Acetylcholine (ACh) was the first neurotransmitter discovered. It was described in 1915 by Henry Hallett Dale for its actions on heart tissue and later confirmed as a neurotransmitter by Otto Loewi, who initially gave it the name “vagus stuff” because of its ability to mimic the electrical stimulation of the vagus nerve. Both scientists received the 1936 Nobel Prize in Physiology or Medicine for their work.

Ach is synthesized in certain neurons by the enzyme choline acetyltransferase (CAT). This enzyme is produced in the soma of cholinergic neurons and transported through the axon to nerve terminal where it is synthesized ACh from the compounds choline and acetyl-CoA. Coenzyme A is synthesized in mitochondria and accesses CAT following transport across the mitochondrial membrane into the cytoplasm. In contrast, choline comes from the liver and dietary sources and is captured from plasma into nerve terminal through a membrane transporter.

However, much of the choline used for ACh synthesis comes from the recycling of choline from metabolized Ach and the breakdown of the phospholipid, phosphatidylcholine.

The rate of ACh synthesis is regulated by precursor availability and by product inhibition, because ACh can bind at an allosteric site on choline acetyltransferase and inhibit the enzyme activity. Once it has been synthesized, a specific transporter uptakes the neurotransmitter from the cytoplasm into vesicles. These vesicles fuse with nerve terminal membrane when an action potential at the presynaptic neuron terminal causes and influx of Ca²⁺. This way, ACh diffuses into the synaptic cleft and can bind to postsynaptic receptors. Finally, the neurotransmitter is rapidly inactivated by cholinesterase enzymes, mainly by neuronal acetylcholinesterase but also by glial butyrylcholinesterase.

There are two classes of receptors that bind ACh: nicotinic and muscarinic. Nicotinic receptors bind nicotine and are located at the neuromuscular junction, autonomic ganglia, and sparsely in the central nervous system (CNS). They are ionotropic receptors linked directly to ionic channels and consist of five polypeptide subunits. Their activation causes the opening of the channel, which increases the Na + movement into the cell and leads to depolarization and generation of the action potential.

Muscarinic receptors bind muscarine and are located at parasympathetic autonomic innervated visceral organs, on the sweat glands and piloerector muscles, and both at postsynaptic and presynaptic level in the CNS. They are G protein-coupled receptors composed of a single polypeptide. Their activation in postsynaptic cells can be either excitatory or inhibitory and is always slow in onset and long in duration.

ACh has excitatory actions at the neuromuscular junction, autonomic ganglion, and glandular tissues and in the CNS. It has inhibitory actions at certain smooth muscles and at cardiac muscle.

Central cholinergic neurons can be subdivided into interneurons and projection neurons. The interneurons are present in the caudate–putamen nuclei, in the hypothalamus, and in the spinal cord. The projection neurons have two important clusters in the brain: the forebrain cholinergic complex (Chaps. 1, 2, 3, and 4) and pontomesencephalotegmental complex (Chaps. 5 and 6).

The forebrain cholinergic complex is formed by medial septum, horizontal and vertical diagonal band of Broca, and nucleus basalis of Meynert. Neurons of the medial septum innervate predominantly the hippocampus; those of the vertical and horizontal diagonal band project to the anterior cingulate cortex and olfactory bulb, respectively; and those of the nucleus basalis of Meynert provide afferents to the amygdala and throughout the rest of the cortical mantle (Bigl et al. 1990). Therefore, the cholinergic neurons of the forebrain complex are important in memory and cognition. The pontomesencephalotegmental complex is formed by the pedunculopontine and laterodorsal tegmental nuclei. Their neurons innervate the pontine reticular formation, the thalamus, the limbic system, the superior colliculus, and the basal ganglia. Consequently, the cholinergic neurons of the pontomesencephalotegmental complex are involved in the rapid eye movement sleep and eye movements, sleep–wake cycle and arousal, stimulus-reward learning, visual orienting, and sensory–motor patterns (Schliebs and Arendt 2006).

Physiologically, the brain cholinergic system is involved in many functions in the central nervous system. It plays a role in controlling cerebral blood flow, cortical activity, sleep–wake cycle, conscious awareness, behavior, and modulating cognitive function and cortical plasticity (Schliebs and Arendt 2006).

