**4. Prevention and treatment of Alzheimer's disease**

The main risk factors for dementia are age and genetics (see more information about AD risk factors at http://www.alz.org/alzheimers\_disease\_causes\_risk\_factors.asp), although other risk factors may also influence the onset of dementia. For instance, since the brain is nourished by a rich network of blood vessels, cardiovascular alterations are considered a risk factor for neurological disorders. In fact, vascular dementia is linked to morphological changes to blood vessels which are in turn present in other types of dementia like AD. Indeed, a healthy cardiovascular system is frequently linked to brain protection [93]. In this context, the control of blood cholesterol levels, blood pressure, and body weight is recommended to maintain good brain health. In fact, high-fat diets and sedentary lifestyles are becoming major concerns in terms of their contribution to the high incidence of dementia in Western society, whereas regular physical exercise and heart-healthy diets are also good habits to lower the risk of dementia [35].

Only two types of drugs are currently available to treat Alzheimer's disease: acetylcholines‐ terase inhibitors (often shortened to just "cholinesterase inhibitors") and NMDA receptor antagonists. Cholinesterase inhibitors (donepezil, rivastigmine, and galantamine) bind to and reversibly inactivate cholinesterases, inhibiting acetylcholine hydrolysis. Such inhibition results in increased acetylcholine concentrations at cholinergic synapses and indeed, AD involves a substantial loss of cholinergic neurons in the neocortex and hippocampus, which in turn contributes to the AD symptomatology and to memory impairment in particular. Therefore, increased levels of acetylcholine are thought to protect against the death of cholinergic neurons, alleviating AD symptoms [94]. Memantine is a low-affinity voltagedependent antagonist of glutamatergic NMDA receptors. By binding to the NMDA receptor, memantine inhibits the sustained influx of Ca2+ ions from the extracellular milieu, thereby preventing neuronal death by excitotoxicity. Such a pathogenic mechanism can be mediated by the Aβ oligomers that bind to NMDA receptor as agonists, favoring Ca2+ influx and neuronal excitotoxicity [95]. Interestingly, memantine preserves physiological receptor activity, such that released glutamate can still mediate receptor activation leading to neuronal depolarization in postsynaptic neurons [96]. However, neither cholinesterase inhibitors nor NMDA antago‐ nists have disease-modifying effects in AD and they are generally viewed as palliative treatments with marginal to minimal clinical efficacy, either alone or in combination. There‐ fore, only a small percentage of patients with AD respond to these treatments and these responders normally undergo a short period of cognitive stabilization after which they again suffer from the cognitive decline associated to largescale neuronal degeneration [97, 98]. This scenario highlights the unmet clinical need for the treatment of AD and related conditions.

These subjects were diagnosed as AD during a 5-year follow-up even though they displayed no cognitive impairment at entry. The PC species identified were diacyl PC 36:6, 38:0, 38:6, 40:1, 40:2, 40:6, PC acyl-alkyl 40:6, and LPC 18:2, as well as the acylcarnitines (ACs) propionyl AC (C3), and C16:1-OH [89]. It is noteworthy that control subjects (not previously diagnosed with AD) did not display any of these modifications, while already diagnosed patients with AD also showed decreased levels of these PC species. Moreover, downregulation of this panel of lipids predicted phenoconversion from healthy to MCI/AD within a 2–3 year time frame with 90% accuracy [89]. These data were supported by independent studies showing decreased levels of PC 38:4, 38:6, and 40:6 in the plasma or serum of AD subjects [86, 90]. In addition, a variety of peripheral lipid changes were also reported that might potentially be useful for early AD diagnosis, such as lower levels of SM and increased levels of Cers in the plasma or serum of patients with AD. In particular, there were significantly fewer SM species containing long chains (e.g., 22 and 24 carbon atom acyl chains) in AD subjects [86, 91]. In parallel, increased Cer levels were reported in the plasma of patients with AD [91, 92]. SM can be metabolized into Cers, second messengers that regulate cellular differentiation, proliferation and apoptosis. Upregulated levels of Cers were concomitant with significant reductions in SM in the plasma of patients with AD. A correlation between the decrease in SM and the increase in Cers was particularly robust in the ratios of SM and Cer species with identical fatty acyl chains. Cer alterations were particularly evident in mild-to-moderate stages of AD [91]. Moreover, it is noteworthy that upregulated Cer levels were significantly correlated with the onset of memory

