Strategies for Continued Successful Treatment in Patients with 
Alzheimer’s Disease: An Overview of Switching Between Pharmacological Agents

Abstract

Introduction:

Alzheimer’s disease (AD) is the most common cause of dementia, characterized by a progressive decline in cognition and function. Current treatment options for AD include the cholines-terase inhibitors (ChEIs) donepezil, galantamine, and rivastigmine, as well as the N-methyl-D-aspartate receptor antagonist memantine. Treatment guidelines recommend the use of ChEIs as the standard of care first-line therapy. Several randomized clinical studies have demonstrated the benefits of ChEIs on cogni-tion, global function, behavior and activities of daily living. However, patients may fail to achieve sus-tained clinical benefits from ChEIs due to lack/loss of efficacy and/or safety, tolerability issues, and poor adherence to the treatment. The purpose of this review is to explore the strategies for continued successful treatment in patients with AD.

Methods:

Literature search was performed for articles published in PubMed and MEDLINE, using pre-specified search terms. Articles were critically evaluated for inclusion based on their titles, abstracts, and full text of the publication.

Results and Conclusion:

The findings of this review indicate that dose up-titration and switching between ChEIs may help to improve response to ChEI treatment and also address issues such as lack/loss of effica-cy or safety/tolerability in patients with AD. However, well-designed studies are needed to provide robust evidence.

1. Introduction

Alzheimer’s disease (AD) is a neurodegenerative disorder that is often associated with aging [1]. AD is the most prevalent cause of dementia and is characterized by progressive decline in cognition and global function, thereby affecting activities of daily living (ADL) [2, 3].

Management of AD continues to remain a challenge for both patients and their caregivers given that the available pharmacological treatment options are only able to provide symptomatic relief. Tacrine, a potent cholinesterase inhibitor (ChEI), was the first drug approved for the treatment of AD. However, its use was discontinued due to a poor safety profile, particularly hepatotoxicity [4]. Currently, the mainstay of treatment approved worldwide for AD includes ChEIs (e.g. donepezil, galantamine, and rivastigmine) and an N-methyl-D-aspartate receptor antagonist (memantine). In addition, many new symptomatic treatment options are under development [5].

Besides symptomatic treatments, many disease-modifying therapies (DMTs) which aim to halt or reverse disease progression, are under extensive research. These DMTs aim to prevent Aβ aggregation, promote Aβ clearance, or target Tau phosphorylation, which are considered the pathogenic mechanisms leading to neuronal death. To date, all Phase III studies evaluating the efficacy of the available DMTs have failed. It may be some time before the first DMT shows proven clinical efficacy and becomes available as an approved treatment. Until the emergence of new therapies (including new symptomatic drugs), the mainstay for AD treatment is limited to the currently available drugs which may slow the progression of symptoms. Even after the emergence of DMTs, symptomatic therapies may still be a viable treatment option for patients with AD experiencing disease progression, therefore the argument could be made that it becomes imperative that these drugs must be used as optimally as possible to maximize the potential clinical outcomes in patients with AD [6].

In this review, we discuss the available therapeutic options for symptomatic treatment of AD, focusing on dose up-titration and switching among approved treatment options for optimal treatment outcomes. In addition, we have attempted to discuss the rationale of within-class switching based on the pharmacokinetics (PK) and pharmacodynamics (PD) of drugs and patient-related factors.

2. Methods

A literature search was performed for articles in the English language on AD, published in PubMed and MEDLINE, using the search terms “Alzheimer’s disease/Alzheimer’s dementia”, “cholinesterase inhibitors”, “donepezil”, “galantamine”, “rivastigmine”, “switch”, “clinical”, “efficacy”, and “safety” (cut-off date: January 2017). All the studies retrieved from this search were critically evaluated for inclusion based on their titles, abstracts, and full text of the publications.

3. Approved treatment options for AD

3.1. ChEIs

Donepezil is indicated for the treatment of all stages of AD in the United States and Japan, and for mild-to-moderately severe AD in Europe [7-10].

Galantamine is indicated for the treatment of mild-to-moderate dementia of the Alzheimer’s type in the US and Japan, and for mild-to-moderately severe dementia of the Alzheimer’s type in Europe [11-13].

Rivastigmine is available as a transdermal patch and as an oral formulation [14-17]. Oral rivastigmine is approved in the US for the treatment of mild-to-moderate dementia of the Alzheimer’s type, and in Europe for mild-to-moderately severe Alzheimer’s dementia [14, 16]. The rivastigmine patch is approved in the US across all stages of AD, in Europe for the treatment of mild-to-moderately severe AD, and in Japan for mild-to-moderate AD [15, 17-19].

3.2. N-methyl-D-aspartate Receptor Antagonists (Memantine)

Memantine is approved for the treatment of moderate-to-severe dementia of the Alzheimer’s type in the US, Europe and Japan [20-22].

The approved treatment options for AD have been described in detail in Table 1.

Table 1.

Approved treatment options for Alzheimer’s disease.

