Computational Drug Repositioning in Cardiorenal Disease: Opportunities, Challenges, and Approaches

Abstract

There is currently increased interest in drug repositioning programs, namely the identification of new therapeutic areas for already approved drugs, both in academia as well as in the biotech and pharmaceutical industry. Since 2012, the number of publications indexed in MEDLINE on drug repositioning or drug repurposing is exponentially increasing with a peak in the year 2021 due to the worldwide search for therapeutic options to combat the COVID-19 pandemic [1]. Drug repositioning, however, is not new, and pharmaceutical companies have ever since been looking for additional market opportunities for their products, in particular when patents expire and generics manufacturers enter the market of the therapeutic areas of initial approvals [2, 3]. In the pharma world, the term indication expansion is also often used instead of drug repositioning or drug repurposing. In particular for patients suffering from a rare disease who are lacking any approved therapies, drug repositioning represents a very interesting and efficient way of bringing new treatment options to the patient fast [4]. This has also been stressed in a recent position paper from the International Rare Disease Research Consortium [5].

Several international consortia have recognized the trend toward drug repurposing. Two US-based endeavors focusing on drug repositioning are the Drug Repurposing Hub as well as EveryCure. Researchers from the Broad Institute have created the Drug Repurposing Hub with the aim to construct and curate a library of FDA approved drugs that can be used for systematic drug repositioning screenings [6]. EveryCure's mission is to identify novel treatment options for patients with rare diseases via computational drug repositioning. Two European initiatives in the context of drug repositioning are the Repo4EU (https://repo4.eu/) and the REMEDi4ALL (https://remedi4all.org/) consortia, both being public–private partnerships with the aim to develop tools for computational drug repositioning but to also apply these tools and develop novel therapeutic options for selected indications. Next to the worldwide drug repositioning effort in the context of COVID-19, there are at least three additional reasons why drug repositioning programs are gaining momentum. First, the molecular characterization of disease processes is continuously improving, and we understand more about key molecular pathways and disease-modifying proteins, forming the basis to find drugs counterbalancing these dysregulations on the molecular level. Second, the arsenal of computational tools, methods, and workflows is getting better at matching disease pathobiology and drug mechanism of action (MoA), identifying novel connections and thus potential targets for therapeutic intervention. And third, the list of successful repositioning cases is getting longer. Even the current blockbuster drugs of GLP1 agonists can be seen as positive drug repositioning examples, both scientifically and commercially. Initially being developed for the treatment of diabetes mellitus, drugs from this compound class are in the meantime also being approved for the treatment of obesity and are in clinical development for several indications across different therapeutic areas.

In this viewpoint article, we will discuss (i) computational and experimental approaches to discover repositioning opportunities, (ii) challenges in the further development of the discovered compounds, and (iii) repositioning approaches in the context of kidney and cardiovascular disease (CVD).

Next to observation-driven drug repositioning in the context of clinical trials or clinical practice as well as experimental methods such as binding assays or experimental phenotype screens, several computational methods and approaches have been developed to systematically search for new drug repositioning opportunities. These computational approaches make use of information on direct drug targets, affected molecular pathways and biological mechanisms, drug side effects, omics signatures on disease pathobiology and drug mechanism of action, but also on data from clinical trials or electronic health records (EHRs) from patient registries [7-10]. Key computational and experimental approaches as given in Figure 1 will be discussed in the following sections.