With respect to sleep–wake cycle or circadian rhythm, cholinergic relevance is explained because high cholinergic background activity occurs during wakefulness and rapid eye movement (REM) sleep. This is important also in cognition because REM sleep appears strongly related with episodic memory and circadian rhythms exert important influences on cognitive processes and different sleep stages may support in particular the development of memory consolidation (Van der Zee et al. 2009; Brankačk et al. 2009). Consequently, pharmacological blockade of cholinergic receptors interferes with REM sleep, and restoration of cholinergic activity may yield normalization of sleep–wake patterns. What remains to be confirmed, however, is whether such improved sleep patterns may directly counteract memory deterioration.

Considering conscious awareness, the role of ACH it is easy to understand because 90 % of brainstem projections to the thalamus, one of the most important structures involved, are cholinergic (Bentivoglio 1990).

With regard to behavior, it should be taken into account that the limbic system is a major target for cholinergic innervations (Mesulam 1995). Therefore, cholinergic pathways are related with vegetative and survival behaviors, emotions, learning, and memory (Mega et al. 1997).

Regarding cognitive function, different studies in humans indicate that cholinergic pathways have important functional roles in attention, working memory, and a number of additional mnemonic processes (Perry et al. 1999).

With reference to cognitive function, it has been described the involvement of cholinergic system in learning, memory, and attention (Schliebs and Arendt 2006).

However, the role of ACh in learning and memory is complex and still not fully understood. ACh affects not only one, but possibly all memory systems in different ways, and that it modulates the distinct phases of learning and memory differentially: favoring memory encoding and attention efforts while hampering memory consolidation and retrieval (Van der Zee et al. 2011).

In relation to attention, psychopharmacological, neuroimaging, and psychological studies of cholinergic system functioning in humans show that the cholinergic system has a specific modulatory role in this cognitive function (Perry et al. 1999; Sarter et al. 2006). Deficiencies in attention processing impair discriminatory processes and responsiveness to relevant and new stimuli and, as a result, can cause cognitive deficits. Moreover, some authors hypothesize that central cholinergic impairment delineates a specific central cholinergic deficiency syndrome of behavioral and psychological symptoms in dementia of the Alzheimer’s type characterized by psychosis, restlessness, agitation, and mood symptoms (Lemstra et al. 2003).

Finally, the function of acetylcholine in cortical plasticity is possible because it establishes synaptic contacts in networks of cells that will perform complex cognitive functions in adulthood (Berger-Sweeney 2003). Therefore, cholinergic system has been implicated in mediating plasticity in the brain in response to experience or injury. In fact, various animal studies have demonstrated the beneficial effects of cholinergic agonists on enhancing recovery and minimizing neuronal damage in various injury models, the impaired experience-dependent plasticity in the cortex and hippocampus in cholinergic depletion states, the modulation of neurotrophic factors that play a major role in neuronal survival and plasticity in adulthood by Ach, and the interaction between acetylcholine and estrogen in supporting hippocampal plasticity in aging females (Craig et al. 2011).

Although neuronal cell loss was found predominantly in pathological aging, such as AD, normal aging is accompanied by dendritic, synaptic, and axonal degeneration with nearly no cell loss (Burke and Barnes 2006; Coleman 2005; Rapp and Gallagher 1996; Rasmussen et al. 1996; Ypsilanti et al. 2008). These findings suggest that functional decline associated with aging across species does not primarily result from cell loss but other mechanisms including decrements in gene expression, impairments in intracellular signaling, and cytoskeletal transport that may mediate cholinergic cell atrophy leading to age-related functional decline in the brain (De Lacalle et al. 1996; Niewiadomska et al. 2006; Small et al. 2004; Williams et al. 2007).

ACh cells in basal forebrain are the most affected cholinergic cells in Alzheimer’s disease. Different studies have shown that these neurons are more susceptible to toxic agents as compared to those in the striatum and brain stem, indicating that brain cholinergic neurons demonstrate differential sensitivity to pathogenic insults (Fass et al. 2000; Julka et al. 1995). Some explanations of the particular vulnerability could be their dependency of acetyl-CoA not only for energy production but also for acetylcholine synthesis, their higher demand for energy production which cause them to be more sensitive to aging-related energy (glucose) deprivation (Szutowicz et al. 2006), AChE-induced expression with acute stress (Li et al. 1996), more susceptibility of transcription factors which are activated by cholinergic stimulation to oxidative stress, the relationship between glucose metabolism and cholinergic transmission (Schliebs 2005), and cholinergic cell susceptibility to inflammatory conditions (Wenk et al. 2000).

