Abstract
Although more than 100 years have passed since Alois Alzheimer reported a case of Alzheimer’s disease (AD), a definitive answer to the causes of cognitive impairment in the disease remains elusive. Despite significant enthusiasm and investment from the pharmaceutical industry, clinical trials of many disease-modifying drugs for AD have been largely unsuccessful. Drug repositioning (DR) or repurposing approaches are relatively inexpensive and more reliable compared to de novo drug development in AD. About 30% of clinical trials for AD in progress around the world use the DR method and hold potential in halting the current deadlock in treatment options. By using drugs approved for other indications, these clinical trials target dysregulated pathways in AD with different or a combination of modes of action, including anti-amyloid, cardiovascular, anti-tau, anti-inflammatory, immunomodulatory, metabolic, neuroprotective, and neurotransmission-based approaches. For instance, anti-diabetic drugs, such as insulin, metformin, liraglutide, and dapagliflozin, and cardiovascular drugs, such as cilostazol, candesartan, telmisartan, prazosin, and dabigatran, could serendipitously provide previously unearthed benefits in AD. This is in line with recent thinking, which views AD as a complex multifactorial disorder, not dominated by one dominant biological factor, such as amyloid-β, and likely a confluence of many pathobiological mechanisms, including vascular dysregulation. Such increasingly available knowledge of phenotyping may be used to design ‘tailor-made’ DR and relatively homogeneous AD subpopulations specifically targeted with existing drugs based on known modes of action. It is thus expected that DR approaches will create a major paradigm shift in AD research and development.
Drug repositioning (DR), also known as drug repurposing, rediscovery, or rescue, refers to the discovery of efficacy in a disease from existing drugs whose safety and pharmacokinetics have already been confirmed in another disease in prior clinical studies, or discontinued drugs repositioned for development. Although DR itself is not a new concept, it has gained attention in recent years because of the slowing in the development of drugs and stricter approval standards. From the initial discovery of a compound, a budget of several tens of millions of dollars and a period of 12–14 years is usually required before clinical application [1]. However, the probability of a new drug being sent to medical practice is thought to be 0.02% to 0.01% [2]. In the current situation, where R&D investment has stagnated in the face of a sluggish world economic outlook [1], cheaper drug development with high certainty, is warranted [3]. The biggest advantage of DR is that the pharmacological and safety profiles are already known. DR eliminates the early phase of clinical development and can therefore significantly reduce development time. Investigational drugs must be manufactured in accordance with good manufacturing practice (GMP) standards, and quality control and reproducibility of production are strictly policed. GMP standards are also required in manufacturing placebo drugs as well. However, the strict regulatory procedures involved in the development of GMP-standard investigational drugs act as a barrier to involvement of academia in all stages of drug discovery research. However, as DR approaches negate safety study requirements and may proceed from Phase II, involvement of academic institutions remains feasible.
Since DR may produce sufficient information on pharmacokinetics and side effects, it may be possible to minimize the risk of failure of clinical trials due to factors relating to pharmacokinetics and safety; new adverse effects are less likely to be unearthed. Many drugs have already been applied clinically via DR [4], and an increasing number of corresponding papers have been published in the neurology field [5]. Success rates of DR have been estimated to be more than twice that of conventional drug discovery [6].
POTENTIAL OF DR IN ALZHEIMER’S DISEASE
In geriatric diseases, such as Alzheimer’s disease (AD), patients frequently have multiple comorbidities, meaning polypharmacy is more apparent than in younger patients [7]. However, an unexpected consequence of this situation has been the serendipitous discovery of potential novel treatment strategies from DR in the daily clinical practice of geriatric diseases. A high incidence of side effects is also characteristic of geriatric diseases [8] so the use of existing drugs with well-established safety profiles represents a lower risk strategy, when compared to newly-developed drugs. Taken together, it is apparent why DR represents a useful approach in the search for new treatments for AD.
Usually, the normal process of drug development takes a total of 12 years or more. The time required for each step is as follows [3, 9]: 2–3 years for target discovery (e.g., bioinformatics, expression analysis) 0.5–1 year for screening and discovery (e.g., combined chemistry) 1–3 years for lead optimization (e.g., rational drug design) 1–2 years in ADMET testing (e.g., bioavailability)
ADMET: absorption, distribution, metabolism, excretion, toxicity [10] 5–6 years for development (e.g., Phase I study) 1–2 years for registration (e.g., US Food and Drug Administration, European Medicines Agency, Japan Agency for Medical Research and Development).