Compound	Geographical area	Approved indication	Titration scheme	Additional remarks (if any)
Donepezila	US	All stages of AD	5 mg/day for 4-6 weeks -->10 mg/day for at least 3 months--> 23 mg/day	A 23 mg sustained-release tablet formulation is approved in the US for treatment of moderate-to-severe AD; administered once patients have been on a dose of 10 mg o.d. for at least 3 months.
	EU	Mild-to-moderately severe AD	5 mg/day for 1 month -->10 mg/day	-
	Japan	Mild-to-severe AD	3 mg/day for 1-2 weeks -->5 mg for at least 4 weeks--> 10 mg/day*	*10 mg approved only for severe AD.
Galantamineb	US	Mild-to-moderate AD	4 mg b.i.d for 4 weeks-->8 mg b.i.d. over at least 4 weeks. Dose may be increased up to 12 mg b.i.d., if tolerated, after a minimum of 4 weeks at 8 mg b.i.d	-
	EU	Mild-to-moderately severe AD
	Japan	Mild-to-moderate AD
Rivastigminec	US	Oral: Mild-to-moderate AD Patch: All stages of AD	Oral: 1.5 mg b.i.d. --> 3 mg b.i.d. --> 4.5 mg b.i.d.--> 6 mg b.i.d., if tolerated with a minimum of 2 weeks at each dose Patch: 4.6 mg/24h -->9.5 mg/24h --> 13.3 mg/24h, if tolerated for a minimum of 4 weeks at each dose*	A target maintenance dose of 9.5 mg/24h patch or 6 mg b.i.d. oral rivastigmine has been approved in most countries for the treatment of mild-to-moderate AD. *In the EU, patients may be switched to the 13.3 mg/24 h patch only after a minimum of 6 months of treatment with the 9.5 mg/24 h patch.
	EU	Oral and patch: Mild-to-moderately severe AD
	Japan	Patch: Mild-to-moderate AD	4.5 mg --> increasing dose by 4.5 mg at 4-week intervals up to the target size of 18 mg. or 9 mg --> 18 mg after 4 weeks. (9 mg=4.6 mg/24 h, 18 mg=9.5 mg/24 h)	-
Memantined	US	Moderate-to-severe AD	5 mg o.d.--> 10 mg o.d.--> 15 mg o.d. --> 20 mg o.d. memantine, if tolerated for a minimum of 1-week at each dose	In the US, memantine is available in extended release capsule form and is administered at an initial dose of 7 mg o.d. and increased in 7-mg increments to reach a maintenance dose of 28 mg o.d. with a minimum treatment period of 1 week at each dose level.
	EU
	Japan

^a7-10; ^b11-13; ^c14-19, 32; ^d20-22. AD, Alzheimer’s disease; b.i.d., twice daily; o.d., once daily.

4. Initiation of Treatment with CHEIS

Guidelines by the European Federation of Neurological Societies (EFNS) and National Institute for Health and Excellence (NICE) recommend ChEIs as the standard of care for AD. They recommend treatment initiation at a lower dose, with gradual up-titration to higher approved doses for optimal treatment outcomes. Patients can be switched to other ChEIs based on the adverse event (AE) profile, adher-

ence, possibility of drug interactions, and dosing profiles [23, 24]. Similarly, guidelines developed at the 4^th Canadian Consensus Conference on the Diagnosis and Treatment of Dementia recommend treatment with ChEIs for patients with mild-to-severe AD (grade 1A) [25]. Moreover, guidelines in Japan recommend the use of ChEIs as the initial treatment for AD; switching is preferred among the ChEIs in case of intolerability and/or lack of efficacy [26].

5. Dose Up-titration

To maximize the therapeutic benefit of treatment with ChEIs, there is growing scientific evidence that the treatment dose should be up-titrated and tailored to individual patients’ needs based on their disease stage and other clinical characteristics [27]. ChEIs have been reported to show significant benefits on cognitive, global, functional, and behavioral outcomes in a dose-dependent manner in clinical studies [27-31].

The ACTION (ACTivities of daily living and cognitION in Patients with Severe Dementia of the Alzheimer’s Type) study (N=716) was a 24-week, randomized, double-blind study that compared the efficacy, safety, and tolerability of rivastigmine patches (13.3 mg/24h and 4.6 mg/24h) in patients with severe AD. The high-dose patch demonstrated significantly less decline in overall cognition (p<0.0001) and function (p=0.025) as compared with the 4.6 mg/24h patch [32]. In a 24-week, open-label extension of the ACTION study (N=397), there were no clinically relevant differences in safety and tolerability between patients who were up-titrated from the 4.6 mg/24h rivastigmine patch to receive the 13.3 mg/24h patch, and patients who continued on the 13.3 mg/24h patch. However, a greater decline was observed in patients with delayed up-titration to the high-dose 13.3 mg/24h patch compared to patients who continued on the high-dose patch from the double-blind phase to the extension phase [33]. This indicates the importance of early and sustained intervention with the high-dose patch to achieve maximum clinical benefit. In another 24-week, double-blind study, Japanese patients (N=859) were randomized to receive either the 4.6 mg/24h or 9.5 mg/24h rivastigmine patch or placebo. Patients receiving the 9.5 mg/24h patch reported delayed deterioration on the Japanese version of the Alzheimer’s Disease Assessment Scale-cognitive subscale (ADAS-J cog; p=0.005) and the Clinician’s Interview-Based Impression of Change plus Caregiver Input (CIBIC plus-J; p=0.067) [19]. Furthermore, in a global Phase III study conducted in patients with moderate-to-severe AD (N=1467), patients receiving donepezil 23 mg/day showed significantly greater cognitive benefits, as assessed by the Severe Impairment Battery (SIB) score, compared to patients who continued treatment with donepezil 10 mg/day (p<0.001) [28, 34]. Another study conducted in 61 Japanese patients reported no statistically significant difference in cognitive decline at any time after starting donepezil 10 mg/day [35]. These observations clearly highlight the dose-dependent efficacy of ChEIs. Therefore, it is suggested to maintain ChEI therapy on a higher dose, as long as it is tolerable, as it may provide a greater chance of slowing/delaying symptomatic disease progression [36].

