Lithium,a Treatment Option for Alzheimer’s Disease? A Review of Existing Evidence and Discussion on Future Perspectives

LITHIUM USE IN ALZHEIMER’S DISEASE: TREATMENT OF NEUROPSYCHIATRIC SYMPTOMS

Literature on the use of lithium to treat NPS in AD is quite scarce. In 2017, Devanand et al. published a case series on the effect of lithium treatment on three AD patients with agitation and/or aggression [7]. Patients showing no response or a partial response to antipsychotics improved on lithium, as determined by clinical judgement and/or objective rating scales (Clinical Global Impression-Severity and -Change, and the Neuropsychiatric Inventory, NPI). Adverse effects were absent to tolerable at low oral doses (300–600 mg) with low serum levels at the final visit (0.30–0.32 mmol/L). These outcomes led to a Phase II placebo-controlled 12 week trial to further explore the efficacy and side effects of the same low dose lithium in the treatment of agitation/aggression with or without psychosis in 80 patients with AD [8].

Although lithium was not significantly superior to placebo in improving symptoms of agitation/aggression, with 31.6% of responders on lithium compared to 17.9% on placebo (χ² = 1.26, p = 0.26), it did permit moderate or marked improvement in global Clinical Global Impression-Change in 10/38 = 36.8% of patients taking lithium compared to none of those taking placebo (0/39 = 0%, Fisher’s exact test p < 0.001). In addition, exploratory analyses showed greater improvements with lithium than with placebo in NPI scores for delusions and irritability/lability (ps < 0.05) [9]. It remains unclear whether narrowing to one NPS (agitation/aggression) could have impeded the observation of a more global positive effect of lithium. One interesting question warranting further investigation is whether a specific patient profile combining a particular set of NPS may benefit most from lithium.

LITHIUM USE IN ALZHEIMER’S DISEASE: DISEASE-MODIFYING TREATMENT

Hallmarks of neurodegenerative diseases, such as AD, include a decrease in neurogenesis and an increase in neurotoxicity. Several studies have been conducted both in vitro and in vivo to investigate a putative positive effect of lithium on neurogenesis and a neuroprotective effect that could be translated into it being considered as a therapeutic option for dementia.

A Danish epidemiological study in an observational cohort explored the effect of continuous use of lithium on the incidence of dementia. They identified a sample of 16,238 individuals who purchased lithium at least once and a further 1,487,177 from the general population who did not purchase lithium, acting as control. They observed that lithium administration lowered the risk of dementia [10]. They also investigated a possible association between lithium treatment in patients with mania or bipolar disorder and subsequent dementia. They reported that continued treatment with lithium was associated with a reduced rate of dementia in contrast to continued treatment with anticonvulsants, antidepressants, or antipsychotics [11].

Macdonald et al. [12] conducted a pilot study assessing the feasibility and tolerability of lithium carbonate at therapeutic levels (targeted serum levels between 0.3 and 0.8 mM) in mild to moderate AD over an extended period (up to one year). This open label observational study enrolled 22 patients, of which 8 completed the full year of therapy. 21% withdrew treatment due to adverse events, 21% due to investigator’s decision. High rate of discontinuation, outside of the ones due to adverse events, seems to be influenced by change in concomitant treatment leading to contraindication and/or loss of compliance to study procedure or study drug. Interestingly, no serious adverse event nor irreversible events related to treatment were encountered during the trial. Side effect appeared to be mild and those estimated to be due to lithium were not the cause of withdrawing in most of the cases.

Hampel et al. [13] conducted a placebo controlled trial on the effect of 10 weeks lithium treatment in 71 mild AD patients (Mini-Mental State Examination (MMSE) score > or = 21 and < or = 26). The 10-week treatment included a 6-week titration phase to reach the target serum level of lithium (0.5–0.8 mmol/L). The treatment showed no effect on GSK3β activity in lymphocytes or cerebrospinal fluid-based biomarker concentrations (p-tau, total tau, and Aβ_1 - 42; p > 0.05), on global cognitive performance as measured by the Alzheimer’s Disease Assessment Scale-Cognitive Subscale (ADAS-Cog) (p = 0.11) or on depressive symptoms (NPI score). However, in a second analysis of this trial, Thomas et al. observed a significant increase in BDNF serum levels and a significant decrease in ADAS-Cog scores among patients in the lithium treated group in comparison to placebo-treated patients and found a correlation between outcome and serum lithium concentration. Noteworthy is the cognitive improvement (ADAS-Cog increase >4 points) that was observed in 28.6% of lithium-treated patients versus 14.3% of the placebo group, although this difference did not reach statistical significance [14].