In geriatric diseases such as AD, the regulatory procedures for above-mentioned Phase I study, i.e., safety study, are extremely stringent. However, with DR, clinical trial failure for pharmacokinetic and safety reasons may be minimized as a result of shorter development time of between 3 and 12 years comprising [9]: 1–2 years for compound identification (e.g., new drug identification, special screening) 0–2 years for acquisition of compounds (e.g., license, internal source) 1–6 years for development (e.g., Phase I / II study) 1–2 years for registration (e.g., US Food and Drug Administration, European Medicines Agency, Japan Agency for Medical Research and Development).
PROBLEMS RELATING TO DR IN ALZHEIMER’S DISEASE
Concerns about individual differences
It is well known that pathogenic mechanisms behind AD and other geriatric diseases vary greatly from individual to individual, meaning symptomatic treatment response varies greatly. DR has been frequently performed in vitro and in silico [11] but cannot accurately predict the therapeutic effect of elderly patients, where individual differences are large. Vascular lesions, such as cerebral infarction and hemorrhage, also frequently co-exist with senile plaques and neurofibrillary tangles in AD [12]. Numerous cases of severe AD pathology have been reported postmortem despite an absence of antemortem cognitive impairment [13]. Recent reports suggest closely related neurodegenerative and cerebrovascular diseases with resultant synergistic cognitive dysfunction [14].
There is controversy about the extent to which sporadic disease states, such as in AD, can be reproduced in experimental systems using genetically engineered animals [15]. DR based on drug reprofiling is considered to be an extremely useful technique for rare genetic diseases due to a single pathogenic factor and relatively small individual differences. However, the effectiveness of DR in geriatric diseases such as sporadic AD, which has large individual differences due to multiple factors, is open to debate.
Safety concerns
Patients with AD must invariably receive treatment for a long period of time because of the slow progressive clinical nature of the disease. Therefore, the safety of repositioned drugs for AD may require re-examination. Many drugs targeted for DR were initially marketed as cancer drugs [16]. Additional indications of cimetidine for pulmonary adenocarcinoma, digoxin for prostate cancer, and ritonavir (anti-HIV drug) for ovarian cancer have been found to have novel effects through DR. Unlike geriatric diseases, DR for cancer may have relatively limited safety concerns due to the nature of the disease. Indeed, safety tests for DR agents may not always be omitted in geriatric diseases.
Cost concerns
Even if it is possible to omit initial safety procedures using DR, a clinical trial to prove efficacy is, of course, necessary. AD intervention studies are usually costly, both in time and expense, with a Phase III trial generally costing about 400–500 million USD [17]. Only pharmaceutical companies can apply for regulatory affairs, and their cooperation on clinical applications of DR is required.
Pharmaceutical companies are generally receptive to drug rescue, i.e., drugs still under development, patents remaining or discontinued drugs, but are often reluctant to repurpose drugs with expired patents [18]. A patent right allows patented invention exclusively for a certain period of time (in principle, 20 years from the filing of the application). Pharmaceutical companies recover investment and costs spent on drug development during this time, while obtaining funding for future drug discovery projects. Still, upon first indication, often fewer than 10 years of the 20-year substance patents remain, meaning the next indication approval and marketing for profit with the same drug would be difficult. Pharmaceutical companies may seek several different patents, such as those related to efficient production of chemical substances and formulation for compositions of chemical substances combined with additives. By applying this “Evergreen Strategy”, which seeks to extend the monopoly of pharmaceuticals through staggering applications, the goal is to maximize the benefits of monopoly, though profits may not be guaranteed through the process [19].
Further, even if an existing drug can be applied to a certain disease, it will not be widely sanctioned for use in clinical practice unless proven to possess effects better than other treatments. The exclusive sales period is given only for the reexamination period (post-marketing investigation period) of 3 years in the United States through an Abbreviated New Drug Application (ANDA) [20] and 4 years in Japan, designated by the Pharmaceutical Affairs Law [21]; after this period, generic drugs will automatically accommodate the new indications. In the EU, however, a harmonized regulatory procedure for generic drugs has not yet been established and drug policies differ between European countries, providing obstacles for DR implementation [22]. Even in USA and Japan, the 3 to 4-year period may not tempt pharmaceutical companies to commit to DR as it is extremely difficult to recover the investment within such a short period of time.
Furthermore, even if indications are expanded, drug costs may be reduced based on the price for existing indications. Even if approved as an appropriate drug price, off-label use of cheaper drugs for existing indications is permitted in clinical practice [23], making it harder for companies to realize profit.