6. Switching between CHEIS

Although continuous dose optimization is an option for treating all stages of AD, certain patients may fail to achieve sustained clinical benefits from ChEIs, sometimes resulting in discontinuation of the treatment. In these patients, switching between ChEIs is a reasonable therapeutic option because it is crucial to not give up on treatment after the first therapy has failed owing to a lack of clinical benefit [37, 38]. In a multicenter, 2-year prospective study, the incidence of switching between ChEIs was 9.2 per 100 person-years among 611 patients treated at baseline [39].

The rationale for switching to alternative ChEIs is based on the lack of clinical response owing to inappropriate drug distribution and the complex molecular mechanisms involved in the changes occurring in the brain. Moreover, differences in individual pharmacological properties of ChEIs make switching between ChEIs an attractive option. Thus, patients who are not able to tolerate or benefit from one ChEI may tolerate or benefit from another [38]. Most published studies have explored switching from donepezil to galantamine or rivastigmine. However, only a few studies have investigated the switch from galantamine to donepezil or rivastigmine and from rivastigmine to donepezil. In the subsequent sections of this article, we focus on the effects of switching from one ChEI to another owing to lack of efficacy, tolerability, or adherence. Studies investigating switching have been summarized in Table 2.

Table 2.

Studies showing switching options as a therapeutic strategy for Alzheimer’s disease.

Reference	Study Design	Switch Type	Total Number of Patients; Duration	Results
Auriacombe et al, 2002a	Open-label, prospective	DPZ to oral RVG	382; 6 months	Oral rivastigmine well-tolerated 56.2% stabilized or improved on the CGIC 48.9% stabilized or improved on the MMSE 57% stabilized or improved on the IADL scale
Edwards K et al. 2004b	Retrospective chart review	DPZ/RIV to GAL	16; 6 months	50% of patients reported stabilization/improvement in cognition, behavior and ADL after switch
Wilkinson DG et al. 2005c	Double-blind, open-label	DPZ to GAL	105; 52 weeks	Galantamine was generally well-tolerated; no change in cognitive performance with either a 4-day or 7-day washout period
Bartorelli et al. 2005d	Observational, prospective	DPZ to RVG oral; GAL to RVG oral	225, 3 months	66.7%-67.7% of patients stabilized or improved after the switch
Sadowsky et al. 2005e	Open-label	DPZ to RIV	61; 28 days	Rivastigmine was well tolerated after switching from donepezil without a wash out period
Figiel et al. 2008f	Open-label	DPZ to RVG	270; 26 weeks	69.7% patients showed improvement or no further decline in global functioning
Grossberg et al. 2009g	Randomized, controlled	Oral RIV to RIV patch	870; 28 weeks	Switching to the rivastigmine patch was well tolerated; ≤2.5% reported nausea and ≤1.9% reported vomiting
Sadowsky et al. 2009h	Open-label, prospective	DPZ+/-MEM to RVG patch	261, 5 weeks	Both immediate and delayed switches were well-tolerated with similar rates of discontinuation
Sadowsky et al. 2010i	Open-label, prospective	DPZ to RVG patch	234, 25 weeks	Both immediate and delayed switches were well tolerated. Cognitive, behavioral and global outcomes were maintained in both groups
Han HJ et al. 2011j	Open-label, prospective	Oral ChEIs to RVG patch	164; 24 weeks	82.8% and 64.3% of patients reported improvement or no decline on CGIC and the Korean version of MMSE scores, respectively
Tian et al. 2013k	Retrospective cohort study	DPZ to RVG patch	772, 12 months	Adherence was slightly improved in patients who switched from oral donepezil to rivastigmine patch
Sasaki and Horie 2014l	Outpatient	DPZ to GAL	44; 3 months	NPI scores improved significantly on BPSD Significant improvement in patients with moderate AD
Spalletta G et al. 2014m	Observational, longitudinal	Oral ChEIs to RVG patch	423; 6 months	Switching from oral ChEI to the rivastigmine patch showed favorable effects as compared to those switching from the rivastigmine patch to oral ChEI
Cagnin A et al. 2015n	Observational, prospective	Oral ChEIs to RVG patch	174; 6 months	56% of patients stabilized or increased the MMSE score as compared to baseline

^a50; ^b48; ^c61; ^d40; ^e58; ^f43; ^g56; ^h41; ⁱ57; ^j46; ^k55; ^l47; ^m45; ⁿ44. AD, Alzheimer’s disease; ADL, activities of daily living; BPSD, behavioral and psychological symptoms of dementia; CGIC, clinical global impression of change; ChEIs, cholinesterase inhibitors; DPZ, donepezil; GAL, galantamine; IADL, Instrumental ADL; MEM, memantine; MMSE, mini-mental state examination; NPI, neuropsychiatric inventory; RVG, rivastigmine.