Repositioning is often described as a shortcut to reduce the time to market for a drug and drastically decrease the development costs. Widely used figures in the drug discovery area report that it takes 10000 candidates to start with to end up with 1 drug on the market and that this process takes on average 10–17 years and comes with a mean price tag between 1.6 and 2.8 billion USD [69, 70]. The exact savings achieved by drug repurposing in time, risk, and money can be unclear, with some conflicting evidence. Some reviews suggest that about 30% of repurposing efforts are successful and lead to a product approved for marketing, compared to about 10% for new drug applications (NDAs) in general, while other studies argue that repurposed drugs do not necessarily have higher success rates than new drugs, with efficacy often being the limiting factor rather than safety [71]. Once a candidate compound for repurposing has been discovered, the path to clinical utilization and marketing remains a cost-intensive challenge, often referred to as the “valley of death” between basic and clinical research [72]. Numerous experts state that the benefit of repurposing lies in the availability of an established safety profile and that for these compounds preclinical animal models to test safety and even clinical (safety) trials may be skipped up to Phase II or even Phase III. There may be criteria, based on which such a shortcut seems straightforward and even obvious, such as: the drug was shown to be safe in multiple human studies, whereas the animal model does not adequately recapitulate the human disease for which there is substantial unmet medical need, data from in vitro or in silico experiments are supportive, and dosing and administration are consistent with prior human experience for the therapeutic agent of interest [73]. However, in many cases either funders or regulatory authorities may still require the generation of partly redundant data, thereby reducing potential cost savings. Major steps and key concepts in the drug repositioning development path are shown in Figure 2.

A significant factor along this way is patentability of the repurposed drug candidates which is crucial to ensure market exclusivity and protect intellectual property. Only when some form of patent protection or market exclusivity can be guaranteed, the costs of providing the necessary preclinical and clinical data to obtain marketing authorization can be earned back. De Visser et al. discuss the challenge of pricing in drug repositioning, with some compounds having the opportunity of gaining a monopoly leading to exorbitantly high prices as seen for colchicine in the US whereas other compounds are hardly reimbursed to an extent that makes it attractive for companies to invest in drug repurposing in the first place [74]. The authors further advocate government policies to adapt the regulations of appropriate and/or exclusive reimbursement to make drug repurposing more attractive and more predictable for companies. Although better definitions and appropriate pricing models may be introduced by regulators in the future to avoid such “hijacking” while providing a predictable case for companies to (co)-invest in drug repurposing, currently protection is the most promising way to secure a business case for repurposing. Without any incentives, compounds, which could be potentially repurposed, usually do not make it past the academic exercise of identifying their potential alternative use. Although novel molecules, so-called new chemical entities, are typically patented on the substance level, repurposed drugs more often rely on the protection of the product (formulation, composition, route of administration, etc.), the method of use (indication), or a combination of two or three of the before mentioned categories. Second medical use patents, sometimes also referred to as Swiss-type patents, typically involve claims like “substance X for the treatment of condition Y”. Novelty and non-obviousness are the two main criteria by which a repurposed drug is deemed patentable [75], leading to the situation that many potential therapies which would rely on a repurposed compound are mentioned—sometimes even in a speculative fashion—in research articles and are thus either obvious to the expert or at least not novel anymore. According to the current situation regarding necessary IP protection to cover the costs of developing these drugs, these compounds are basically burned ground and will (excluding self-medication or supported by non-profit programs) likely never make it to the patient.

Additional concepts may help especially in the case of repurposing. Formulation patents can be obtained for new formulations of existing drugs, such as extended-release versions or new delivery mechanisms. Combination patents can be filed for new combinations of existing drugs that provide a synergistic effect. If the repurposed drug involves a new chemical entity or a novel combination of active pharmaceutical ingredients, even a new composition of matter patent can be issued. In some regions, supplementary protection certificates (SPCs) can extend the patent life of a drug beyond the usual term, providing additional market exclusivity [76], which is especially attractive for the current owner or manufacturer of a drug to perform indication expansion as part of the business strategy to explore new markets, or as part of the life-cycle management of a drug product. Libraries of compounds that are available for repurposing can be divided into immediately available “on-market” drugs which are either still under active patent or exclusivity protection (“on-patent” drugs) or where this protection has expired (“off-patent”). Off-patent drugs are the prime group for repurposing efforts, after which they at some point become “off-market” drugs that have been discontinued, in some cases due to safety concerns, rendering them less ideal for repurposing [77]. For drugs repurposed to treat rare diseases, obtaining orphan drug status can provide market exclusivity for a certain period, along with other incentives.

The Orphan Drug Designation Pathway is designed to encourage the development of drugs for rare diseases, which affect a small percentage of the population [78]. To be eligible, the drug must be intended to treat a rare disease or condition affecting fewer than 200,000 people in the U.S., or it must be unlikely to recover the costs of development and marketing [79]. Additionally, the drug must provide a significant benefit over existing treatments. Incentives for this pathway include 7 years of exclusive marketing rights in the U.S. after approval, tax credits up to 25% of the clinical trial costs, funding for clinical trials, and exemption from certain FDA fees.