PRACTICE OF DR IN ALZHEIMER’S DISEASE
Despite the various difficulties mentioned above, many AD trials have been conducted worldwide using DR techniques, which have many advantages over conventional drug development [24]. We searched clinical trial information on Clinicaltrials.gov as of January 2, 2020, and identified those that used the DR method for drug discovery of AD. DR accounted for approximately 30% of all AD trials currently deployed worldwide [25]. Phase III (Table 1), Phase II (Table 2), and Phase I (Table 3) were listed. Several drugs have been registered as Phase I/II and Phase II/III and have an asterisk in the Clinicaltrials.gov ID. This only covers drug therapy and does not include non-drug therapy or cognitive therapy. Past trials (completed trials) are not included.
Clinical trials (phase III) for Alzheimer’s disease with drug repositioning
GABAA, γ-aminobutyric acid type A. *Phase II/III study.
Clinical trials (phase II) for Alzheimer’s disease with drug repositioning
BACE, β-site amyloid precursor protein cleaving enzyme; HSV, herpes simplex virus.
Clinical trials (phase I) for Alzheimer’s disease with drug repositioning
GABAA, γ-aminobutyric acid type A; mTOR, mechanistic target of rapamycin; PPARγ, peroxisome proliferator-activated receptor γ.
Repositioned drugs can be roughly classified into seven groups (described below) when broadly classified according to their mechanism of action. Some are assumed to have dual or triple mechanisms of action (Fig. 1). Given most AD is sporadic and multifactorial in nature, with the exception of familial cases, drugs with variety of mechanisms of action are desirable [26, 27]. Furthermore, as drug interaction profiles are already known, clinical trials using a combination of drugs, such as the COGNITE study with cromolyn and ibuprofen (Table 1), or the rrAD study with losartan, amlodipine, and atorvastatin (Table 1), may be facilitated more easily, especially when pitted against combination therapy for de novo drugs. [For more details than described in the Table 1–3, go to https://clinicaltrials.gov and enter the ID number starting from NCT in the box of ‘Other terms’].
Targets for drug repositioning in developing new treatments for Alzheimer’s disease. In addition to classical therapies targeting the production and aggregation pathways of amyloid-β (Aβ), therapies directed at the Aβ clearance pathway are attracting attention. Since the clearance efficiency of Aβ decreases due to atherosclerosis and cerebral amyloid angiopathy, the development of a therapeutic method to prevent and improve these vascular conditions with cardiovascular drugs, a neurovascular approach, is desired. In addition, therapies for anti-tau, anti-inflammation, metabolism (e.g., suppression of glucose toxicity), neuroprotection, and improvement of neurotransmission are being developed in recent drug repositioning approaches. BACE, β-site amyloid precursor protein cleaving enzyme; CAA, cerebral amyloid angiopathy.
Anti-amyloid
In line with the Amyloid Hypothesis, drugs expected to suppress the production and aggregation, or promote its clearance, of amyloid-β (Aβ) have been trialed clinically. Examples include the iron chelator deferiprone [28], which may inhibit Aβ aggregation, and some cardiovascular drugs (see below), which may promote Aβ clearance by preserving cerebral blood flow (CBF).
Cardiovascular
Antithrombotic agents, cilostazol [29] and dabigatran [30], have been reported to preserve CBF and promote Aβ clearance. In retrospective analyses, cilostazol was found to suppress cognitive decline in mild cognitive impairment (MCI) or early AD patients whose Mini-Mental State Examination score ranged from 22–26 [31, 32], which led to the COMCID study (Table 2). Several cardiovascular drugs have also been explored for CBF-preserving and Aβ-clearing potential in AD, including three types of angiotensin receptor II blockers, losartan, candesartan, and telmisartan, and angiotensin-converting enzyme (ACE) inhibitor perindopril. The Aβ-clearing potentials of losartan [33], candesartan [34], and telmisartan [35] have been proven in mouse models of Aβ amyloidosis. In a meta-analysis, use of angiotensin receptor II blockers was associated with reduced risk of incident AD (HR 0.72; 95% confidence interval, 0.58–0.88; p < 0.001) [36], providing a clinical background for the rrAD (Table 1), SARTAN-AD and CEDAR (Table 2), and HEART studies (Table 3). In addition, perindopril, a brain-penetrating ACE inhibitor, was shown to be protective against cognitive impairment in mice with intracerebroventricular injection of Aβ, compared to non-brain-penetrating counterparts [37]. In a small clinical study, brain-penetrating ACE inhibitors were found to suppress cognitive decline compared to those not passing the blood-brain barrier in AD patients [38], providing additional context to the SARTAN-AD study (Table 2).
