1) Which anti-Alzheimer drug - donepezil, rivastigmine, or galanthamine - is the most appropriate c

Fråga: 1) Which anti-Alzheimer drug - donepezil, rivastigmine, or galanthamine - is the most appropriate choice in Alzheimer patients with moderate to severe impairment of renal function? 2) What is known about the metabolism of rivastigmine, donepezil, and galanthamine, and the pharmacological activity of drug metabolites? 3) The question relates to an 80-year-old male patient on hemodialysis and a discussion about the effects of hemodialysis on drug dosage is also requested.

Sammanfattning: None of the anti-alzheimer drugs donepezil, rivastigmine or galanthamine, have been properly studied in patients with impaired renal function. Pharmacokinetic data speak in favour of using rivastigmine, especially in cases of hemodialysis. Importantly however, the only published clinical study relates to donepezil, suggesting no change of donepezil kinetics in cases of impaired kidney function. The tolerability for cholinergic side effects might be useful in dose titration of all three drugs.

Svar: Donepezil, rivastigmine and galanthamine, are all acetylcholine esterase inhibitors, with a similar modest effect on progression of symptoms in Alzheimer´s disease (1-3). There is very little information in the literature concerning the treatment of patients with reduced renal function. However, the available data indicate that some potentially important differences between the drugs with regard to pharmacokinetics and drug metabolism, which might be of clinical relevance in patients with renal dysfunction.

Donepezil has an almost complete oral bioavailability, a long half-life (70 h), and exhibits a high protein binding (93-96 per cent), predominantly to albumin but also orosomucoid. The volume of distribution has been estimated to approximately 14 L/kg in a small material of young healthy volunteers (mean 23 years), compared to approximately 22 L/kg in a small group of elderly patients (mean 74 years) (4). Donepezil is subject to extensive metabolism in the liver by cytochromes P4502D6 and 3A4, as well as glucuronosyl transferase. The CYP2D6-metabolite O-desmethyl-donepezil has been described as equally potent in acetylcholine esterase inhibition compared to donepezil, but only reaching one fifth of donepezil plasma levels (1, 5). From a dose of donepezil, 80 per cent can be recovered in urine mainly as donepezil-metabolites but also as unchanged donepezil (10-17 per cent of urinary content). Unfortunately, we have not found any data on the specific urinary contents of the active O-desmethyl-donepezil-metabolite. In addition to renal excretion, around 20 per cent of the donepezil dose is subject to faecal elimination (1, 4, 6).

The pharmacokinetics of a single dose of donepezil in patients with moderately to severely impaired renal function (n=11, creatinine clearance less than 30 ml/min/1.73 m2, but not in dialysis) was compared to the same number of age-, and weight-matched controls (6). Importantly, no significant differences were found in kinetic parameters of donepezil and the drug was equally well tolerated in the two groups. However, the levels of active metabolites were not determined (6).

In contrast to donepezil, rivastigmine has a low protein binding (40 per cent) and a short half-life (0.5-3h) due to a rapid hydrolysis by hepatic esterases (1, 7). Furthermore, rivastigmine has a lower oral bioavailability than donepezil (35 per cent), and a smaller volume of distribution (around 2 L/kg), but this is somewhat greater for the hydrolysis product (around 5 L/kg). Importantly, this decarbamylated metabolite exhibits less than 10 per cent of the activity of rivastigmine. The metabolite is conjugated with sulfate and efficiently eliminated in urine (more than 90 per cent of the dose eliminated within 24 hours), with less than one per cent detected in faeces (1, 7, 8). Rivastigmine has been described as a pseudo-irreversible inhibitor of acetylcholine esterase, forming a complex with the target enzyme (7). Theoretically, with regards to the pharmacological activity, the formation of such a complex should to some extent counterbalance the otherwise rapid elimination of free drug.

There are no published studies on how renal function impairment affects concentrations of rivastigmine or its metabolites, and the data-on-file provided by the manufacturer based on single-dose experiments are not conclusive. For example, they report that Cmax and AUC of rivastigmine were doubled in patients with moderate impairment of kidney function, but not different from healthy controls in patients with severely reduced kidney function (3). Less surprisingly, levels of decarbamylated rivastigmine-metabolite are higher in cases of moderate kidney function impairment (7), with an increased elimination half-life from 3.4 to 6h (8). This has not been given any clinical significance since doses of rivastigmine are titrated to maximal tolerability for cholinergic adverse effects (7). Interestingly, symptoms such as nausea, vomiting and dizziness have been reported as more frequent with rivastigmine than with donepezil (8).

Galanthamine is the most recently approved drug for the use against Alzheimer´s disease in Sweden (2), and documentation on this indication is very limited. It has almost 100 per cent oral bioavailability, negligible protein binding and a volume of distribution similar to rivastigmine (9). In a group of Alzheimer patients, the plasma half-life was estimated to around 8h, some 30 per cent higher compared to younger healthy volunteers. Galanthamine undergoes desmethylation and subsequent glucuronidation. CYP2D6 is responsible for the generation of a pharmacologically active metabolite, O-desmethyl-galanthamine, but this metabolite is very unstable and rapidly inactivated. Overall, galanthamine metabolites are not considered to contribute to clinical effects (1, 9). Around 50 per cent of the dose is recovered in urine after 72h, of which half is excreted as unchanged drug (1). We have not found any specific data about the possible effects of impaired kidney function on pharmacokinetics of galanthamine.

Some of the above information is essential when predicting the effect of hemodialysis on drug plasma concentrations. Donepezil, rivastigmine, and galanthamine, share a similar low molecular weight (Mr 379, 250, 285, respectively), meaning that diffusion over the dialysis membrane of the unbound drug molecules should be similar for all three drugs. However, the low extent of rivastigmine and galanthamine protein binding, together with the rather small volume of distribution, indicate that hemodialysis would lead to a substantial reduction of rivastigmine or galanthamine plasma levels, in contrast to donepezil. In conclusion, accumulation of galanthamine and rivastigmine can be avoided by hemodialysis.

In summary, all three drugs are extensively metabolised. This means an efficient inactivation of rivastigmine and galanthamine but not donepezil, where at least one active metabolite has been characterised. The kinetics and effects of this donepezil-metabolite in kidney disease are not known, which limits the value of a published study on single-dose kinetics on donepezil in kidney disease. Rivastigmine is subject to the fastest inactivation and elimination of the three and both rivastigmine and galanthamine should be eliminated efficiently by hemodialysis, in contrast to donepezil. There is less documentation on galanthamine than rivastigmine or donepezil. The dosage of all three drugs might be possible to titrate from the presence of cholinergic side effects.