6.1. Switching Owing to Lack of Efficacy

Lack of efficacy is defined as significant deterioration despite the use of symptomatic medication at an effective dose for at least 6 months. In this case, switching can be performed overnight with quicker titration until the minimal effective dose is reached [37].

A prospective, multicenter, 3-month observational trial in patients with mild-to-moderately severe AD (N=225) was conducted to determine the response to switching from another ChEI to rivastigmine when patients experienced deterioration (loss of at least two Mini-Mental State Examination [MMSE] points in the past 6 months) with initial treatment. A total of 188 patients switched from donepezil to rivastigmine, 33 switched from galantamine to rivastigmine, and four switched from donepezil to galantamine. Overall, 67.7% and 66.7% of patients in the donepezil-rivastigmine and galantamine-rivastigmine switch groups, respectively, responded to rivastigmine (Clinical Global Impression of Change [CGIC] score ≤4). Among the non-responders, >80% of patients had minimal worsening of the disease (CGIC score 5). MMSE scores also improved after switching from both donepezil (p=0.008) and galantamine (p=0.05) to rivastigmine; however, this was observed when patients with an absolute change from baseline in MMSE score >5 were excluded [40].

A 5-week core-phase of the prospective, parallel-group, open-label SWAP study (SWitch from Aricept to Patch) evaluated the effects of “immediate” (n=131) or “delayed” (n=130) switching from 5-10 mg/day donepezil tablets to the 4.6 mg/24h rivastigmine transdermal patch following a 7-day withdrawal period [41]. This was followed by a 20-week extension phase with 9.5 mg/24h rivastigmine transdermal patch treatment [42]. Results from this study suggest that both switching strategies were well-tolerated. Global function remained stable during the course of the study and the mean change in CGIC scores was similar in both the immediate and delayed switch groups (4.1; 95% confidence interval [CI], 3.9-4.4 and 4.3; 95% CI, 4.1-4.4, respectively). At the end of the study, cognitive, behavioral, or global outcomes were maintained in the switch groups with a modest decline in the Alzheimer's Disease Cooperative Study-Activities of Daily Living Scale (ADCS-ADL) scores at Week 25 (-3.7; 95% CI, -4.9 to -2.4; p<0.0001). Most patients (55%) preferred the rivastigmine patch to the oral formulation. Results from this study indicate that the majority of patients receiving donepezil may be switched safely to the rivastigmine patch without a withdrawal period [42]. In another, 26-week open-label study (N=270), patients with mild-to-moderate AD not responding to donepezil were switched to receive rivastigmine 3-12 mg/day. Improvement/stabilization of AD was reported in 69.7% of patients (136/195, observed case analysis) who did not respond to the treatment or declined while taking donepezil and were immediately switched from donepezil to rivastigmine [43].

In a 6-month observational study conducted in 174 patients with AD who switched from oral ChEIs to the 9.5 mg/24h rivastigmine patch, the MMSE score increased or stabilized in 56% of patients as compared to baseline. The main reasons for switching were lack/loss of efficacy with previous oral ChEIs (57%), tolerability concerns (33%), or both (10%) [44]. Another study (EVOLUTION, N=423) assessing the effectiveness of switching in patients with mild and moderate AD showed favorable effects of switching from oral ChEIs to the rivastigmine patch as compared to those switching from the rivastigmine patch to oral ChEIs. In this study, the reasons for switching therapy included loss of efficacy with previous ChEIs (41.4%), lack of response (28.8%), reduced tolerability (14.2%), and poor compliance to treatment (9.9%) [45]. A 24-week, prospective, open-label, single-arm, multicenter study in patients with probable AD (N=164) was conducted to assess the effects of switching to the rivastigmine patch after a poor response from initial treatment with oral ChEIs (donepezil, galantamine or rivastigmine capsules). Poor response was defined as a decrease of at least two points on the Korean version of the MMSE (K-MMSE). At Week 24, 82.8% and 64.3% of patients reported an improvement or no decline on the CGIC and K-MMSE scores, respectively. Poor responders to oral ChEIs experienced improvement in symptom or disease stabilization when switched to the rivastigmine patch [46].

The efficacy of galantamine in patients with mild cognitive impairment/mild-to-moderate AD, who switched from donepezil to galantamine without a washout period or dose titration was elucidated in an outpatient study by Sasaki and Horie. Neuropsychiatric inventory scores improved significantly on behavioral and psychological symptoms of dementia, particularly in terms of delusions, agitation, and aberrant motor activity in patients with AD (p=0.027). Remarkable improvements were noted in patients with moderate AD (MMSE scores 10-19; p=0.007) while no improvement was noted in patients with mild cognitive impairment (MMSE scores ≥24; p=0.648) [47]. A retrospective chart review of 16 patients with AD who had been on donepezil or rivastigmine and were switched to galantamine (4 mg/day, n=5; 8 mg/day, n=7; 12 mg/day, n=4) reported stabilization/improvement in cognition, behavior and ADL in 50% of patients even 6 months after switching treatment [48]. Another 52-week observational study evaluated the efficacy of galantamine on cognition in patients with mild-to-moderate AD, who were either naïve to ChEIs (naïve group, n=42) or failed to respond to donepezil and were switched to galantamine (switch group, n=24). At the end of the study, no significant difference between naïve and switch groups was observed on the Korean version of ADAS-cog (p=0.162). However, the results of the study suggest that, as the efficacy of galantamine on cognition was not inferior in the switch group than in the naïve group, switching between ChEIs may be a viable option for patients not responding to treatment [49].

