Development of a pharmacological evidence‐based anticholinergic burden scale for medications commonly used in older adults

The present study aimed to develop a pharmacological evidence‐based anticholinergic burden scale (ABS) through a direct assessment of muscarinic receptor‐binding activities of 260 medications commonly used in older adults.


Introduction
An increased life expectancy has been associated with better access to health care and treatment for many diseases. This has resulted in more elderly patients with multidrug regimens prescribed for multiple morbidities. Polypharmacy with anticholinergic agents is common in nursing home residents and markedly increases the risk of anticholinergic toxicity. 1 More than 600 medicinal products with a broad therapeutic range and adverse effect profiles are recognized to exhibit anticholinergic activity, 2,3 and include psychoactive drugs, such as hypnotic, antipsychotic, antiparkinsonian, and antidepressant drugs, and non-psychoactive drugs, such as cardiac, corticosteroid, and antibiotic drugs. 4,5 Drugs taken by patients may also exaggerate the anticholinergic effects of typical anticholinergics used frequently to treat overactive bladder and chronic obstructive pulmonary disease. Therefore, anticholinergic accumulation results in dry mouth, constipation, blurred vision, dry eyes, tachycardia, urinary retention, cognitive impairment, agitation, paranoia, anxiety, and delirium. 6 Anticholinergic effects are caused by the significant blockade of muscarinic receptors in the parasympathetic and central nervous systems. 7,8 More receptors may be additively blocked when several agents are simultaneously administered, which is known as anticholinergic accumulation, and occurs because of polypharmacy and/or changed pharmacokinetics. 4,[9][10][11] This accumulation may enhance anticholinergic adverse effects.
A rating scale to quantify the anticholinergic burden may be useful for reducing the risk of anticholinergic adverse events in elderly patients. Anticholinergic drug rating scales have been developed through collaborations between a number of investigators. 10,[12][13][14][15][16][17][18][19][20][21][22][23] In these scales, drugs are rated in an original manner from 0 to 3, with 0 signifying no known anticholinergic activity and 3 indicating marked anticholinergic activity. The scores of all medications are then summed to obtain the total score. The majority of previous studies evaluated the rating of anticholinergic scores using practitioners' knowledge of a list of drugs with known anticholinergic effects or serum anticholinergic activity, [23][24][25][26][27] but not by the direct and quantitative assessment of the anticholinergic activity of individual drugs using pharmacological methods. DOI: 10.1111/ggi.14619 558 | A reliable pharmacological assessment for clinically relevant anticholinergic activity involves directly elucidating whether drugs exhibit binding activity for muscarinic receptors and the functional antagonism of the cholinergic agonist-induced response. 7,8 Furthermore, the findings obtained need to be confirmed by the presence of typical antimuscarinic adverse drug effects, such as dry mouth and constipation, in clinical studies. As an initial step, assessments of the presence or absence of muscarinic receptorbinding activity are crucial for confirming the anticholinergic adverse event of each drug. A radioreceptor-binding assay with a selective radioligand may be a simple and powerful pharmacological tool for evaluating the presence or absence of the anticholinergic activity for each drug. 8 The population of Japan is aging rapidly, and although a number of drugs are clinically used in older adults, some are not included in previously developed scales, such as the anticholinergic cognitive burden. 19,27 Therefore, the present study aimed to develop a pharmacological evidence-based anticholinergic burden scale (ABS) through direct assessments of the muscarinic receptor-binding activities of medications commonly used in elderly Japanese.

Animals
Eight-to 10-week-old male Sprague-Dawley rats (250-300 g) were purchased from Japan SLC (Shizuoka, Japan). They were housed in the laboratory with free access to food and water and maintained on a 12-h light/dark cycle in a room with a controlled temperature (24 AE 2 C).
Experimental protocols received approval from the Ethics Committee for Research at the University of Shizuoka (registration number: 136038), and were performed in accordance with the guidelines for the Care and Use of Laboratory Animals, which conformed to the provisions of the Declaration of Helsinki (as revised in Tokyo 2004).