Forlenza et al. [15] evaluated the effect of long-term lithium treatment on cognitive and biological outcomes in patients with amnestic mild cognitive impairment (MCI) in a 2-year double blinded, randomized, controlled trial followed by a single blinded extension phase for an additional 24 months. In an interim analysis at 12 months of treatment for 45 patients (24 received lithium (0.25–0.5 mmol/L)), study group reported that lithium treatment was associated with a reduced cognitive decline and significant decrease in p-tau (p = 0.02) compared to placebo. In the final analysis [16], 61 patients were randomized to receive either lithium (50.8%) or placebo (49.2%), of these 52 completed the double blinded phase and 40 completed extension phase. Over 24 months, patients treated with lithium were cognitively and functionally stable, whereas patients treated with placebo had statistically significant cognitive and functional decline on ADAS-Cog and CDR-SoB scores. There was a trend toward a lower conversion rate from MCI to dementia in patients treated with lithium compared to placebo. Overall tolerability and adherence were good. The same team re-assessed the global cognitive and functional state of the patients enrolled in the initial trial 11 to 15 years later [15]. Of the 36/61 participants the team was able to reach, 22 had received lithium and 14 placebo. They found statistically significant differences in MMSE score and a better performance in the phonemic verbal fluency test for patients having previously received lithium [17].

Nunes et al. [18] evaluated over a 15-month period the effect of daily administration of micro-dose of lithium (300μg) on patients with AD. The lithium-treated group showed no decreased performance on the MMSE test in contrast to lower scores observed for the control group, with differences becoming significant from three months of treatment and increasing progressively.

In 2014, Mauer et al. [19] conducted a systematic review on twenty-four clinical, epidemiological, and biological reports assessing lithium at standard or low doses for dementia or other behavioral or medical purposes. They reported that most studies on lithium at standard doses showed notable cognitive benefits or effectiveness in preventing dementia.

Included epidemiological studies were numerous, large, and replicated and the randomized controlled trials also showed clear biological and/or clinical benefits. Aligned with these outcomes, Matsunaga et al. [20] conducted a meta-analysis of randomized placebo-controlled trials testing lithium as a treatment for patients with AD and/or individuals with MCI (3 clinical trials, 232 participants). They reported that lithium treatment appeared to be associated with significant decrease in cognitive decline as compared to placebo (standardized mean difference = –0.41, 95% confidence interval = –0.81 to –0.02, p = 0.04, I² = 47%).

MECHANISM OF ACTION OF LITHIUM IN AD

Lithium is the smallest and lightest solid element and yet it has a wide diversity of effects. Proposed pathways involved in the pleiotropic pharmacological effects of lithium in neurodegenerative disorders such as AD illustrates the different mechanisms underlying the observed efficacy of lithium (Fig. 1).

OPEN IN VIEWER

Fig. 1

Pleiotropic pharmacological effects of lithium in neurodegenerative disorders such as AD. [6]; [18 –35].

In vitro and in vivo studies have identified several pathways through which lithium exerts its effects.

In AD-specific pathways [6 , 21–23]:

Lithium acts as an inhibitor of the enzyme GSK3β implicated in AD pathogenesis via its role in tau hyperphosphorylation, neuroinflammation, oxidative stress, apoptosis, and modulation of gamma-secretase and beta-secretase both required for the production of Aβ protein.

Lithium also acts on disease nonspecific pathways:

Inhibition of oxidative stress and the inflammatory cascade following lithium treatment can be observed with an increase in antioxidants and a decrease in lipid peroxidation, free radicals, inflammatory cytokines and microglia activation [21 , 24].

This broad spectrum of lithium actions could explain the positive results observed in animal models with regards to AD biomarkers and cognitive improvement.