The 505(b)(2) pathway is a streamlined process for NDAs that allows for the use of existing data [https://www.fda.gov/media/156350/download]. It is also eligible to rely on data not developed by the applicant for the application, such as published literature or studies conducted by other entities. This pathway is used for drugs that are modifications of existing drugs, such as new formulations, combinations, or new indications. Incentives include reduced development time by leveraging existing data, which can significantly shorten the development timeline, and cost savings due to the reduced need for extensive clinical trials.

Although the Orphan Drug Designation is specifically for rare diseases and is not limited to repurposed compounds, the 505(b)(2) pathway is for any drug that can leverage existing data, thus focusing on repurposed drugs. Orphan drugs receive specific incentives like market exclusivity and tax credits, which are not inherent to the 505(b)(2) pathway. Both pathways involve rigorous FDA review processes to ensure safety and efficacy, and both can utilize existing data to support the application, though the 505(b)(2) pathway explicitly allows for this reliance. These pathways provide valuable mechanisms to bring repurposed drugs to market more efficiently, addressing both rare and common conditions.

As part of the IP protection efforts, or due to specific requirements of the alternative indication, product maturation involves optimizing the repurposed drug's formulation, dosing, and delivery to enhance its efficacy, safety, and marketability [80]. Re-dosing involves optimizing the dosage to achieve the best therapeutic effect with minimal side effects, which may require new clinical trials to determine the optimal dose for the new indication. It also includes developing new dosing schedules, such as once-daily or extended-release formulations, to improve patient compliance and outcomes. Re-formulation focuses on creating new formulations, such as oral, injectable, or transdermal, to improve drug delivery and patient convenience. It also involves reformulating the drug to enhance its stability, shelf-life, and bioavailability. Drug combinations aim to achieve synergistic effects by combining the repurposed drug with other drugs, enhancing therapeutic outcomes [81]. This also includes developing fixed-dose combination products that simplify treatment regimens and improve adherence [82]. Companion diagnostics are developed to identify patients who are most likely to benefit from the repurposed drug, enabling personalized medicine. This involves identifying biomarkers that can predict response to the drug, allowing for more targeted and effective treatments. Challenges include the requirement for regulatory approval of each new formulation, dosing regimen, or combination, which may involve new clinical trials and data submissions. Finally, ensuring market acceptance involves gaining the approval and adoption of new formulations and combinations by healthcare providers and patients [4].

For almost two decades, antihypertensive medication has been the only treatment regimen with renoprotective effects. This has significantly changed in the last few years with new drugs from different drug classes showing good results regarding renal but also cardiovascular outcomes. These mainly include sodium glucose cotransporter-2 inhibitors (SGLT2i), non-steroidal mineralocorticoid receptor antagonists (MRAs), selective endothelin receptor antagonists (ERAs), and also glucagon-like-peptide 1 receptor agonists (GLP1RAs) [83, 84]. All these drugs can be considered as examples for drug repositioning or indication expansion as they have initially been developed for other diseases, in the case of SGLT2i and GLP1RAs diabetes mellitus. Current research efforts are focusing on identifying responders for these new drugs to optimize treatment and update therapy guidelines in the context of cardiorenal disease. A better understanding of disease pathophysiology and the identification of predictive clinical and molecular markers are essential for this task. A better understanding of the molecular mechanisms of disease is also crucial for identifying novel therapeutic targets as well as additional novel treatment options.