The loss of cholinergic activity in AD was demonstrated at various levels. Specific cholinergic deficit, involving the nucleus basalis of Meynert projections, hippocampus, and the temporal and the frontal non-motor areas, was consistently found in autopsy material from Alzheimer’s patients (Dournaud et al. 1995; Geula and Mesulam 1996). However, cholinergic innervations of the striatum (originating from striatal interneurons) and of the thalamus (originating in the brainstem) remained relatively intact in the disease (Geula and Mesulam 1999). Moreover, the activity of the enzyme responsible for the synthesis of acetylcholine, choline acetyltransferase (ChAT), was found to be remarkably decreased in pathological samples from the cortex and hippocampus of Alzheimer’s patients (Bowen et al. 1976; Davies and Maloney 1976; Perry et al. 1977). Also, other specific markers of the function of cholinergic synapses, the acetylcholine vesicular transport essential to replenish synaptic vesicles (VAChT), depolarization-induced acetylcholine release, and choline uptake in nerve terminals to replenish the acetylcholine synthetic machine, were reduced in the same tissues (Efange et al. 1997; Nilsson et al. 1986; Rylett et al. 1983). Although muscarinic receptor subtypes were not significantly changed in Alzheimer’s disease brains (Nordberg et al. 1992; Waller et al. 1986), at least one type of nicotinic receptors, the α-4 subtype, was described as consistently reduced in patient’s brains (Burghaus et al. 2000; Schröder et al. 1991). Moreover, trkA, the high affinity receptor of the nerve growth factor (NGF), a survival factor for cholinergic neurons of the nucleus basalis of Meynert, was found to be decreased in these neurons in the brains from Alzheimer’s disease patients (Salehi et al. 1996). Furthermore, in aged rats, forebrain cholinergic neurons demonstrated striking reductions in the retrograde transport of NGF, and cholinergic cells that were no longer capable to transport NGF appeared severely shrunken (De Lacalle et al. 1996; Cooper et al. 1994). Correspondingly, neuropathological studies have suggested that the structural components of the limbic network are the primary site of neurofibrillary tangle formation in patients with AD (Braak and Braak 1991; Brun and Gustafson 1976). Thus, according to Braak and Braak theory of progression of AD pathology, neurofibrillary tangles first appear in the transentorhinal cortex and then progress to other structures of the limbic system before finally spreading to neocortical structures in the end-stages of AD (Braak and Braak 1991).

The association of cholinergic hypofunction with cognition impairment was suggested by different studies showing a correlation of clinical dementia ratings with the reductions in a number of cortical cholinergic markers such as ChAT, muscarinic and nicotinic acetylcholine receptor binding as well as levels of Ach (Bierer et al. 1995; Gsell et al. 2004; Nordberg 1992). Furthermore, it was described that damage of the basal forebrain cholinergic system in autoptic brains was related to the dementia score evaluated from the patient during life (Perry et al. 1996; Wilcock et al. 1982).

Moreover, pharmacological and lesional studies corroborated this role of the neurotransmitter. After first postulation of Deutsch 1971 (Deutsch 1971), other studies with antimuscarinic agents, selective muscarinic antagonists, and centrally acting nicotinic-cholinergic antagonists have shown to impair memory performance in a variety of behavioral paradigms in rodents (Decker and McGaugh 1991; Hunter and Roberts 1988; Levin 1992). Both muscarinic antagonists and nicotinic antagonists have also shown to impair memory performance in monkeys and humans (Terry et al. 1993; Vitiello et al. 1997; Elrod and Buccafusco 1991; Newhouse et al. 1994; Hristensen et al. 1992; Molchan et al. 1992). Similarly, lesion-induced damage to basal forebrain cholinergic system and cholinergic projections to the neocortex in animal models induce cognitive impairments, especially on attention, as well as on learning and memory processes (Sarter and Bruno 1997; McKinney 2005). It should be noted that damage to similar basal forebrain regions in humans (as a result of arterial aneurysms or resection of an arteriovenous malformation) has also been associated with severe memory deficits (Damasio et al. 1985). On the contrary, drugs enhancing central cholinergic function improve the performance of aged patients (Drachman and Leavitt 1974; Drachman 1977) and reverse deleterious effects of anticholinergic drugs (Bartus et al. 1982).