Muscarinic receptor-binding assay
The muscarinic receptor-binding assay was performed using [ 3 H] NMS, a selective radioligand for muscarinic receptors, as previously described. 8 Animals were exsanguinated by taking blood from the descending aorta under anesthesia with pentobarbital sodium (50 mg/kg, i.p.), and the whole brain, except for the cerebellum, was dissected. Brain tissue was carefully minced and homogenized using a Kinematica Polytron homogenizer in 19 volumes of ice-cold 30 mM Na + /HEPES buffer (pH 7.5). The homogenate was then centrifuged at 40 000 g for 20 min. The resulting pellet was resuspended in the same buffer for the binding assay. In the competition experiment, the tissue homogenate was incubated with [ 3 H] NMS (0.3 nM) in the presence of each drug. Incubations were performed at 25 C for 60 min and the reaction was terminated by rapid filtration (Cell Harvester; Brandel Co., Gaithersburg, MD, USA) through Whatman GF/B glass filters. Filters were rinsed three times with 3 mL of ice-cold 50 mM Na + /K + phosphate buffer (pH 7.4). Tissue-bound radioactivity was extracted from the filters overnight in scintillation fluid, and radioactivity was measured using a liquid scintillation counter. Specific [ 3 H]NMS binding was assessed experimentally from differences in counts between the absence and presence of 1 μM atropine. All steps were performed at 4 C to minimize the dissociation of drugs from receptor sites.

Data analysis
The assessment of muscarinic receptor-binding activity was based on the displacement of specific [ 3 H]NMS binding in brain tissues by drugs. [ 3 H]NMS binding data were subjected to a non-linear regression analysis using GraphPad PRISM (ver. 5; GraphPad Software, San Diego, CA, USA). The ability of non-labeled drugs to inhibit specific [ 3 H]NMS binding (0.5 nM) was estimated from IC 50 , the molar concentration of drugs needed to displace 50% of specific [ 3 H] NMS binding. The maximum blood concentration (C max ) of drugs after their administration to subjects was cited from their interview forms. The interview form is "Prescription drug information forms" provided by pharmaceutical companies from Pharmaceuticals and Medical Devices Agency (PMDA) in Japan, https://www.pmda.go.jp. Data were presented as the means of four to seven experiments, with the standard error of mean (SEM) in footnotes of tables.

Assessment of muscarinic receptor-binding activity
A radioreceptor binding assay with [ 3 H]NMS 8 was used to assess the anticholinergic activities of 260 medications commonly used in elderly Japanese. The significant competitive inhibition of specific [ 3 H]NMS binding in the rat brain was observed at a concentration of 100 μM by 96 of the 260 drugs examined. Based on these preliminary results, the concentration-dependent displacement of specific [ 3 H]NMS binding in the rat brain by five or six lower concentrations of 96 drugs was examined to estimate quantitatively their muscarinic receptor-binding activity (IC 50 ). These drugs displaced specific [ 3 H]NMS binding at concentrations between 0.01 nM and 100 μM in a concentration-dependent manner.