impairment, supporting the role of Cers as potential AD biomarkers [92].

**4. Prevention and treatment of Alzheimer's disease**

the onset of AD.

140 Update on Dementia

dementia [35].

In conclusion, a wide range of peripheral fluid changes have been described that could be used as biomarkers for early AD diagnosis. However, many of the clinical studies involved are crosssectional in nature and some of them do not reveal reliable biomarkers to test disease pro‐ gression. Nevertheless, longitudinal studies with several years of follow-up do identify promising biomarkers for early AD diagnosis that reliably predict cognitive impairment and

The main risk factors for dementia are age and genetics (see more information about AD risk factors at http://www.alz.org/alzheimers\_disease\_causes\_risk\_factors.asp), although other risk factors may also influence the onset of dementia. For instance, since the brain is nourished by a rich network of blood vessels, cardiovascular alterations are considered a risk factor for neurological disorders. In fact, vascular dementia is linked to morphological changes to blood vessels which are in turn present in other types of dementia like AD. Indeed, a healthy cardiovascular system is frequently linked to brain protection [93]. In this context, the control of blood cholesterol levels, blood pressure, and body weight is recommended to maintain good brain health. In fact, high-fat diets and sedentary lifestyles are becoming major concerns in terms of their contribution to the high incidence of dementia in Western society, whereas regular physical exercise and heart-healthy diets are also good habits to lower the risk of Developing disease-modifying drugs (DMDs) capable of preventing neuron degeneration and thereby counteracting AD progression is one of the most pressing challenges of modern pharmacology. Since the pathological process of AD begins many years before its clinical diagnosis, the optimal time for a disease-modifying therapy may be during the prodromal stage of AD. Therefore, clinical diagnosis of AD must be achieved when patients show no relevant clinical signs. Indeed, the development of DMDs will require the concomitant incorporation of reliable biomarkers to identify early stages of AD (see Section 3.3). Hitherto, no DMDs are available for AD and although several have been tested up to phase 3, none has yet achieved marketing approval. The recurrent failures in clinical trials raise a number of questions about our understanding of AD pathophysiology. In this sense, the amyloid cascade hypothesis has not only influenced the study of AD pathophysiology over the past 2 decades but also, the choice of drug targets (see Section 2). Therefore, most clinical trials have set out to prevent Aβ accumulation, either by inhibiting its production/aggregation or enhancing its clearance, as well as reducing tau phosphorylation [99, 100]. However, it remains unclear if these two hallmarks of AD are a cause or consequence of the disease. In fact, they could lie downstream of previous molecular/cellular alterations, as a result of the disease pathology (damage response proteins) and/or as products of an endogenous protective response to disease-induced damage. Nonetheless, over the past 20 years the main focus of biomedical research and the associated drug discovery programs for AD have targeted brain amyloid or tau hyperphosphorylation, and the associated formation of neurofibrillary tangles [18].