Anti-tau
Given the lack of success in Aβ treatment, drugs targeting tau protein remain highly desirable. DRs using the anticancer drug nilotinib [39], which may suppress tau formation, and nicotinamide [40], which may suppress tau-dependent microtubule depolymerization, have been tested. In addition, the GSK3β inhibitory effect of lithium has been shown to prevent tau phosphorylation in mouse models of tauopathies [41, 42]. Indeed, a systematic review and meta-analysis has suggested lithium treatment possesses beneficial effects on cognitive performance in subjects with MCI and AD [42]. Salsalate, an NSAID shown to suppress tau acetylation, rescued tau-induced memory deficits, and prevented hippocampal atrophy in a tauopathy mouse model [43], consequently raising expectations in tau-based therapy.
Anti-inflammatory and/or immunomodulatory
The relationship between inflammation and AD has long attracted attention [44]. Long-term use of NSAIDs, ibuprofen in particular, were found to be protective against AD in a study from a large healthcare database [45]. In addition, although some controversies still exist [46], a recent meta-analysis from cohort studies showed NSAID exposure may reduce the risk of AD [47]. Therefore, anti-inflammatory drugs such as leukotriene inhibitor montelukast [48], asthma drug cromolyn [49], and NSAID ibuprofen [49] have all been touted as agents against AD, as has the antiviral drug valacyclovir [50] based on the hypothesis that herpes simplex virus may be etiologic or contribute to the pathology of AD. Rifaximin, an antibacterial agent, could ameliorate dysbiosis in the gut and prevent AD progression based on the microbiota-brain-gut axis in AD [51]. In addition, several drugs indicated for multiple myeloma or leukemia, including daratumumab (anti-CD38 antibody) [52], dasatinib (tyrosine kinase inhibitor) [53], lenalidomide (thalidomide analog; TNF-alpha inhibitor) [54], and sargramostim (granulocyte macrophage colony stimulator) [55], have been explored for efficacy in AD based on putative immunomodulatory and disease-modifying effects in cellular or animal models of AD.
Metabolism
The link between AD and diabetes is well established [56], with some declaring AD as ‘type 3 diabetes’ [57]. Therefore, among drugs acting on the metabolic system, the repositioning of antihyperglycemic as anti-AD drugs is currently receiving attention [56]. Such agents include insulin nasal therapy [58], metformin [59], liraglutide (glucagon-like peptide 1 receptor agonists) [60], and dapagliflozin (sodium glucose co-transporter 2 antagonist) [61]. A systematic review of past seven studies of insulin nasal treatment (293 patients) showed improved story recall performance of ApoE ɛ4 (–) patients with MCI or AD [62], suggesting potential for intranasal glulisine as an agent in the treatment of amnestic MCI and probable mild AD patients (Table 2). Other studies have used benfotiamine (a vitamin B1 derivative) or nicotinamide (vitamin B3) as clinical trial drugs because of the potentially positive effects on multiple cellular processes. For instance, benfotiamine has been shown to prevent hyperglycemic damage via multiple pathways [63] and improve cognitive impairment in a mouse model of AD [64]. Nicotinamide inhibited histone deacetylase and phosphorylated tau in a mouse model of AD [65]. Similarly, vorinostat used for cutaneous T-cell lymphoma, also a histone deacetylase inhibitor, may show promise as a therapeutic agent for AD [66]. Interestingly, an anti-retroviral drug efavirenz has been shown to modulate cerebral cholesterol metabolism [67] and thus may prove effective in AD patients supposed to have impaired cholesterol metabolism in the brain [68]. Rapamycin, a lipophilic macrolide antibiotic and well-established inducer of autophagy, can reduce Aβ and tau pathologies [69], and is now under clinical trial for AD [70].
Neuroprotection
Most of the disease-modifying drugs mentioned are thought to result in neuroprotection through the mechanisms described above. In addition, at least five clinical trials are being conducted using levetiracetam, an anti-epilepsy drug that modulates the function of the synaptic vesicle glycoprotein 2A, to suppress neuronal hyperexcitability associated with AD [71, 72]. Intriguingly, the anti-epileptic drug levetiracetam has been reported to suppress silent hippocampal seizures and spikes by foramen ovale electrodes in a case of AD, suggesting effectiveness in suppressing occult hippocampal hyperexcitability contributing to the AD pathogenesis [73]. These scientific and clinical findings have prompted several clinical trials of levetiracetam, such as the LEV-AD, ILiAD, and LAPSE studies (Table 2). Clinical trials have also been conducted using ethyl icosapentate, an ω3 fatty acid important in the normal function of nerve cells [74]. Leuprolide, a luteinizing hormone releasing agonist used for prostate cancer, has been explored for its neuroprotective benefits, based on promising early results [75].