Definition of anticholinergic burden scale in consideration of blood concentrations of drugs in subjects after their administration at clinical doses
The evaluation of muscarinic receptor-binding activity under in vitro conditions may be limited in part because the drug concentrations tested may not reflect those in biological fluids under clinical conditions. To assess the clinical significance of the ABS rating based on muscarinic receptor-binding activity, we compared the IC 50 (Tables 1 and 2) of specific [ 3 H]NMS binding with C max in subjects after the administration of each drug at clinical doses. C max was cited from the interview form of each drug as described in the Methods section. As shown in Table 4, IC 50 values for the majority of drugs were ≤C max , while IC 50 values for 6 drugs (nortriptyline, clomipramine, diphenhydramine, propiverine, levomepromazine, difenidol) were greater than C max ;    however, this difference was less than three-fold. Generally, comparisons showed an approximate similarity between muscarinic receptor-binding activity (IC 50 of inhibition of specific [ 3 H]NMS binding) and C max for 33 drugs after their administration at clinical doses. Therefore, we defined these drugs with IC 50 ≤C max or within three-fold of C max as ABS 3, in consideration of their blood concentrations in subjects after their administration at clinical doses (Table 4). Notably, IC 50 for 10 drugs (disopyramide, nortriptyline, clomipramine, imipramine, diphenhydramine, propiverine, levomepromazine, zotepine, quetiapine, and difenidol) in Table 2, were approximately similar to their C max . Therefore, we included these drugs as ABS 3 in consideration of C max (Table 4). Furthermore, as shown in Table 5, 37 drugs with three-fold greater IC 50 than C max were rated as ABS 2. These included 8 drugs (desloratadine, tiotropium, glycopyrronium, ipratropium, butropium, mepenzolate, pridinol, and cloperastine) presented in Table 1, and 29 drugs presented in Table 2. Furthermore, 26 drugs presented in Table 3 were defined as ABS 1. The remainder (164 drugs) of 260 drugs, which displayed slight (<50%) or no inhibition of specific [ 3 H]NMS binding at 100 μM were defined as drugs with ABS 0 (negligible or no binding activity of muscarinic receptor) (Table S1). We compared muscarinic receptor-binding activity-based ABS of 33 drugs rated as ABS 3 (Table 4) with their previously reported scale rating in 12 studies. 10,[12][13][14][15][16][17][18][19][20][21][22] As shown in Table S2, there were marked similarities for 28 drugs between the present ABS data (Table 4) and previous scoring data in literatures, except for disopyramide, olanzapine, and quetiapine, which showed large variations (0-3) among studies. The scale of ABS 3 for aclidinium and umeclidinium (bronchodilators) was the first report.

