pre-dicted risk reductions on dapagliflozin were very similar to the long-term effects of SGLT2 inhibition as observed in recent cardiovascular outcome trials in patients with type 2 diabetes at high cardiovascular risk. A large ongoing dapagliflozin outcome trial (DAPA-CKD; NCT 03036150) will provide a more definitive answer as to whether the pre-dicted effects of dapagliflozin in patients with chronic kidney disease are accurate and will thus validate or refute these predictions.

Future perspectives

In this thesis, we investigated strategies based on predictive and dynamic biomarkers to tailor pharmacological therapy in patients with DKD. Uti-lizing the PRE score, which translates short-term changes in biomarkers into a predicted long-term drug response, appears to be a promising strategy to personalize drug therapy. The PRE score may be used in clin-ical practice to monitor individual drug response and guide treatment decisions. Additionally, it may be applied in future clinical trials to es-timate treatment effects on renal outcomes during early stages of drug development that can support the decision as to whether the drug should be tested in phase III clinical trials. Finally, the PRE score can aid en-richment of clinical trials. During an enen-richment phase, patients can be selected based on estimated drug effects on multiple risk markers calcu-lated by the PRE score. This approach enables the inclusion of patients who are likely to respond to the drug, as well as the exclusion of patients who are prone to experience harmful effects of the new intervention.

What is needed to integrate the PRE score in future practice? The PRE score adequately predicts long-term drug response on a population

127

7

level for several drug classes. Prospective, pre-specified trials in which the PRE score is used to monitor and adjust pharmacological therapy in clinical practice are required to test whether the PRE score can also be used to improve patient outcomes on an individual patient level. In such a study patients would be randomly assigned to a PRE score guided treatment regimen and care as usual and be followed for long-term clinical outcomes. Treatment recommendations would be given in both treatment arms according to clinical practice guidelines but the PRE score guided treatment arm will be informed by an individuals predicted drug response based on changes in multiple biomarkers. The intricacies of this type of study would have to be well thought through before trial implementation, e.g. the selection of clinically relevant end-points with sufficient level of evidence and whether to randomize par-ticipants at the practice or patient level. Such prospective studies will provide evidence necessary for the PRE score to be used for optimally informing treatment decisions.

Implementation of biomarker-based response scores such as the PRE score is only one aspect of personalized medicine. Personalized treat-ment encompasses many other facets such as advancing treattreat-ment ad-herence and stimulate shared decision making. Regardless of the dif-ferent aspects of personalized medicine, its implementation requires a multi-faceted approach involving many stakeholders. Different stake-holders have different priorities and focus areas for personalized med-icine. For example, regulatory agencies would have to develop models to assess efficacy and safety, and pharmaceutical companies would have to market drugs for specifically targeted patient populations. Health-care providers would focus on the development of new guidelines and implementation of biomarker based approaches in clinical practice and clinicians would require education in the topic of personalized medicine, e.g. on which diagnostics and drugs to test and prescribe for specific patient populations. Importantly, the implementation of personalized treatment in DKD would require close involvement of patients since they are the most important end-users of healthcare.

This thesis was conducted under the umbrella of the BEAt-DKD project. One of the important aims of BEAt-DKD consortium is to promote implementation of personalized medicine. BEAt-DKD works towards this by supporting stakeholder engagement and collaboration.

128

Chapter 7

Discussions with multiple stakeholders are ongoing to identify and eventually align different interests, to overcome obstacles, and to strate-gize solutions to move personalized medicine forward in DKD. For in-stance, patients and patient organizations support the implementation of a drug response score such as the PRE score since it may provide ac-cess to individualized therapy that improves quality of life. Health care providers would be generally willing to use the PRE score in clinical practice as it would support individualized, evidence based treatment decisions. However, models of therapy response must be translated in reliable and efficient applications for clinical practice. Regulatory agen-cies and health policy makers require prospective clinical trial evidence demonstrating that implementation of the PRE score results in addi-tional benefit as compared to existing treatment decision tools in terms of gain in healthy patient years. These ongoing joint efforts to reach a consensus on the implementation of personalized methodologies such as the PRE score are essential to translate personalized treatment in patients with DKD from research to clinical practice.

References

1. Heerspink HJ, Ninomiya T, Persson F, Brenner BM, Brunel P, Chaturvedi N

et al. Is a reduction in albuminuria

as-sociated with renal and cardiovascu-lar protection? A post hoc analysis of the ALTITUDE trial. Diabetes Obes Metab 2016; 18: 169–177.

2. Chin MP, Wrolstad D, Bakris GL, Chertow GM, de Zeeuw D, Golds-berry A et al. Risk factors for heart failure in patients with type 2 diabe-tes mellitus and stage 4 chronic kidney disease treated with bardoxolone me-thyl. J Card Fail 2014; 20: 953–958. 3. Hoekman J, Lambers Heerspink HJ,

Viberti G, Green D, Mann JF, de Zeeuw D. Predictors of congestive heart failure after treatment with an endothelin receptor antagonist. Clin J Am Soc Nephrol 2014; 9: 490–498. 4. Heerspink HJL, Parving HH,

An-dress DL, Bakris G, Correa-Rotter

R, Hou FF et al. Atrasentan and renal events in patients with type 2 diabe-tes and chronic kidney disease (SO-NAR): a double-blind, randomised, placebo-controlled trial. Lancet 2019; 393: 1937–1947.

5. Hellemons ME, Persson F, Bakker SJ, Rossing P, Parving HH, De Zeeuw D et al. initial angiotensin receptor blockade-induced decrease in albu-minuria is associated with long-term renal outcome in type 2 diabetic pa-tients with microalbuminuria: a post hoc analysis of the IRMA−2 trial. Di-abetes Care 2011; 34: 2078–2083. 6. Eijkelkamp WB, Zhang Z, Remuzzi G,

Parving HH, Cooper ME, Keane WF

et al. Albuminuria is a target for

reno-protective therapy independent from blood pressure in patients with type 2 diabetic nephropathy: post hoc analy-sis from the Reduction of Endpoints

129

7

in NIDDM with the Angiotensin II Antagonist Losartan (RENAAL) trial. J Am Soc Nephrol 2007; 18: 1540–1546.

7. Apperloo EM, Pena MJ, de Zeeuw D, Denig P, Heerspink HJL. Individual variability in response to renin angi-otensin aldosterone system inhibition predicts cardiovascular outcome in patients with type 2 diabetes: A pri-mary care cohort study. Diabetes Obes Metab 2018; 20: 1377–1383.