6.2. Switching Owing to Lack of Safety and Tolerability

In case of intolerance to initial therapy, it is recommended to wait for complete resolution of side effects before switching to another treatment [37]. Efficacy and safety of switching from donepezil to oral rivastigmine in patients not responding or not tolerating donepezil was evaluated in a prospective study with AD patients (N=382). A total of 74.4% and 54.5% of patients who were switched because of tolerability issues and lack of efficacy with donepezil (as assessed by CGIC), respectively, responded to rivastigmine. Discontinuation due to AEs with rivastigmine in patients who experienced tolerability problems or lack of efficacy with donepezil was 15.4% and 11.5%, respectively. Nausea (30.1%) and vomiting (14.1%) were the most common AEs reported during the study [50].

ChEIs are associated with a range of side effects, including gastrointestinal symptoms, particularly during the dose titration phase [51, 52]. It is believed that these side effects are related to high maximum concentration (C_max) and a short time to C_max (t_max) following oral administration, thereby leading to an increase in acetylcholine levels in the brain and periphery. Thus, strategies that lower C_max and prolong t_max may be expected to improve the tolerability of ChEIs [51, 53]. The transdermal patch formulation may be beneficial for AD patients who are being switched from the initial treatment because of tolerability reasons.

In a 6-month observational study (N=174) evaluating the effectiveness of switching from oral ChEIs to the rivastigmine patch, 56% of patients reported improvement or no further deterioration of the MMSE score compared to baseline. The main reasons for switching were lack/loss of efficacy (57%), tolerability problems (33%) or both (10%). The most frequently reported AEs were skin reactions (16%) and gastrointestinal symptoms (7%); only 9% and 3% of patients discontinued because of these AEs, respectively [44], indicating that the rivastigmine patch can be considered as a therapeutic strategy to improve treatment persistence [44, 50]. This is consistent with the observation in the double-blind randomized controlled trial determining efficacy and safety of the rivastigmine patch and capsule in which one-third of gastrointestinal AEs were observed with the patch compared with the capsule [54].

In the 5-week core phase of the SWAP study, patients were randomly assigned 1:1 to either an immediate switch or delayed switch group (7-day withdrawal period) from 5-10 mg/day donepezil to the 4.6 mg/24h rivastigmine patch. Both an immediate or delayed switch from donepezil to rivastigmine was safe and well-tolerated in patients with mild-to-moderate AD. Only 3.8% of patients from the immediate switch group and 0.8% from the delayed switch group who received the rivastigmine patch reported nausea. Furthermore, no discontinuations were reported due to nausea and vomiting. Results from this study indicated that most patients were able to switch directly to the rivastigmine patch without a washout period [41]. In addition, data from the extension-phase of the study revealed similar results. At least one AE (from application site reaction, agitation and fall) was reported in 70.5% (184/261) of the overall patient population, with a greater number of patients from the immediate switch group (n=96 [73.3%]) in comparison with the delayed switch group (n=88 [67.7%]). Only ten (3.8%) and 11 (4.2%) patients experienced both nausea and vomiting in the immediate and delayed switch groups, respectively [42].

Studies investigating the switch from donepezil to the rivastigmine patch indicate that this transition has a good safety profile, is well tolerated and can be performed without a washout period [50, 55].

The IDEAL (Investigation of TransDermal Exelon in ALzheimer’s disease) study was a 24-week randomized controlled trial evaluating efficacy and safety of rivastigmine patches (9.5 mg/24h and 17.4 mg/24h) versus oral rivastigmine (3-6 mg b.i.d.) or placebo. All patients who completed the double-blind phase were eligible to enter the 28-week extension phase of the study. These patients (N=870) were switched directly to the 9.5 mg/24h patch, irrespective of their treatment during the double-blind phase, and up-titrated to the 17.4 mg/24h patch. During the first four weeks of the extension phase, the rivastigmine patch was well tolerated by patients previously randomized to rivastigmine; nausea and vomiting were reported in ≤2.5% and ≤1.9% of the patients, respectively [56]. The rivastigmine patch demonstrated good skin tolerability in both the IDEAL and SWAP studies [42, 56].