Mutations in the BACE-1 gene have not been related to AD but elevated levels of this enzyme have consistently been found in both the brain and CSF of patients with AD [101–103]. Since β-secretase activity is pathologically elevated in AD, BACE1 inhibition has been addressed as a potential therapeutic approach to combat AD. In fact, both genetic deletion of BACE-1 and administration of a BACE-1 inhibitor rescued cognitive deficits and lowered brain Aβ production in AD mouse models. Interestingly, although BACE-1 has other substrates, its inhibition was apparently free of side effects in AD mice [104, 105]. The latest generation of small molecule BACE-1 inhibitors has achieved satisfactory brain penetration and a robust reduction in cerebral Aβ in preclinical animal models. Furthermore, administration of most of these inhibitors in humans also reduced Aβ and sAPPβ levels, whereas sAPPα (the α-secretase cleavage product) was enhanced in the CSF. This observation is consistent with BACE-1 inhibition since β- and α-secretase compete for APP processing (see **Figure 1**). Many of these BACE-1 inhibitors are still in phase-1 clinical trials where safety and tolerability are tested but some of them are currently in phase 2/3, although no clinical efficacy data are as yet available (**Table 1**). Interestingly, one such drug (LY2886721 from Eli Lilly Company) was discontinued in a phase-2 trial because a number of subjects developed hepatic toxicity, although they were not associated with the mechanism of action of BACE1 [106].



Mutations in the BACE-1 gene have not been related to AD but elevated levels of this enzyme have consistently been found in both the brain and CSF of patients with AD [101–103]. Since β-secretase activity is pathologically elevated in AD, BACE1 inhibition has been addressed as a potential therapeutic approach to combat AD. In fact, both genetic deletion of BACE-1 and administration of a BACE-1 inhibitor rescued cognitive deficits and lowered brain Aβ production in AD mouse models. Interestingly, although BACE-1 has other substrates, its inhibition was apparently free of side effects in AD mice [104, 105]. The latest generation of small molecule BACE-1 inhibitors has achieved satisfactory brain penetration and a robust reduction in cerebral Aβ in preclinical animal models. Furthermore, administration of most of these inhibitors in humans also reduced Aβ and sAPPβ levels, whereas sAPPα (the α-secretase cleavage product) was enhanced in the CSF. This observation is consistent with BACE-1 inhibition since β- and α-secretase compete for APP processing (see **Figure 1**). Many of these BACE-1 inhibitors are still in phase-1 clinical trials where safety and tolerability are tested but some of them are currently in phase 2/3, although no clinical efficacy data are as yet available (**Table 1**). Interestingly, one such drug (LY2886721 from Eli Lilly Company) was discontinued in a phase-2 trial because a number of subjects developed hepatic toxicity, although they were

not associated with the mechanism of action of BACE1 [106].

Lilly & Co.

Idec

Squibb

Begacestat GSI-953 Pfizer Notch-sparing

Semagacestat LY450139 Eli Lilly & Co. γ-secretase

Merck β-Secretase

**action** 

inhibitor

inhibitor

inhibitor

Notch-sparing γ-secretase inhibitor

γ-secretase inhibitor

β-secretase inhibitor

β-Secretase inhibitor

**Result of study** 

Discontinued in phase 2

Ongoing in phase 2/3

Ongoing in phase 2/3

Ongoing in phase 2

Discontinued in phase 3

Discontinued in phase 3

Phase-1 trial completed

**Clinical trial ID\* Observations** 

NCT01561430 Altered liver

–

NCT00890890 Lack of clinical

NCT02245737 –

NCT02322021 –

NCT00547560 –

NCT01739348 NCT01953601

NCT01035138 NCT00762411 NCT00594568 biochemistry

Lack of clinical improvement Increased risk of skin cancer and infections.

improvement Increased rate of skin cancers

**Drug Synonyms Company Mechanism of**

LY2886721 – Eli Lilly & Co. β-Secretase

AZD3293 LY3314814 Astra Zeneca/ Eli

MK-8931-009

E2609 – Eisai/Biogen

Avagacestat BMS-708163 Bristol-Myers

Verubecestat MK-8931

142 Update on Dementia


Some data in this table are available at http://www.alzforum.org/therapeutics/.

\*Clinical trial IDs were obtained from Clinicaltrials.gov unless specified.