LIMITATIONS TO CURRENT USE OF LITHIUM IN AD

Lithium is a monovalent cation that shares certain properties with sodium, potassium, and calcium. The active agent in lithium salts is the lithium-ion Li⁺, which having a smaller diameter, can easily displace K⁺ and Na⁺ and even Ca^2 + from important enzyme or neurotransmitter receptor sites [29]. The safety profile of lithium and its well-known narrow therapeutic index are directly correlated to its circulating concentration and cell penetration mechanisms.

In the treatment of bipolar disorder, the administered daily dose varies between 500 mg and 1,250 mg of lithium carbonate (10–20 mEq/day). To cross cell membranes, lithium uses mainly the sodium pump, sodium leak channel, and sodium-lithium counter transporter.

All these systems involve other cations and require a certain circulating concentration (>0.5 mEq/L varying between subject) in order to shift the transport towards getting lithium inside the cells in accordance with the pharmacokinetics plateau principal [29 –32].

Signs of lithium toxicity are tremor, ataxia, dehydration, nystagmus, renal impairment, confusion, and convulsion. They are mostly observed at blood concentrations >1.5 mEq/L [30 –32].

The narrow therapeutic index of lithium, between 0.6 and 1.2 mEq/L, imposes the routine monitoring of its serum concentrations: serious toxicity can be observed at concentrations exceeding 1.5 mEq/L. The maintenance level is between 0.4 and 0.8 mEq/L [30, 32].

Lithium has no clinically important protein binding properties and no metabolites. It is excreted almost entirely by the kidneys, although small amounts are also lost in sweat and feces. A substantial amount of filtered lithium is reabsorbed (primarily in the proximal tubules), explaining an observed five-fold higher renal clearance rate of creatinine compared to lithium [32]. The elimination half-life of lithium is about 18 to 24 h, although it is considerably longer in the elderly because of the age-related decrease in glomerular filtration rate (and correspondingly shorter in young patients). Lithium is not liver metabolized [30].

As a direct consequence of lithium’s pharmacokinetic profile, any conditions affecting circulating concentration (sodium depletion, dehydration) can trigger its toxicity. All adverse events are positively correlated with dose and duration of exposure and are usually reversible with a dose decrease or by discontinuing treatment. Long-term adverse effects on the thyroid, parathyroid gland and kidney are observed with continuous high-dose treatment. Tolerance may be low in some patients due to the moderate side effects, including gastrointestinal symptoms such as dyspepsia, nausea, vomiting, and diarrhea, as well as weight gain, hair loss, acne, tremor [29, 32].

Lithium as a therapeutic agent in neurodegenerative diseases other than AD has also been widely studied, particularly in Huntington’s disease, dementia other than AD, Parkinson’s disease, Alexander’s disease, Down syndrome, Kleine-Levin syndrome, fragile×syndrome, amyotrophic lateral sclerosis, Machado-Joseph disease, and prion disease [6]. In most of these clinical investigations, lithium was administered at a dose ranging from 150 to 1200 mg per day which can lead to serious side effects, including renal toxicity. Results have increased toxicity awareness and a desire to develop low dose lithium formulations to provide therapeutic efficacy without side effects. A certain number of research groups [9 , 19] have studied the effect of micro doses of lithium in the treatment of neurodegenerative disease and observed very good long-term tolerance. Clinical trials aiming to demonstrate the efficacy of low dose lithium in the treatment of AD, used doses between 150 and 600 mg/day targeting a circulating concentration of 0.25–0.8 mEq/L.

The narrow therapeutic index being difficult to manage in clinical practice led to a general decrease in lithium use, particularly in frail patient populations such as the elderly, often more subject to renal disorders, polypharmacy, and dehydration and therefore more sensitive to lithium toxicity [33]. On the other hand, the narrow therapeutic index could also explain some inconsistency in the positive outcomes observed with lithium in neurodegenerative diseases, as the use of a micro dose that fell just below the plateau would preclude the pharmacological benefit.