Thanks to cholinergic hypothesis, the primary therapeutic approach to address the cognitive loss associated with AD was the cholinergic replacement strategy. Although studies with muscarinic and nicotinic–cholinergic ligands were unsuccessful, acetylcholinesterase inhibitors (AChEIs) demonstrated a slight reverse of memory impairment in AD patients and finally became the first and almost unique disease specific treatments.

However, later this cholinergic hypothesis has been questioned, and it is no longer believed that the cholinergic depletion alone is responsible for causing AD.

The reasons for this questioning are diverse. Firstly, studies of AD patients indicate that the loss of cholinergic markers cannot be detected in individuals with mild AD and that the cholinergic deficit is not present until relatively late in the course of the disease (Davis et al. 1999). In contrast to patients with advanced AD, in autoptic brain samples of patients with mild cognitive impairment (MCI) and early AD, no decrease in ChAT activity has been observed in a number of brain regions studied (Tiraboschi et al. 2000). Similarly, the number of ChAT-positive and VAChT-positive cells was unaltered in MCI as compared to non-demented controls (Gilmor et al. 1999). In contrast, in hippocampus and frontal cortex of MCI patients, even an increased activity of ChAT has been observed, indicating that the cognitive deficits observed are seemingly not interrelated with ChAT activity (DeKosky et al. 2002). Also in vivo PET studies of MCI and early forms of AD have observed only mild loss of AChE, as revealed by ligands that label AChE (Rinne et al. 2003). Secondly, the benefit of pro-cholinergic therapy on cognitive function in AD and age-related cognitive deficits is modest, and it is not able to stop disease process (Lleó 2007; Drachman et al. 1982). Thirdly, models of muscarinic cholinergic blockade in normal older adults did not replicate all of the cognitive deficits in AD. Specifically, research with muscarinic and nicotinic cholinergic manipulations in healthy subjects and in animals showed that the cholinergic system primarily contributes to effortful attention processes more than memory, the primary deficit in AD (Sarter et al. 2006; Newhouse et al. 2001). Fourthly, age-related changes in sex hormones such as estradiol can affect cholinergic integrity and cognitive processes (Gibbs 2010). Finally, to further complicate the issue, novel data in primates and humans confirmed that cortical cholinergic activity was also decreased by aging (Smith et al. 1999) and that cholinergic deficits were also found in neurodegenerative diseases others than Alzheimer’s disease, although not in hippocampus (Perry et al. 1985; Murdoch et al. 1998).

Nevertheless, some of the conclusions of aforementioned studies appear premature, and there are important facts that can explain these results and that should be taken into account.