\*\*Information regarding the clinical use of lithium was obtained from [121, 122] and Clinicaltrials.gov.

**Table 1.** Developed disease-modifying drugs for AD treatment in clinical trials.

Clinical mutations in PS1 are supposed to induce a loss of γ-secretase function that in turn prevents Aβ generation and increases the Aβ 42/40 ratio (an increase in the longer vs. shorter Aβ isoforms) [31]. Such loss of function is then translated into increased neuronal Aβ produc‐ tion, which is further potentiated with the ageing in AD mice harboring FAD mutations [23, 58]. This pathological mechanism is associated with accumulation of autophagic vesicles in axonal dystrophies surrounding amyloid plaques, which are principally formed by long hydrophobic isoforms of Aβ like Aβ42. Therefore, γ-secretase inhibition or modulation has also been studied as a plausible therapeutic approach against AD, although non-specific effects hinder the development of γ-secretase inhibitors (GSI) as DMDs given that γ-secretase also cleaves several type-I transmembrane proteins such as the Notch receptor, N-cadherin, ErbB4, and p75NTR (see Section 2).

**Drug Synonyms Company Mechanism of**

Abbott Laboratories

Squibb

Biosciences

Therapeutics Ltd

Therapeutics Ltd

Neuroscience SE

Some data in this table are available at http://www.alzforum.org/therapeutics/. \*Clinical trial IDs were obtained from Clinicaltrials.gov unless specified.

**Table 1.** Developed disease-modifying drugs for AD treatment in clinical trials.

TauRx

TauRx

Janssen

Axon

RG7345 RO6926496 Roche Tau-targeted

Public institutions Tau

Valproate Depakote,

144 Update on Dementia

Lithium \*\* Lithium

Depakene

carbonate

Epothilone D BMS-241027 Bristol-Myers

TPI 287 – Cortice

Methylene Blue Rember TM TRx-0014

Methylthioninium

LMT-X Methylene

AADvac1 Axon

Blue TRx-0237

ACI-35 – AC Immune SA

peptide 108 conjugated to KLH

(MT)

**action** 

inhibitor

phosphorylation

phosphorylation inhibitor

Microtubule stabilizer

Microtubule stabilizer

Tau aggregation inhibitor

Tau aggregation inhibitor

Tau-targeted active

Tau-targeted active

passive immunotherapy

\*\*Information regarding the clinical use of lithium was obtained from [121, 122] and Clinicaltrials.gov.

immunotherapy

immunotherapy

Tau

**Result of study** 

Discontinued in phase 3

Ongoing in phase 2

Discontinued in phase 1

Ongoing in phase 1

Discontinued in phase 2

Phase 3 completed

Phase 1 completed

Ongoing phase 1

Discontinued in phase 1

**Clinical trial ID\* Observations** 

NCT00071721 Lack of clinical

NCT01492374 No reasons reported regarding discontinuation in

phase 1

Discrepant results reported

Blinding of phase-2 trial has been questioned

No results available

as yet

–

NCT02281786 No reasons reported regarding discontinuation in

phase 1

ISRCTN72046462 (see at isrctn.com) NCT01055392 NCT02129348 NCT00088387

NCT01966666

NCT00684944 NCT00515333

NCT01689233 NCT01689246 NCT01626378

ISRCTN13033912 (see at isrctn.com)