The finding that the JAK/STAT signaling pathway is activated in diabetic kidney disease (DKD) progression based on the analysis of omics data for example has led to the investigation of the JAK inhibitor baricitinib. Baricitinib beneficially impacted albuminuria levels in patients with type 2 diabetes and DKD in a Phase II clinical trial [85]. Baricitinib in addition reduced levels of inflammatory markers such as CCL2, TNFR1/2, ICAM1, or serum amyloid A, however, leading to a higher number of patients experiencing episodes of anemia. It is unclear whether this was the primary reason why no follow-up studies have been conducted for baricitinib in DKD. Anti-inflammatory drugs have also been identified as potential novel therapeutic options for DKD in a transcriptomics signature-based drug repositioning approach by Klein et al. making use of drug expression profiles being available in the Connectivity Map [86]. Dimethylaminoparthenolide, a water-soluble analogue of parthenolide, has been selected for preclinical validation in vivo and had a beneficial impact on the degree of glomerulosclerosis and tubulointerstitial fibrosis. Connectivity mapping has also been used to identify compounds of interest to tackle glomerulopathies. Chung et al. consolidated publicly available transcriptomics signatures including data from focal segmental glomerulosclerosis (FSGS) patients, minimal change disease patients, and IgA nephropathy patients and identified a set of drugs beneficially interfering with endoplasmic reticulum stress and unfolded protein response, mechanisms that they have identified as relevant in disease progression. The EGFR inhibitor neratinib was found to be cytoprotective in a glomerular cell culture model resembling the in-vivo situation of glomerular damage [87]. Other drug repositioning studies identified AZD5438, a CDK2 kinase inhibitor, for the treatment of cisplatin-induced acute kidney injury (AKI) [88] or compounds interfering with molecular mechanisms in the context of nephropathic cystinosis [89]. Interestingly, autosomal-dominant polycystic kidney disease (ADPKD) is a prime target indication for drug repositioning approaches in the field of nephrology. This is probably due to the fact that the only currently approved drug tolvaptan, a selective vasopressin V2 receptor antagonist, cannot fully halt disease progression and is also associated with significant side effects. The high medical need to find and develop novel therapeutic strategies for ADPKD is imminent. Both experimental screening approaches [64, 90] and computational drug repositioning methods [91, 92] have been performed in recent years to identify compounds beneficially interfering with key molecular pathways in the development and progression of ADPKD. Key molecular mechanisms and potential drug targets of ADPKD have recently also been reviewed by Zhou and Torres [93]. This review on drug repurposing in ADPKD is in fact one of the first manuscripts that has been published in the special section on drug repurposing published by Kidney International [94]. Other publications in this series address drug repositioning candidates in the context of podocyte dysfunction [95], therapeutic options for proximal tubulopathies [96], as well as small molecules for the treatment of nephronophthisis and related renal ciliopathies [97]. Despite the fact that long lists of potential treatment options are presented for individual renal diseases, not all will make their way to the stage of being tested in clinical trials or all the way to the clinic.

We have recently identified clopidogrel as a promising therapeutic option for patients with FSGS following a computational network-based drug repositioning approach and subsequent preclinical validation [98, 99]. Clopidogrel significantly reduced proteinuria levels in the adriamycin mouse model for FSGS and also ameliorated histopathological damage in renal tissue. Due to these positive in vivo data and clopidogrel's favorable safety profile, it appears as an attractive option for testing in human clinical trials. We have therefore set up the ClopiD4FSGS clinical trial to determine whether the effect that has been observed in the preclinical setting also holds true in the human setting [100]. Clopidogrel is indicated for a number of CVDs and specifically approved for atherosclerosis, myocardial infarction, peripheral arterial disease, and stroke.

Next to anti-platelet drugs, other standard-of-care compound classes for patients with CVD include anti-hypertensives, statins, beta-adrenergic blockers, calcium channel blockers, or cardiac anti-arrhythmic drugs. There are roughly a thousand different CVD entities within the MeSH ontology and for a large number of diseases there are still no effective treatments available. Lal et al. focused on atrial fibrillation, a CVD with poor treatment options, and used a transcriptomics-based systems biology approach to computationally screen for drugs beneficially interfering with dysregulated molecular processes in atrial fibrillation. The anti-diabetic medication metformin was among the top compound hits and was subsequently validated in the preclinical setting [101]. Wu et al. also made use of transcriptomics data to identify novel treatment options for patients with hyperlipidemia and hypertension. The strength of this study comes from the fact that they were able to assess the impact of top-ranked compounds on LDL cholesterol levels, a marker for hyperlipidemia, and systolic blood pressure, a marker for hypertension, using EHR data from two large cohort studies [102]. Currently approved drugs for the treatment of hyperlipidemia and hypertension were found to significantly reduce the two parameters and served as positive controls for this approach. A number of new compounds beneficially impacting these two parameters were in addition identified. Although it was not a prospectively planned clinical trial, the approach nevertheless is a very efficient way to identify compounds with beneficial impacts on relevant parameters in the human setting of CVD. Another compound that was recently identified with a beneficial impact on blood pressure levels is 5-aminosalicylic acid. This compound positively influences gut energy metabolism and microbiota dysbiosis, two processes that have previously been linked to hypertension on a mechanistic level [103]. Last but not least, there is the aforementioned case of colchicine that was initially developed for the treatment of gout and is currently approved in the US for CVD. A thorough review of drug repurposing options in the context of CVD was recently published by Abdelsayed et al. [104] with Ghosh et al. discussing drug repurposing opportunities in the context of stroke intervention with a focus on compounds in clinical testing [105].