The first one is that that upregulation of ChAT in the surviving cholinergic synapses can compensate for early deficits of the neurotransmitter (Craig et al. 2011). The upregulation of hippocampal ChAT in MCI cases may be due to the replacement of denervated glutamatergic synapses by cholinergic input arising from the septum (Mufson et al. 2008). This way, nearly normal cholinergic levels can be detected in the cortex in spite of significant loss of cholinergic neurons. In fact, one study revealed that cognitive deficits are detectable not earlier before at least 30 % of the total cholinergic basal forebrain cells have degenerated (Arendt 1999). The second one is that ChAT or AChE is not rate-limiting cholinergic enzymes. Therefore, they do not reflect exactly the cholinergic function in the living patients (Terry and Buccafusco 2003). The third one is that although the number of cholinergic neurons can be preserved, the dysfunction of these can be detected in early stages. Thus, parameters of cholinergic function such as trkA and p75NTR receptors, acetylcholine release, high-affinity choline uptake, and expression of mAChRs and nAChRs are altered in MCI and early AD (Mufson et al. 2007; Auld et al. 2002; Picciotto and Zoli 2002). Moreover, neurotrophic factors like NGF and BDNF are dysregulated in MCI and AD indicating an enhanced vulnerability of the cholinergic system in AD (Cuello et al. 2007). Accordingly, neurofibrillary degeneration and cell volume loss have been detected in early stages of AD (Sassin et al. 2000; Mesulam 2004). The fourth one is that taking into account that there is an age-related cholinergic denervation, age-matched control subjects used in studies already show cholinergic depletions of the hippocampus, and it can be difficult to demonstrate the loss of cholinergic neurons in AD (Kuhl et al. 1996). Finally, it should be kept in mind that neurochemical analysis in human tissue samples are compromised by unavoidable delays in their postmortem collection that do not exist in animal studies. That is why in vivo imaging methods are important. In fact, they have shown to support the cholinergic hypothesis. Specifically, positron emission tomography (PET) studies indicate that cortical acetylcholinesterase activity is reduced in AD patients, nicotinic receptor deficits are present in early stages of AD and correlate with the level of cognitive impairment, and muscarinic receptors decrease with age and AD in neocortical regions (Kuhl et al. 1999; Nordberg 2001; Zubieta et al. 2001). Moreover, single photon emission computerized tomography (SPECT) studies indicate that the vesicular acetylcholine transporter is reduced throughout the entire cerebral cortex and hippocampus in early onset AD patients.

Given all this information, it seems that although dysfunction of cholinergic neurons is relevant to explain symptoms seen in AD, it cannot account for all the manifestations of the illness (Pinto et al. 2011). This fact was the origin of the reformulation of the hypothesis by Craig et al. (2011). Its framework is the cofactor theory of McDonald in 2002 that predicts that different risk factors associated with AD have converging effects on hippocampus causing neuronal damage and death accompanied by progressive cognitive decline (McDonald 2002). It is also based on the Ach role in plasticity of brain through mechanisms like neurogenesis, neurotrophic factors, and changes in dendritic branching, which are involved in learning and memory as well as in functional recovery from injury. Considering the demonstrated cholinergic loss in early stages of the disease, this hypothesis proposes that this cholinergic depletion reduces the ability of the brain to compensate for the accumulation of risk factors, whose frequency increase with age. Therefore, in a healthy individual, sub-threshold injury can be unnoticed because of Ach-mediated compensatory mechanisms. However, memory impairment can appear after a major insult (stroke) or when a minor insult (mild ischemia, elevated glucocorticoids, epileptiform activity) occurs in an individual with poor cholinergic projections to the hippocampus (Craig et al. 2011).

Finally, it should be said that the cholinergic hypothesis has not only been described in AD but also in several brain diseases like psychiatric disorders and brain traumatic injury and also in sleep regulation (Arciniegas 2003; Battaglia 2002; Dilsaver and Coffman 1989; Hshieh et al. 2008; Luppi et al. 2006; Raedler et al. 2007). However, no specific treatments following this theory have been approved in these disorders.

There are some evidences of a link between amyloid precursor protein (APP), the originator of neuritic plaques characteristic of AD, and cholinergic transmission. Before describing them, it is appropriate to clarify that there are two alternatives ways in APP processing. One is amyloidogenic and generates β-amyloid peptide by the sequential action of β- and γ-secretases. The other one is non-amyloidogenic and generates soluble APPα (sAPPα) by the action of α-secretase.

First evidence of the aforementioned link emerged when it was observed that acetylcholinesterase (AChE) colocalized with β-amyloid deposits in Alzheimer’s brains (Morán et al. 1993). Another evidence came from studies showing that M1/M3 muscarinic cholinergic agonists increased sAPPα secretion and decreased total β-amyloid formation both in and in vivo in patients with Alzheimer’s disease (Müller et al. 1997; Hock et al. 2003). Regarding the relationship between nicotinic–cholinergic agonists and Aβ deposition, it is complex and incompletely understood (Oz et al. 2013). Agreeing with these mentioned reports, inhibitors of AChE were found to increase secretion of sAPPα in both cortical rat brain slices and cell culture (Mori et al. 1995; Racchi et al. 2001), and scopolamine treatment of transgenic Tg2576 mice resulted in increased levels of fibrillar β-amyloid and decreased α-secretase activity (Liskowsky and Schliebs 2006). Neurotrophic growing factor (NGF) signaling has also been shown to influence expression and metabolism of APP and to modulate the cholinergic control of APP processing (Isacson et al. 2002; Haring et al. 1995).