NCT02031198 –

improvement Brain volume loss

Discrepant results reported

Apparently effective in early AD (amnestic MCI) but not in mild-tomoderate AD

Semagacestat was the first GSI to undergo clinical trials, and it reduced Aβ concentrations in the mouse CNS and human plasma [107, 108]. Two large phase-3 trials with semagacestat were prematurely interrupted due to serious adverse events, including hematological alterations, and an increased risk of skin cancer and infections that were attributed to inhibition of the Notch signaling pathway. Furthermore, a worsening of cognition was observed in AD-treated patients [109]. Notch-sparing GSIs (second generation inhibitors) and modulators (agents that shift γ-secretase cleavage from longer to shorter Aβ species without affecting Notch cleavage) were then designed for clinical development. Avagacestat and begacestat were first conceived as notch-sparing GSIs that supposedly display greater selectivity for APP than for Notch cleavage [10], although this was recently reported not to be the case [31]. Therefore, these drugs are also likely to fail and indeed, the poor clinical efficacy of Avagacestat was coupled to an increased rate of skin cancers, again suggesting side effects attributable to Notch signaling inhibition (see **Table 1**). Finally, some non-steroidal anti-inflammatory drugs (NSAIDs) modulate γ-secretase (GSMs), decreasing the abundance of Aβ42 while increasing that of Aβ38. Tarenflurbil (the R-enantiomer of flurbiprofen) was tested in a phase-3 trial but it did not slow cognitive decline in patients, while it did increase the frequency of dizziness, anemia, and infection. This failure of tarenflurbil was attributed to its low potency and poor brain penetration [10, 99].

Aggregation of Aβ monomers into higher molecular weight oligomers is thought to be a key neurotoxic event leading to neurodegeneration in the amyloid pathology [7]. For this reason, some DMDs also target this conversion to fight AD. Tramiprosate and scyllo-inositol are two compounds that prevent the transition from Aβ monomers to oligomers, thus favoring Aβ clearance from the brain by insulin-degrading enzyme (IDE) and neprilysin [110]. In addition, scyllo-inositol can also directly bind to Aβ oligomers, promoting their dissociation. Both these drugs have been involved in phase-2 clinical trials and both reduced Aβ42 levels in the CSF of treated patients. In a larger phase-3 study, tramiprosate failed to induce clinical improve‐ ment, and thus, further clinical evaluation is still necessary. Scyllo-inositol, also failed to produce significant clinical improvement in a phase-2 trial. Rosiglitazone is an anti-diabetic drug that improves spatial learning and memory abilities, and it mildly decreases Aβ42 brain levels by activating PPARγ and upregulating IDE in AD mice [111]. This drug was involved in phase-2 and phase-3 clinical trials, although the inconclusive results in phase 2 were followed by a lack of clinical efficacy in a larger phase-3 study [99, 112].

Another therapeutic approach to promote Aβ clearance was based on immunization toward Aβ. Active immunization by vaccination stimulates the immune response to promote antibody formation against pathogenic forms of Aβ, such as Aβ42. Active Aβ immunotherapy has been studied since 1999 when the generation of Aβ antibodies was shown to produce clearance of cerebral Aβ by phagocytic microglia in animal models [113]. Unfortunately, this revolutionary approach soon suffered its first setback in a phase-2 trial to test active immunization using full length human Aβ42 peptide, with some patients developing brain inflammation with aseptic meningoencephalitis and provoking the termination of the clinical study [99]. Passive immu‐ notherapy is an alternative strategy and recent approaches were based on shorter Aβ immu‐ nogens, such as the humanized monoclonal antibody to Aβ1–5, bapineuzumab, which binds to both soluble and fibrillar forms of Aβ. Despite the evidence of adverse effects in phase-1 trials, bapineuzumab advanced to phases 2 and 3 where it failed to demonstrate clinical efficacy in patients with AD. Another antibody against Aβ is Solanezumab, a humanized monoclonal antibody against Aβ16–24 that preferentially binds to soluble Aβ. In phase-2 trials, solanezu‐ mab was found to be safe while increasing plasma and CSF levels of Aβ40 and Aβ42, an indication of decreased plaque load in the brain. However, solanezumab had no effect on behavioral outcomes. Despite the lack of efficacy in phase 2, the antibody advanced to phase-3 trials in patients with mild-to-moderate AD where the primary endpoints, both cognitive and functional, were not achieved [18]. Many other humanized antibodies have been developed, directed at different regions of the Aβ peptide, some entering phase-3 trials (Gantenerumab and Aducanumab) and others having been discontinued (Ponezumab; **Table 1**).