ONGOING STUDIES ON THE USE OF LITHIUM IN ALZHEIMER’S DISEASE

Lithium still holds great promise in the treatment of neurodegenerative diseases and its potential benefits for AD continue to motivate researchers and pharmaceutical companies in their investigations:

CONCLUSIONS AND CLINICAL PERSPECTIVES [6 , 48–50]

An estimated minimum of 55 million people worldwide currently suffer from dementia and this number is set to rise to 139 million by 2050. AD is the most commonly diagnosed form of dementia, accounting for an estimated 60% to 80% of cases. NPS, even at early stages of the disease, are the most disruptive aspect for patients, caregivers, and healthcare systems, for which there remains no safe and efficacious drug treatment. This represents an urgent unmet need, particularly when research suggests that interventions to reduce or eliminate AD could significantly reduce the prevalence of dementia in coming decades, and that even a modest delay in disease onset could generate billions in savings to society. The pathophysiology of AD initially involves the misprocessing of amyloid precursor protein leading to the creation of toxic peptides that assemble into oligomers themselves forming deposits or plaques over time, ultimately leading to synaptic degeneration and the emergence of tau phosphorylation inside neurons. Eventually sufficient loss of neurons leads to dysconnectivity in key cognitive networks and the emergence of symptoms. Targeting amyloid protein has been the field of focus and yet despite over two decades of extensive research and investment in AD therapeutic development, only two therapeutics have come to market.

The two anti-amyloid therapies recently approved in the United States offered a maximum 25% slowing of decline, enough to give study participants a few extra months of independent living. Both must be infused and require close monitoring due to the potentially life-threatening amyloid-related imaging abnormalities. These successful developments represent a pivotal milestone for AD treatment research. Results so far suggest that these anti-amyloid therapies appear most likely to impact a certain subset of patients if delivered early enough, thereby highlighting the need for earlier diagnosis and proper patient identification. Considering the involvement of numerous intertwined pathways beyond and independent of amyloid accumulation in AD pathophysiology, it seems likely that anti-amyloid therapies will be used in combination with other drugs developed in the future to tackle different stages of the disease. While it is still too early to assess the impact of the monoclonal amyloid antibodies on daily clinical practice in early AD, it is now possible to design clinical trials where patients are first treated with a monoclonal amyloid antibody, then randomized to an anti-tau agent. This type of combination therapy in sequence is being tested in the DIAN (Dominantly Inherited Alzheimer Network) cohort, using an anti-tau antibody [51]. Other putative anti tau agents such as lithium salts could be considered in future clinical trials.

Administration of lithium-salts for the treatment of psychiatric disorders has now exceeded nine decades in the history of medicine. Its safety profile, although complex, has been well studied and optimized. The multimodal action of lithium in AD detailed above allows it to impact a variety of pathophysiological mechanisms (prevention of Aβ accumulation, prevention of Tau phosphorylation, anti-inflammatory effect, etc.) ultimately achieving a certain homeostasis and finely-tuned dysregulation of the AD disease cascade.

In this literature review, we have outlined the potential of lithium as a therapeutic option in AD, for the treatment of neuropsychological symptoms and as a potential disease-modifying treatment. However, the clinical translation of this potential has yet to be achieved. The challenges and limitations imposed by the narrow therapeutic index, especially in the elderly population with dementia, have greatly impacted its broader development in neurodegenerative disorders. However, the interesting new approaches discussed in this paper could overcome these obstacles. It appears that in order to clearly demonstrate a disease-modifying potential of lithium, a year-long treatment initiated early in the disease course and using micro doses to avoid long term toxicity would be needed. The lack of evidence of a clear benefit on NPS, could be explained by a need to properly identify the set of NPS that might most benefit from lithium treatment.

Precision medicine, also known as personalized medicine, is concerned with finding the right treatment for the right patient at the right time. Recent developments in profiling patients based on their NPS with mild behavioral impairment criteria suggests [48] that it might represent a good novel marker of preclinical disease. Perhaps a subset of patients with early disease and NPS would benefit most from lithium treatment. Lithium would also represent a good candidate for combination therapy, as it targets different mechanisms to those targeted by the anti-amyloids; especially with the optimized formulations developed, bypassing the metabolism pathways source of to drug-drug interaction.

Lithium clearly has a future in AD therapeutics but in order to develop to its full potential it will be necessary to overcome long-standing misbelief surrounding its safety profile and rigorously exemplify the benefits.