AChE intervenes not only in APP processing but also in β-amyloid aggregation itself. It seems that the enzyme forms a complex with the protein and increases the neurotoxicity of Alzheimer’s fibrils (Alvarez et al. 1998; Reyes et al. 2004). Conversely, β-amyloid increases AChE in vitro through α7-nicotinic ACh receptors, with β-amyloid (1–42) being more potent than β-amyloid (1–40) (Fodero et al. 2004). Butyrylcholinesterase (BuChE) seems also involved in β-amyloid aggregation because its levels correlate positively with amyloid plaques and neurofibrillar tangles in Alzheimer’s brains, and there are studies suggesting a role of BuChE in the transformation of β-amyloid into neuritic plaques (Mesulam and Geula 1994; Guillozet et al. 1997).

Several studies have demonstrated that activation of nicotinic Ach receptors, presumably mediated through activation of the α7 subtype, results in a significant increase in tau phosphorylation (Wang et al. 2003). In contrast, muscarinic Ach receptors activation may prevent tau phosphorylation (Wang et al. 2003; Rubio et al. 2006). According with this, chronic nicotine administration to 1-month-old triple transgenic 3xTg-AD mice for 5 months did not change soluble β-amyloid levels but resulted in a striking increase in phosphorylation and aggregation of tau, which appeared to be mediated by p38-MAP kinase (Oddo et al. 2005).

There is abundant evidence that β-amyloid may trigger cholinergic dysfunction through action on α7 nicotinic receptors, by affecting NGF signaling, mediating tau phosphorylation, interacting with acetylcholinesterase, and specifically affecting the proteome in cholinergic neurons (Schliebs and Arendt 2011). In fact, it has been observed that the severity of neurodegeneration in AD correlates best with the pool of soluble β-amyloid than with the number of insoluble β-amyloid plaques (McLean et al. 1999). Thus, in different cell and animal models, prefibrillar assemblies of β-amyloid have been shown to induce neurotoxicity, electrophysiological changes, and disruption of cognitive function, which may explain why early cholinergic dysfunction occurs before there are substantial plaques in AD (Cleary et al. 2005). In particular, there are studies providing evidence that soluble β-amyloid can inhibit release of ACh from hippocampal slices, decrease the intracellular acetylcholine concentration, decrease activity of choline acetyltransferase, impair M1 muscarinic Ach receptors, desensitize α7 nicotinic receptors at high concentration, and inhibit hippocampal long-term potentiation in brain slices and rat brains in vivo (Kar et al. 2004; Hoshi et al. 1997; Pedersen et al. 1996; Kelly et al. 1996; Dineley et al. 2002; Walsh et al. 2002; Wang et al. 2002).

Furthermore, the NGF receptor p75NTR has been shown to increase the susceptibility of cells to β-amyloid toxicity. Considering that the p75NTR is mainly expressed by basal forebrain cholinergic cells, this could explain the particular vulnerability of these cells to β-amyloid in AD (Perini et al. 2002). Moreover, semiquantitative immunohistochemical study in aged Tg2576 mice revealed a β-amyloid-mediated decrease in cholinergic innervation of cortical blood vessels, which may contribute to the alterations of the cerebrovascular system observed in transgenic Tg2576 mice (Bürger et al. 2009).

Oral rivastigmine is rapidly and completely absorbed. Peak plasma concentrations are reached in approximately 1 h. As a consequence of rivastigmine’s interaction with its target enzyme, the increase in bioavailability is about 1.5-fold greater than that expected from the increase in dose. Absolute bioavailability after a 3 mg dose is about 36 % ± 13 %. Administration of rivastigmine with food delays absorption (t _max) by 90 min and lowers C _max and increases AUC by approximately 30 % (http://www.ema.europa.eu/docs/en_GB/document_library/EPAR_-_Product_Information/human/000169/WC500032598.pdf).