According to the amyloid cascade hypothesis, Aβ accumulation precedes and drives tau hyperphosphorylation via the activation of different kinases, including cyclin dependent kinase 5 (CDK5) and glycogen synthase kinase 3β (GSK3β) [14, 114]. Tau hyperphosphoryla‐ tion is thought to destabilizes neuronal microtubules, impairing axonal transport and leading to neurite pathology, finally resulting in deficient synaptic function and neuronal death [115, 116] (see Section 2). In this context, DMDs were developed to inhibit tau phosphorylation, as well as compounds that prevent tau aggregation. GSK3β is the main enzyme involved in tau hyperphosphorylation, and lithium and valproate are both drugs that inhibit GSK3β and reduce tau phosphorylation in animal models [117]. Unexpectedly, valproate impaired the cognitive and functional status, and it was also associated with a reduced brain volume in patients with AD receiving the drug in clinical trials [118]. Lithium is neuroprotective in animal models of AD, not only via the inhibition of GSK-3β but also through the remodeling of Aβ plaques, leading to a decrease in the number of dystrophic axons, reduced neuronal degener‐ ation and improved cognitive scores in AD mice [119, 120]. However, no conclusions have been reached regarding the clinical efficacy of lithium for AD treatment. Some clinical trials failed to demonstrate a protective effect of lithium on cognitive performance, although a more recent clinical study showed that lithium reduced cognitive decline patients with early AD (amnesic MCI) [121, 122]. Tau hyperphosphorylation compromises its ability to bind to microtubules in AD, provoking microtubule instability. In this sense, epothilone D and TPI 287, synthetic paclitaxel-derived microtubule-stabilizing drugs with good BBB permeability, were assessed in phase-1 trials of safety and tolerability. Unfortunately, epothilone D was recently discontinued (see **Table 1**). Tau hyperphosphorylation also provokes tau aggregation which is also considered a key neurotoxic event in AD [123]. LMT-X is a new version of methylene blue, a compound that was tested and discontinued in a phase-2 trial to treat AD. LMT-X is an inhibitor of tau aggregation that specifically disrupts tau-tau interactions in the microtubule binding region. In a phase-2 trial, this new drug slowed down the cognitive decline in a subgroup of patients, and it is now being tested in phase-3 trials, although information about clinical efficacy is not yet available [124, 125]. Finally, two tau-derived peptide vaccines that stimulate active immunization entered phase I studies: AADvac1 and ACI-35. AADvac1 is a synthetic peptide corresponding to a naturally occurring, truncated and misfolded form of tau. ACI-35 is a liposomal vaccine containing a synthetic peptide corre‐ sponding to human protein tau sequence 393–408 (numbering according to the tau 2N4R isoform), with phosphorylated S396 and S404 residues. Vaccination with these peptides improves neurobehavioral deficits in AD rodents while ACI-35 is characterized by a rapid and robust polyclonal antibody response specific to phosphorylated tau in WT and AD mice [125]. In addition, passive immunization has also been investigated using a humanized monoclonal antibody targeting pS422 phospho-tau. In AD mice, chronic administration of this antibody reduced hyperphosphorylated tau accumulation [126], although clinical studies with this antibody were recently discontinued in phase 1 (see **Table 1**).

in phase-2 and phase-3 clinical trials, although the inconclusive results in phase 2 were