Discussion
In the present study, the muscarinic receptor-binding activities of 260 drugs that are commonly used in elderly Japanese were Table 5 Drugs defined as anticholinergic burden scale 2 with their pharmacological classification and muscarinic receptor-binding activity (mean IC 50 ) in consideration of maximum blood drug concentrations (C max ) and clinical doses assessed by the displacement of specific [ 3 H]NMS binding in the rat brain. 8 In total, 96 drugs exhibited muscarinic receptorbinding activity in concentration-dependent manner. We defined the ABS rating of drugs, based on the muscarinic receptor-binding activity (IC 50 ) and blood concentration (C max ) after the administration at clinical doses in humans (Tables 4 and 5), and on quantitative assessments of muscarinic receptor-binding activity (Tables 1-3, S1). To our knowledge, the present study has developed the first comprehensive assessment of the anticholinergic activity of each medication by pharmacological methods. Previous studies evaluated the rating of anticholinergic scores using practitioners' knowledge of a list of drugs with known anticholinergic effects or serum anticholinergic activity. 10,[12][13][14][15][16][17][18][19][20][21][22][23]26,27 Marked variabilities and inconsistencies exist among these anticholinergic scales, including the number and potency score of anticholinergic medication. 23,28,29 The use of serum anticholinergic activity has limitations, such as the inability to assess the anticholinergic effects of each individual drug or the potentially distorting effects of endogenous serum proteins. 26 In previous studies, serum anticholinergic activity was not associated with anticholinergic adverse drug events 24 or four anticholinergic drug scoring systems. 25 None of the previous scales have comprehensively assessed the presence or absence of anticholinergic effects by each medication through pharmacological methods.
We herein demonstrated that 96 of 260 drugs displayed concentration-dependent muscarinic receptor binding in the rat brain. Based on the muscarinic receptor-binding activities (IC 50 ) of 96 drugs, we found 31 drugs with strong binding activity of muscarinic receptors exerting IC 50 less than 100 nM, 39 drugs with moderate binding activity exerting IC 50 ≥100 nM and <10 μM, and 26 drugs with weak binding activity exerting IC 50 ≥10 μM and <100 μM (Tables 1-3). All drugs in Table 1 concentration-dependently inhibited the cholinergic agonistevoked contractions of rat ileal smooth muscle in an organ bath assay, indicating their cholinergic antagonistic effects (our unpublished data).
The in vitro muscarinic receptor-binding activities of drugs may be limited in part because the concentrations tested may not reflect those in the human body under clinical conditions. Considerations of the anticholinergic potency and clinical dose of each drug may be required for the accurate quantification of the anticholinergic burden in assessments of the relative risks versus benefits of prescribing drugs. Therefore, we compared the muscarinic receptor-binding activity (IC 50 ) of each drug, which was presented in Tables 1,2, with its therapeutic blood concentration (C max ) following clinical dosing in humans. Based on the close approximation of IC 50 and C max , 33 drugs were defined as drugs with ABS 3 as shown in Table 4, which included 10 drugs in Table 2, in addition to 23 drugs in Table 1. Furthermore, 37 drugs with three-fold greater IC 50 than C max were finally rated as ABS 2 (Table 5), and 26 drugs presented in Table 3 were defined as ABS 1. The remainder (164 drugs), which displayed slight or no inhibition of specific [ 3 H]NMS binding at 100 μM, were defined as drugs with ABS 0 (negligible or no binding activity of muscarinic receptor) (Table S1).
The present muscarinic receptor-binding activity-based ABS of 33 drugs rated as ABS 3 was compared with their previously reported scale rating in 12 studies (Table S2). 10,[12][13][14][15][16][17][18][19][20][21][22] These scales assign scores to anticholinergic medications, with lower scores (typically 0-1) indicating no or the limited anticholinergic activity of a medication, intermediate scores (generally 2) indicating moderate anticholinergic activity, and high scores (usually 3) indicating strong anticholinergic activity. A cross-sectional study to validate the anticholinergic risk scale in 249 participants showed that higher anticholinergic risk scale scores were associated with greater numbers of anticholinergic adverse events, including confusion, falls, dry mouth, dry eyes, and constipation. 15 As shown in Table S2, there were marked similarities for 28 drugs between the present ABS data (Table 4) and previous scoring data in the literature. This result confirmed that drugs with clinically established anticholinergic effects in previous studies exhibited high affinities for muscarinic receptors.
A limitation of our newly developed ABS is that the anticholinergic activities of active metabolites that formed after the oral administration of drugs were not evaluated. In this case, the summation of the unchanged form and active metabolites needs to be considered. The muscarinic receptor-binding activities of drugs depend on drug concentrations in tissues rather than plasma concentrations. Therefore, pharmacokinetic factors, such as the tissue concentrations of drugs, need to be considered for the precise assessment of anticholinergic activity because some drugs accumulate within tissues through sustained binding to tissue proteins by their repeated administration, which may indicate a markedly higher drug concentration in tissues than in plasma. In older adults, the renal excretion and hepatic metabolism of drugs are reduced, which may allow drugs to accumulate in tissues, indicating enhanced anticholinergic effects. 30 An increasing age may also make the blood-brain barrier more permeable, which may be an issue because 70% of patients treated for overactive bladder with anticholinergics are 61-80 years old. 3 Much caution should be paid for the central adverse effects in case that drugs with ABS 3 and 2 drugs are prescribed in patients. Therefore, the accuracy of the ABS rating may be increased by considering tissue (brain etc.) drug concentrations after drug administration.
In conclusion, this newly developed pharmacological evidencebased burden scale has the potential as a practical tool for assessing the anticholinergic burden, specifically for risk reductions of anticholinergic adverse events in the poly-medicated elderly.