But there is also the other way round, that is, the repositioning of drugs that are approved for CVDs for other indications outside the cardiovascular space. Tripathi et al., for example, discuss the role of statins as anti-cancer therapeutics due to their effect on apoptotic processes via the BCL2 signaling cascade which subsequently has an impact on p53 signaling, mechanisms that are often dysregulated in the context of tumor development and progression [106]. Cancer is also one of the therapeutic areas that is discussed as an indication field for certain CVD compounds in the review by Ishida et al. [107]. Other indications in which CVD drugs are evaluated are, for example, cirrhosis, hemangioma, osteoporosis, Marfan's syndrome, or certain kidney diseases.

Drug repositioning is a hot topic at the moment and the list of positive drug repositioning cases is getting longer. The systematic identification of drug repositioning opportunities, however, is still a young research field and it has become evident that bringing a repositioned product to the market involves much more than just the initial discovery. Whereas academic research groups are key stakeholders in developing new methods or generating new datasets that lead to a better understanding of disease pathobiology and drug mechanism of action, thus forming the basis for the identification of novel repositioning opportunities, they often lack the capabilities of further drug development. Big pharma companies, on the other hand, are primarily focusing on their own products whilst often missing out on opportunities outside their core business areas. In the end, it may be public–private partnerships, small biotech, and “techbio” companies who are the innovators in the drug repositioning space, coming up with novel opportunities and the capabilities for further development. After the positive completion of Phase II clinical trials, pharmaceutical companies are typically in-licensing these assets to run the pivotal Phase III clinical trials and bring the rediscovered products to the market.

Despite some incentives from the regulatory perspective like the Orphan Drug Designation pathway or the 505(b)(2) pathway, some hurdles remain to make drug repositioning programs even more attractive. There is, for example, still a lack of funding for Phase II clinical trials for repurposed compounds as this is often considered not innovative enough. Starting a Phase II clinical trial is thus associated with a significant risk of failure. Having access to large population datasets for in-silico validation of initial repurposed drug candidates would help to further reduce the risk as exemplified in the context of hyperlipidemia and hypertension by Wu et al. who had access to two large cohorts for retrospective data analysis and data validation [102]. Maturation of repurposed drug candidates, that is, re-dosing or re-formulation of existing drugs for the new indication is often necessary, both from the perspective of efficacy as well as from the IP and business perspective to bring a patent-protected product to the market. The development of drug combinations is a very attractive way to bring highly innovative and effective products to the market and leverage a certain period of market exclusivity [108].

Overall, for people suffering from a rare disease, drug repositioning is the option with the highest chance for a treatment becoming available within a reasonable time. With the growing number of successful drug repositioning programs, we strongly believe that this is not merely a passing trend but a viable concept for the long-term to bring new therapies to the market in a fast and efficient way.

P.P. conceptualized and planned this viewpoint article. P.P., M.L., K.K.I, and K.K. wrote the first draft of the manuscript. All authors contributed input, reviewed, and edited the manuscript. All authors approved the final draft.

Paul Perco, Matthias Ley, Kinga Kęska-Izworska, Dorota Wojenska, Enrico Bono, Samuel M. Walter, and Lucas Fillinger are employees of Delta 4 GmbH. Klaus Kratochwill is co-founder of Delta4 GmbH.