Another therapeutic approach to promote Aβ clearance was based on immunization toward Aβ. Active immunization by vaccination stimulates the immune response to promote antibody formation against pathogenic forms of Aβ, such as Aβ42. Active Aβ immunotherapy has been studied since 1999 when the generation of Aβ antibodies was shown to produce clearance of cerebral Aβ by phagocytic microglia in animal models [113]. Unfortunately, this revolutionary approach soon suffered its first setback in a phase-2 trial to test active immunization using full length human Aβ42 peptide, with some patients developing brain inflammation with aseptic meningoencephalitis and provoking the termination of the clinical study [99]. Passive immu‐ notherapy is an alternative strategy and recent approaches were based on shorter Aβ immu‐ nogens, such as the humanized monoclonal antibody to Aβ1–5, bapineuzumab, which binds to both soluble and fibrillar forms of Aβ. Despite the evidence of adverse effects in phase-1 trials, bapineuzumab advanced to phases 2 and 3 where it failed to demonstrate clinical efficacy in patients with AD. Another antibody against Aβ is Solanezumab, a humanized monoclonal antibody against Aβ16–24 that preferentially binds to soluble Aβ. In phase-2 trials, solanezu‐ mab was found to be safe while increasing plasma and CSF levels of Aβ40 and Aβ42, an indication of decreased plaque load in the brain. However, solanezumab had no effect on behavioral outcomes. Despite the lack of efficacy in phase 2, the antibody advanced to phase-3 trials in patients with mild-to-moderate AD where the primary endpoints, both cognitive and functional, were not achieved [18]. Many other humanized antibodies have been developed, directed at different regions of the Aβ peptide, some entering phase-3 trials (Gantenerumab

followed by a lack of clinical efficacy in a larger phase-3 study [99, 112].

146 Update on Dementia

and Aducanumab) and others having been discontinued (Ponezumab; **Table 1**).

According to the amyloid cascade hypothesis, Aβ accumulation precedes and drives tau hyperphosphorylation via the activation of different kinases, including cyclin dependent kinase 5 (CDK5) and glycogen synthase kinase 3β (GSK3β) [14, 114]. Tau hyperphosphoryla‐ tion is thought to destabilizes neuronal microtubules, impairing axonal transport and leading to neurite pathology, finally resulting in deficient synaptic function and neuronal death [115, 116] (see Section 2). In this context, DMDs were developed to inhibit tau phosphorylation, as well as compounds that prevent tau aggregation. GSK3β is the main enzyme involved in tau hyperphosphorylation, and lithium and valproate are both drugs that inhibit GSK3β and reduce tau phosphorylation in animal models [117]. Unexpectedly, valproate impaired the cognitive and functional status, and it was also associated with a reduced brain volume in patients with AD receiving the drug in clinical trials [118]. Lithium is neuroprotective in animal models of AD, not only via the inhibition of GSK-3β but also through the remodeling of Aβ plaques, leading to a decrease in the number of dystrophic axons, reduced neuronal degener‐ ation and improved cognitive scores in AD mice [119, 120]. However, no conclusions have been reached regarding the clinical efficacy of lithium for AD treatment. Some clinical trials failed to demonstrate a protective effect of lithium on cognitive performance, although a more recent clinical study showed that lithium reduced cognitive decline patients with early AD (amnesic MCI) [121, 122]. Tau hyperphosphorylation compromises its ability to bind to microtubules in AD, provoking microtubule instability. In this sense, epothilone D and TPI

The aforementioned therapeutic approaches summarize the attempts to develop DMDs based on the amyloid cascade hypothesis, principally focused on Aβ and hyperphosphorylated tau protein. With several anti-amyloid drugs now having failed in late stage clinical trials, many critical voices in the scientific community have questioned the validity of the amyloid hypothesis to explain the pathophysiology of AD and as platform on which to develop DMDs for AD therapy. Moreover, the incidence of serious side effects observed in human trials is another drawback to the clinical development of these types of drugs, particularly when many of these adverse effects are associated with the mechanism of action of the compounds tested. However, the amyloid hypothesis cannot be disregarded due the lack of reliable biomarkers to detect efficacy at early stages, and because many of the compounds in clinical trials cross the BBB poorly or cause side effects that forced trials to be discontinued before efficacy could be evaluated [18, 127].