Neurotransmission
Drugs affecting neurotransmission are intended to improve mental symptoms, such as behavioral/psychological symptoms (BPSD), not modify the disease, through effects on neurotransmitters such as acetylcholine, GABA, serotonin, dopamine, and noradrenaline. For instance, several clinical trials have been conducted with brexpiprazole, a serotonin/dopamine activity modulator, for agitation in AD, based on promising results in earlier studies [76]. The serotonin reuptake inhibitor, escitalopram, has been examined for agitation [77] and the dopamine/noradrenaline reuptake inhibitor, methylphenidate, for apathy [78], two BPSD frequently observed in AD patients. Prazosin is also under clinical trial to determine whether agitation can be suppressed by the inhibitory effect of alpha-1 adrenoceptor stimulation [79]. Nicotine transdermal patches have also been suggested as cognitive enhancers through nicotinic acetylcholine receptor enhancement. Zolpidem may improve sleep quality by modulating GABAA receptors and may also suppress aberrant excitatory-inhibitory synaptic mechanisms in entorhinal cortex microcircuits in AD patients [80] although there is a concern that the use of zolpidem may be associated with increased risk of AD among older people [81]. Dronabinol, synthetic Δ-9-tetrahydrocannabinol, may suppress agitation by activating endocannabinoid receptors [82]. Beyond direct effects on neurotransmission, allopregnanolone has been explored for its effects on neurogenesis as a GABAA receptor modulator [83] and riluzole for its inhibition of neuronal death as a glutamate modulator [84].
FUTURE PROSPECTS OF DR IN ALZHEIMER’S DISEASE
As etiological and clinical heterogeneity is a common feature of AD, no one therapy is suitable for all AD patients. Phenotyping of AD at network, cellular, genetic, and molecular levels will aid more ‘tailor-made’ therapy strategies. There are ongoing efforts to phenotype AD, such as the Rocky Mountain Alzheimer’s Disease Center Longitudinal Biomarker and Clinical Phenotyping Study [85], to establish a large bank of blood, cerebrospinal fluid, imaging, and clinical data in AD research. Future discovery of multimodal biomarkers will help identify subpopulations with relatively homogeneous pathophysiological signatures among intrinsically heterogeneous AD patient groups [86]. Such subpopulations may be specifically targeted with some of the drugs discussed above according to their respective mechanism of action. Therefore, in future clinical trials for AD, inclusion criteria could be designed from the results of AD phenotyping and suitable DR strategies implemented accordingly.
The US-ADNI study (Alzheimer’s Disease Neuroimaging Initiative) reported the importance of vascular factors in the development of late-onset AD [87]. They performed multifactorial, data-driven analyses based on the results of images, blood, and cerebrospinal fluid biomarkers collected from 1171 ADNI study-enrolled patients. The earliest significant changes in late-onset AD were vascular lesions, including impaired CBF, which preceded abnormalities in Aβ and tau deposition. It is well known that vascular lesions are closely related to aging and environmental factors. The US-ADNI data thus valuably revisits the importance of aging and environmental factors in the onset of late-onset AD.
Consistent with this notion, some ongoing DR strategies intervene with anti-diabetic drugs, including insulin, liraglutide, and dapagliflozin, anti-hypertensive drugs, including losartan, telmisartan, candesartan, amlodipine, and prazosin, and anti-thrombotic drugs, including cilostazol and dabigatran. These studies suggest the importance of the neurovascular approach in treating AD [14]. Arteriolosclerosis due to diabetes and/or hypertension, as well as cerebral amyloid angiopathy, further impair efficiency of Aβ clearance, dependent upon vascular integrity, forming a ‘vicious cycle’ of Aβ accumulation and neurodegeneration (Fig. 1) [88].
Until now, AD research and development has tended to focus on neurodegeneration, but given it is increasingly being understood as a disease with heterogeneous variants in which the causative gene, susceptibility gene, mutation locus, and disease modifier are substantially variable, it is important to have a research culture and financial support that allow drug development targeting various mechanisms of action. Promising multi-faceted DR studies are being conducted based on the various lines of evidence listed in Tables 1–3, but in the background physicians are relentlessly promoting clinical trials, while overcoming drug patent problems. Breaking through existing concepts and persuading against prevailing doubts of DR may be an arduous process. However, the realization of effective DR strategies could herald a major paradigm shift in AD research and development.
CONCLUSION
The limits of drug discovery research have been tested in recent years but existing drugs could represent ‘hidden clues’, repurposing current therapies and reinvigorating approaches to a multitude of disorders. Nevertheless, DR research requires commitment in form of funding and cooperation between industry, academia, and government. The procedures governing DR, and indeed translational research as a whole, must improve to facilitate this process.