A post-hoc analysis of data from three clinical trials [41, 44, 58] conducted by Sadowsky CH et al. compared the tolerability of switching from donepezil to the rivastigmine patch (4.6 mg/24h) or rivastigmine capsules (3-12 mg/day). Results from this analysis indicate better tolerability of the rivastigmine patch versus rivastigmine capsules; there were fewer gastrointestinal AEs (nausea: 3.8% versus 32.9%; vomiting: 4.2% versus 24.1%, respectively) and discontinuations due to AEs (14.6% versus 19.3%, respectively) with the patch [57]. Evidence from an open-label study suggests that switching patients from donepezil to rivastigmine without a washout period is well tolerated [58]. In a more recent observational, retrospective, multicenter study conducted in patients with AD (N=1022), improved ease of administration (56.65%), tolerability (36.79%), efficacy (31.60%), and adherence (18.59%) were the main reasons for switching to the rivastigmine patch from any oral ChEI [59]. Based on the literature review conducted by Sadowsky C, et al., it has been recommended that patients receiving a high dose of oral rivastigmine can be switched directly to the 9.5 mg/24h rivastigmine patch, whereas those on the lower doses of oral rivastigmine should be switched to the 4.6 mg/24h patch and continue treatment for 4 weeks before up-titration to a high-dose patch [60]. A double-blind study (N=105) conducted by Wilkinson et al. investigated the switch from donepezil to galantamine and explored the optimum length of a washout period, given the longer half-life of donepezil than galantamine. Results from this study suggest that galantamine was generally well-tolerated; however, patients reported fewer gastrointestinal AEs if the washout period was 4 days rather than 7 days [61].

To date, no published studies have assessed the clinical effects of switching patients from galantamine or rivastigmine to donepezil because of the lack/loss of efficacy or safety/tolerability issues [62]. From these reported clinical observations, it seems that within-class switch among ChEIs is valuable when the first prescribed ChEI is not tolerable or efficacious. Additionally, switching from oral ChEIs to the rivastigmine patch seems to be well-tolerated.

7. Factors affecting clinical response to CHEIS

In this section, we discuss differences in the pharmacokinetic and pharmacodynamic characteristics of ChEIs and how they may explain the benefit of within-class switching. These differences can be attributed to the diversity of the chemical structures, each of which belongs to an independent chemical class [63] (Fig. 1). Results from meta-analyses described in systematic review papers indicate that there is no difference amongst ChEIs in terms of cognition, ADL or global functions in patients with AD [64]. This may be true when considering collective treatment response in a patient population, but it may not be necessarily true in an individual patient as patients with AD have diverse characteristics and may respond differently to treatment.

Fig. (1).

Chemical structures of cholinesterase inhibitors.

7.1. Pharmacodynamics

Donepezil and galantamine function as rapidly reversible selective inhibitors of acetylcholinesterase (AChE), whereas rivastigmine is a slowly-reversible dual inhibitor of AChE and butyrylcholinesterase (BuChE) [65]. Galantamine, a relatively weak AChE inhibitor (half maximal inhibitory concentration [IC₅₀] value of ~2.8-3.9 µM), appears to have similar clinical efficacy as that of donepezil (IC₅₀: 15-24 nM) and rivastigmine (IC₅₀: 4 nM) and this effect may be attributed to its allosteric potentiation ligand (APL) activity. It increases the probability of nicotinic acetylcholine receptor (nAChR) channel opening induced by nicotinic agonists and potentiates the agonist response of the nAChR subtypes [66]. The levels of nAChR subtypes are significantly reduced in patients with AD, compared with age-matched controls [67].

Rivastigmine exhibits a unique pharmacological property that enhances potency and selectivity in a different manner via dual AChE-BuChE inhibition. The potential clinical relevance of BuChE inhibition by rivastigmine in patients with AD has been demonstrated by the dynamic shift of cholinesterase activity. Following 12 months of treatment with 3-12 mg/day rivastigmine in patients with mild AD, inhibition of AChE and BuChE activities in the cerebrospinal fluid was reported to be 45% and 58%, respectively [68]. In another 13-week open-label study in patients with mild-to-moderate AD, rivastigmine was associated with a decrease in AChE and BuChE activity by 42.6% and 45.6%, respectively [69].

Donepezil is a highly specific AChE inhibitor, designed to exhibit very high selectivity for AChE [70]. Donepezil also shows high affinity for the sigma-1 receptor as well, which is believed to play a role in the pathophysiology of neuropsychiatric diseases, including AD [71]. Although it is not obvious if rivastigmine or galantamine show similar affinity for the sigma-1 receptor, the pleiotropic effect of donepezil may exhibit a unique pharmacological profile other than AChE inhibition.

7.2. Pharmacokinetics

Differences in route of administration (oral/ transdermal) or metabolism may influence the efficacy and safety of a drug. Orally administered drugs are absorbed through the gastrointestinal wall, and their plasma drug-concentration levels increase rapidly to the peak (C_max). Drug plasma levels then drop and may reach their lowest level (C_min) until the next dose is administered [72]. Larger and more frequent plasma level fluctuations may lead to an increased incidence of cholinergic side effects, including gastrointestinal AEs such as nausea and vomiting [73]. Thus, by reducing C_max and slowing t_max, the occurrence of these AEs may be reduced and may help in achieving sustained efficacy [72, 73]. To achieve such a pharmacokinetic profile, a novel rivastigmine transdermal delivery system was developed. The patch shows a steady rivastigmine concentration-time profile as compared with the oral formulations. Systemic exposure (area under the concentration-time curve from zero to infinity [AUC_∞]) for the 9.5 mg/24h patch was approximately five times higher than that with the oral dose of 3 mg rivastigmine, whereas C_max with the patch was 20% lower than that observed with the oral solution [53]. Rivastigmine is metabolized to an inactive metabolite, NAP-226-90, by AChE and BuChE, with no involvement of the cytochrome P-450 (CYP) system. Conversely, donepezil and galantamine are metabolized primarily by these enzymes [74]. This implies that rivastigmine has fewer clinically relevant drug-drug interactions, making it ideal for use in the elderly population being treated with multiple medications for numerous comorbidities [75].