Absorption of rivastigmine from the transdermal patch is slow, with rivastigmine being detected in the plasma after a lag time of 0.5–1 h after the first dose. Approximately 50 % of the drug load is released from a patch during the 24-h application period. Peak plasma concentrations (C _max) are reached in 10–16 h after a single dose, with a slow decrease in concentration over the remainder of the 24-h period. On application of a new patch during multiple dose administration, there is an initial gradual decrease in plasma rivastigmine concentrations until the rate of absorption from the new patch exceeds that of elimination, after which time plasma concentrations increase gradually to reach a peak at 8 h. There is no relevant accumulation of rivastigmine or its metabolite NAP226-90 following multiple dose administration of the patch, with the exception of higher plasma rivastigmine concentrations on the second day than on the first day of patch administration. It should be noted that exposure to rivastigmine and its metabolite NAP226-90 is highest when the patch is applied to the upper back, chest, or upper arm, with exposure levels 20–30 % lower when the patch is applied to the abdomen or thigh (http://www.ema.europa.eu/docs/en_GB/document_library/EPAR_-_Product_Information/human/000169/WC500032598.pdf; http://www.pharma.us.novartis.com/product/pi/pdf/exelonpatch.pdf).

The main differences between transdermal and oral rivastigmine formulations pharmacokinetic properties are that patches have a more gradual and sustained absorption. Patches have less fluctuation between the maximum and minimum plasma concentrations. Thus, steady-state trough concentrations of rivastigmine are 50 % of peak levels after patch administration, while they are almost zero between the two daily doses with oral administration. Moreover, increments in exposure when rising the dose are less pronounced with the patch. So, C _max values are lower, and time to C _max values are longer with patches than with capsules. Lastly, single-dose intersubject variability is smaller with transdermal administration than with the oral form (http://www.ema.europa.eu/docs/en_GB/document_library/EPAR_-_Product_Information/human/000169/WC500032598.pdf).

There is an approximately linear and inverse relationship between exposure to rivastigmine and its metabolite NAP226-90 and bodyweight of AD patients at steady state. Hence, rivastigmine steady-state concentrations are doubled when bodyweight decreases by half, and they are halved when bodyweight doubles (http://www.ema.europa.eu/docs/en_GB/document_library/EPAR_-_Product_Information/human/000169/WC500032598.pdf).

Rivastigmine is weakly bound to plasma proteins (40 %), with an apparent volume of distribution of 1.8–2.7 L/kg. It crosses the blood–brain barrier readily, with peak CSF concentrations observed 1.4–2.6 h post-dose (http://www.ema.europa.eu/docs/en_GB/document_library/EPAR_-_Product_Information/human/000169/WC500032598.pdf).

Rivastigmine is rapidly and extensively metabolized, primarily via cholinesterase-mediated hydrolysis to the metabolite NAP226-90. In vitro, this metabolite shows minimal inhibition of acetylcholinesterase (<10 %). Half-life in plasma is approximately 1 h for oral rivastigmine and 3.4 h after transdermal patch. The longer t _1/2 of the patch is explained because elimination is rate limited by absorption (http://www.ema.europa.eu/docs/en_GB/document_library/EPAR_-_Product_Information/human/000169/WC500032598.pdf).

Based on in vitro studies, no pharmacokinetic interaction is expected with medicinal products metabolized by the following cytochrome isoenzymes: CYP1A2, CYP2D6, CYP3A4/5, CYP2E1, CYP2C9, CYP2C8, CYP2C19, or CYP2B6. Based on evidence from animal studies, the major cytochrome P450 isoenzymes are minimally involved in rivastigmine metabolism. Total plasma clearance of rivastigmine is approximately 130 l/h after a 0.2 mg intravenous dose and decreases to 70 l/h after a 2.7 mg intravenous dose, which is consistent with the nonlinear, overproportional pharmacokinetics of rivastigmine due to saturation of its elimination. Renal clearance of rivastigmine is 2.1–2.8 L/h. Rivastigmine is metabolized to a lesser extent after transdermal than after oral administration, presumably because of the lack of presystemic (hepatic first pass) metabolism. So, the metabolite-to-parent AUC ratio is around 0.7 after transdermal patch administration versus 3.5 after oral administration. No unique metabolic routes have been detected in human skin in vitro. Rivastigmine is eliminated by the kidneys mostly as metabolites. Unchanged rivastigmine is found in trace amounts in the urine and feces. Nicotine use increases the oral clearance of rivastigmine by 23 % in patients with AD (http://www.ema.europa.eu/docs/en_GB/document_library/EPAR_-_Product_Information/human/000169/WC500032598.pdf).

Age has no impact in bioavailability of rivastigmine in AD. Studies in Alzheimer’s patients aged between 50 and 92 years showed no change in bioavailability with age (http://www.ema.europa.eu/docs/en_GB/document_library/EPAR_-_Product_Information/human/000169/WC500032598.pdf).

The C _max of rivastigmine was approximately 60 % higher, and the AUC of rivastigmine was more than twice as high in subjects with mild to moderate hepatic impairment than in healthy subjects. Renal impairment C _max and AUC of rivastigmine were more than twice as high in subjects with moderate renal impairment compared with healthy subjects; however, there were no changes in C _max and AUC of rivastigmine in subjects with severe renal impairment. No study has been conducted with rivastigmine transdermal patches in subjects with hepatic or renal impairment (http://www.ema.europa.eu/docs/en_GB/document_library/EPAR_-_Product_Information/human/000169/WC500032598.pdf).

No relevant differences in the pharmacokinetics of rivastigmine transdermal patch were observed between Japanese and White healthy individuals (Lefèvre et al. 2009).

Cognitive function impairment is the one of the chore features of AD and PDD and has a great impact in patients and caregivers daily life. Therefore, it is one of the main outcomes of the studies analyzing the efficacy of rivastigmine.

There are different scales available to measure this complex brain function. However, the most frequently used and that were analyzed in the aforementioned studies are ADAS-cog and MMSE. ADAS-cog stands for Alzheimer’s Disease Assessment Scale, cognition subscale, and its primary purpose is to be an index of global cognition in response to anti-dementia therapies. It is a clinician-administered 70-point evaluation and assesses multiple cognitive domains including memory, language, praxis, and orientation (Rosen et al. 1984). MMSE stands for Mini-Mental State Examination and is a 30-point scale clinician-administered evaluation and is the mostly used test worldwide for screening and dementia staging (Hashimoto and Mori 2011). Also, other scales that evaluate specific cognitive domains are used as secondary measures in those studies. Trail Making Test part A (TMTA) (Bornstein 1985), a test where consecutive targets numbers on a sheet of paper should be connected in sequential order by the test taker and which evaluates attention cognitive domain, and the Ten-Point Clock-Drawing Test (TCD), which assesses visuospatial and executive functions, are the most used ones (Mendez et al. 1992).

Efficacy of rivastigmine capsules in the improvement of ADAS-cog is shown in the Cochrane review (Birks and Grimley Evans 2015). High-dose rivastigmine (6–12 mg daily) was associated with a two-point improvement in cognitive function on the ADAS-cog score compared with placebo after 26 weeks. Lower-doses capsules (4 mg daily or lower) showed same statistically significant results.

Efficacy of rivastigmine patches was shown in the IDEAL study (Winblad et al. 2007). Treatment with rivastigmine 9.5 mg/24 h patch significantly improved cognitive and global function. The patch was noninferior to rivastigmine 12 mg/day capsules in terms of cognitive improvement. Cognition and global function were assessed through the ADAS-cog at 24 weeks after treatment with rivastigmine patch 9.5 mg/24 h (10 cm²) and capsules of 12 mg/d compared to placebo. At that time, the two routes of the administration of the drug achieved an increase of up to 4 points in the ADAS-cog. In particular, a retrospective analysis of the study indicated improvements of delayed recall of words, constructive praxis, and ideational and recall of test interactions in capsules vs. placebo and showed better results in the areas of delayed recall of words, naming objects, and fingers and ideational praxis in patches vs placebo (Winblad et al. 2007; Grossberg et al. 2010b). There were also significant improvements in secondary efficacy measures of this randomized trial MMSE and TMTA (Winblad et al. 2007; Grossberg et al. 2010b). Both the capsules and the patch showed improvements of these two subscales with no statistically significant difference between them. No improvements were seen in TCD for both routes of administration (Winblad et al. 2007; Grossberg et al. 2010b). The 24-week results for the assessment tools and the results for clinically relevant responders of this study are summarized in Tables 4.1 and 4.2, respectively.