7.3. Patient-Related Factors

CYP2D6 is the key CYP-450 isoenzyme involved in the metabolism of donepezil and galantamine. The genetic polymorphism in CYP2D6 has been well investigated, and a large number of allelic variants are known, which may have been reported to be responsible for decreased, increased, or no enzyme activity. In a 6-month study, CYP2D6*10, a mutant genotype, was found to be associated with a better response to donepezil treatment in patients with AD. Moreover, the steady-state plasma concentration of donepezil in patients with mutant genotypes was higher than that in patients with the wild-type genotype [76]. Another study showed that patients with the single-nucleotide polymorphism (SNP) rs1080985 in the CYP2D6 gene had a rapid metabolism and hence a poor response to donepezil. This rapid metabolism is attributed to high enzymatic activity as a consequence of a higher gene expression associated with the G allele. Hence, the presence of this SNP may influence clinical response to donepezil in patients with AD [77]. There are no reports describing the effect of this particular SNP in the CYP2D6 gene on metabolism and/or efficacy of rivastigmine, as rivastigmine is not metabolized by the CYP-450 isoenzymes [74].

Differential effect of drugs in patients with AD may also be attributed to the SNPs of AChE. In a computational study conducted by Saravanaraman et al., results from molecular dynamics and docking study revealed that various non-specific SNP forms of AChEs exhibit different dynamic properties, which in turn affects their ligand-binding properties. Of the reported 153 SNPs, four non-specific SNPs (A415G, P104A, V302E, and Y119H) that were predicted to be functionally unfavorable were found to be structurally stable. However, their conformational alterations were found to interfere with the binding of AChE inhibitors, suggesting it to be a reason for the differential effect of ChEIs in patients with AD [78]. Similar results were observed in another study conducted in patients with AD who underwent treatment with AChE inhibitors (N=158). Of the 25 SNPs located in 3 cholinergic system genes (CHAT, CHT and ACHE), treatment response in patients with AD was found to have significant association with two SNPs of CHAT (rs2177370, rs3793790) [79]. Moreover, there are reports describing the difference in response in patients with specific SNP of molecules related to the pleiotropic mode of actions of ChEIs. The pleiotropic APL effect of galantamine is explained by the allosteric activation of nAChR, by which galantamine potentiates ACh signals transmitted through the nAChR. The CHRNA7 gene is reported to encode α7 nAChR, one of the major nAChR subunits in the central nervous system (CNS), on chromosome 15q14. In a retrospective study conducted in 233 patients with mild-to-moderate AD, the ratio of responders to non-responders with galantamine treatment was significantly higher in women with the SNP2 of CHRNA7 rs8024987 compared with female non-carriers (p<0.01) [80]. A similar difference was not observed in patients treated with other ChEIs, suggesting the involvement of the APL effect of galantamine through α7nAChR.

Similar observations have been reported in patients carrying a BuChE SNP and their clinical response to rivastigmine. Although many SNPs have been reported for the gene, the most prevalent non-synonymous substitutions SNP of BuChE is the K-variant (A539T). Carriers of this variant have been reported to have 33% lower enzyme activity in plasma. The allele frequency of the K-variant is reported to be around 10%, with some differences due to ethnicity [81]. In a post-hoc analysis of a study in 994 patients [82], BuChE wild-type carriers younger than 75 years showed significantly greater treatment response to rivastigmine over 2 years than patients receiving donepezil. However, BuChE K-variant carriers experienced similar long-term treatment effects with both agents [83], suggesting that the greater effect of rivastigmine may be attributed to the inhibition of BuChE, which is fully expressed in the wild-type carriers. Furthermore, the efficacy of rivastigmine was also shown to be better in patients carrying wild-type BuChE compared with those with the K-variant in an open-label study in 146 patients with AD. The difference was more evident in a subpopulation carrying allele ApoE4 [84].

These findings must be carefully interpreted considering the limitation of these analyses; some are retrospective analyses of studies with different study objectives and results being observed for a specific population (e.g. female subjects or patients younger than 75 years). Further elucidation in prospective studies with sufficient statistical power is required. Together with the polymorphism in the coding genes related to the cholinergic pathway, those present in the non-cording region merits in-depth research as the non-coding microRNAs (CholinomiRs’) are believed to coordinate the cognitive and inflammatory aspects of cholinergic signaling by targeting major cholinergic transcripts including AChE [85].

In addition to the SNPs of molecules related to cholinergic pathway, some other patient conditions may also be implicated for the differential treatment response to ChEIs. BuChE expressed in the neuron and glial cells is reported to be co-localized with senile plaques in AD brain and its enzyme activity positively correlates with the number of senile plaques in human autopsy brain samples [86, 87]. Molecular interactions of BuChE with amyloid beta protein (Aβ) and the role of BuChE in neuro-inflammation have also been reported [88-90]. The inhibitory potency of various ChEIs towards BuChE may also suggest the possibility of differential treatment response to ChEIs in patients with AD [91].

Apart from the intrinsic patient factors which may influence the treatment response to ChEIs, the potential impact of concomitant medications on cholinergic system cannot be ruled out. The negative effect of anti-cholinergic agents on cognition is widely reported [92, 93]. Combination of anti-cholinergic drugs and ChEIs leads to pharmacological antagonism which results in lack/loss of efficacy of ChEIs in patients with AD [94]. Hence, caution should be exercised during concurrent use of anti-cholinergic drugs and ChEIs.

In summary, the main reasons for the potential difference in responsiveness to each ChEI can be attributed to the different chemical structure of the compounds and the diverse pathology of AD, which is not yet fully understood. Although the individual responses to each ChEI vary due to their PK/PD properties and patient factors including genetics, response of an individual patient to a particular compound with the help of biomarkers is still not predictable. Responders and non-responders can be identified only after a treatment trial period with careful observation. In this context, it is recommended to initiate therapy with any of the ChEIs, taking into account expected therapeutic benefits and potential safety issues [95]. Moreover, patients should be reviewed regularly using cognitive, global, functional and behavioral assessment(s), which may help physicians to detect lack of efficacy of a treatment.

8. Switching and patient adherence

Achieving maximal benefits from treatment is dependent upon patient adherence to the type of treatment used. A decline in cognition, mood, and behavior can pose a challenge for patient adherence to a given medication [96]. Switching may positively affect treatment adherence through enhanced motivation of patients and caregivers by the expectation of the treatment option. A retrospective cohort study examined patient adherence using “proportion of days covered” (PDC) in 772 patients with AD who were new donepezil users and were subsequently switched to the rivastigmine patch [55]. Results from the analysis indicated that adherence improved in patients switching from oral donepezil to transdermal rivastigmine. Patients who switched within the first year of initiating donepezil to rivastigmine patch exhibited increased adherence (PDC, 60.6% versus 69.3%; p=0.0004). Patients who switched from donepezil to the rivastigmine patch within the first 3 months (PDC, 80.4% versus 90.7%; p=0.04) exhibited even better adherence than those who switched between 7 and 9 months (PDC, 61.3% versus 71.0%; p=0.05). Switching after 2 years did not result in increased patient adherence. Hence, the time to switch between the rivastigmine patch and donepezil tablets was a predictor of difference in PDC. This could be attributed to better tolerability with the transdermal patch, and/or perceived convenience and ease of use of transdermal patches. It is suggested that an early decision to switch to another ChEI is beneficial for a patient’s outcome from the viewpoint of treatment persistence [55]. Improvement of tolerability by switching may also contribute to improved efficacy attributed by improved adherence and/or up titration to higher doses. Side effects of oral ChEIs, such as nausea and vomiting, are practical concerns while up-titrating to efficacious doses and may negatively affect drug adherence as well as motivation of patients and caregivers to stay on drug treatment. In a population-based cohort study using British Columbia claims data, approximately 50% of patients receiving ChEIs were reported to discontinue therapy within 12 months of treatment initiation. A total of 3231/24,526 patients (new ChEI users) switched to a second ChEI within 90 days of discontinuation of the first ChEI [97]. Similar results were reported in an Austrian cohort study (N=15,809) [98]. Considering the long disease course of AD, and the high rate

of treatment discontinuation in clinical practice, regular assessment of disease progression and appropriate switching among ChEI may present a useful measure for better clinical outcomes by avoiding discontinuation of the treatment.

Conclusion

As donepezil, galantamine, rivastigmine, and memantine are the only available therapeutic options for the treatment of patients with AD, it is vital that clinicians optimize the use of available treatments until new preventive DMTs or symptomatic medications become available. Guidelines recommend ChEIs as the first choice of treatment, particularly for mild-to-moderate AD. Based on the disease stage and clinical characteristics, the therapeutic dose should be up-titrated to the maximum approved dose, as long as it is tolerable, as it may help to improve response to ChEI treatment. Switching between ChEIs may also help to address issues such as lack/loss of efficacy or safety/tolerability in patients with AD. However, most of the switching studies referred to in this review are of an open-label design with potential bias, therefore well-designed studies are needed to provide robust evidence. In addition, future therapies will need to address multiple aspects of AD, for example, different pathogenic mechanisms and convergence of symptoms that may occur during the natural course of dementia.

Disclosures

Rafael Blesa has received consulting fees from Novartis, Janssen-Cilag, Lundbeck and Nutricia, and lecture fees from Novartis and Nutricia. George T. Grossberg has served as a consultant for Acadia, Allergan, Avanir, Axovant, GE, Genentech, Lundbeck, Novartis, Otsuka, Roche, and Takeda Pharmaceuticals; he has received research support from Cognoptix, Janssen, and the National Institutes of Health, provided safety monitoring for EryDel, Merck, and Newron, and is on the speaker’s bureau for Acadia. Kazuhiro Toriyama and Kengo Ueda are full time employees of Novartis. Sean Knox was full time employee of Novartis at the time of development and finalization of the manuscript.

Consent for Publication

Not applicable.

conflict of interest

The authors declare no conflict of interest, financial or otherwise.