Library › Pharmacology › Drug interactions and dosing

Cardiovascular drug interactions, renal dosing, and dosing in older adults

Pharmacokinetic and pharmacodynamic interactions of cardiovascular drugs, dose adjustment for kidney and liver disease, pharmacogenomic variation, and prescribing and deprescribing in older adults.

Adverse drug events from cardiovascular medicines

Cardiovascular drugs combine three properties that make interactions dangerous: many have a narrow therapeutic index, many share a small number of metabolic and transport pathways, and the patients who take them are older, take several other drugs, and have declining kidney function. Anticoagulants, antiplatelet drugs, digoxin, antiarrhythmic drugs, and potassium-raising drugs account for a large share of preventable harm.

In a national surveillance study of emergency hospitalizations for adverse drug events in Americans aged 65 years and older, warfarin was implicated in about one-third of hospitalizations, and warfarin, insulins, oral antiplatelet agents, and oral hypoglycemic agents together accounted for about two-thirds. Drugs designated as high-risk or potentially inappropriate by quality measures caused only a small fraction. Harm therefore concentrates in drugs that are clearly indicated and difficult to use well, which is why dose and interaction review matter more than lists of drugs to avoid.

Population studies link specific interactions to hospitalization. Among older adults in Ontario, admission for digoxin toxicity was strongly associated with recent clarithromycin exposure, hypoglycemia with glyburide plus trimethoprim-sulfamethoxazole, and hyperkalemia with an angiotensin-converting enzyme (ACE) inhibitor plus a potassium-sparing diuretic. Each pair has a mechanism (P-glycoprotein inhibition, CYP2C9 inhibition, and additive potassium retention), and each was predictable at the time of prescribing.

Five factors determine whether a combination is contraindicated, managed with a dose change, or of no clinical consequence. They are the magnitude of the exposure change, the therapeutic index of the affected drug, the time course of the interaction, the patient's organ function, and the availability of a monitoring test.

Pharmacokinetic interaction mechanisms

Pharmacokinetic interactions change the concentration of a drug at its site of action. They are classified by the step they affect: absorption, distribution, metabolism, or excretion.

Absorption. Bile acid sequestrants (cholestyramine, colestipol) bind digoxin, warfarin, amiodarone, and thyroid hormone in the gut; separation of doses by several hours reduces the effect, and cholestyramine also shortens the elimination half-life of amiodarone. Antacids, sucralfate, and metoclopramide lower digoxin absorption. Gastric pH matters for a few drugs: dabigatran etexilate is formulated with a tartaric acid core to ensure dissolution, and acid suppression lowers its bioavailability modestly without a labeled dose change. Food changes absorption of rivaroxaban at the 15- and 20-mg doses, which are taken with the evening meal because fasting bioavailability is incomplete. Intestinal P-glycoprotein (P-gp) and intestinal CYP3A4 together limit the oral bioavailability of many substrates, so inhibitors and inducers act partly at the gut wall; grapefruit juice acts almost entirely there.

Distribution and protein binding. Displacement of a highly protein-bound drug from albumin raises the free fraction transiently, but clearance of the free drug rises in parallel and the steady-state free concentration returns to baseline. Displacement alone rarely causes clinical harm. Interactions once attributed to displacement, such as warfarin with sulfonamides, are explained by inhibition of CYP2C9. Displacement matters mainly for interpretation: in hypoalbuminemia or uremia, a total (bound plus free) concentration of phenytoin or another highly bound drug underestimates the active free concentration.

Metabolism. Enzyme inhibition is the most common mechanism of serious cardiovascular interactions. Reversible competitive inhibition begins with the first doses of the inhibitor and resolves as it is cleared. Mechanism-based (irreversible) inhibition, produced by clarithromycin, erythromycin, diltiazem, verapamil, and the furanocoumarins of grapefruit, inactivates the enzyme; recovery requires synthesis of new enzyme and takes days. Induction (rifampin, carbamazepine, phenytoin, St John's wort) increases transcription of CYP3A4, other CYP enzymes, and P-glycoprotein through the pregnane X receptor. Its effect develops over 1 to 2 weeks and persists for about 2 weeks after the inducer is stopped. The dose of the affected drug is therefore adjusted both when the inducer is started and when it is withdrawn.

The US Food and Drug Administration (FDA) classifies inhibitors by their effect on the area under the concentration-time curve (AUC) of a sensitive index substrate. Strong inhibitors raise the AUC 5-fold or more, moderate inhibitors 2- to less than 5-fold, and weak inhibitors 1.25- to less than 2-fold. Strong inducers lower the AUC by 80% or more, moderate inducers by 50% to less than 80%, and weak inducers by 20% to less than 50%. Prescribing information uses these categories, so "avoid with strong CYP3A inhibitors" refers to the FDA list.

Excretion. Renal elimination of cationic drugs depends on tubular secretion by organic cation transporter 2 (OCT2) and the multidrug and toxin extrusion proteins (MATE1 and MATE2-K). Cimetidine, trimethoprim, dolutegravir, and ranolazine inhibit this pathway. Dofetilide, which is about 80% renally excreted, is contraindicated with cimetidine, trimethoprim, and dolutegravir for this reason, and metformin exposure rises with the same inhibitors. Renal P-glycoprotein secretes digoxin, which explains much of the rise in digoxin with amiodarone, dronedarone, verapamil, and quinidine.

Several drugs inhibit tubular secretion of creatinine without changing glomerular filtration: trimethoprim, cimetidine, dolutegravir, cobicistat, and dronedarone raise serum creatinine, typically by about 0.1 mg/dL with dronedarone. The rise appears within days, stabilizes, and reverses on withdrawal. It is distinguished from true loss of kidney function by its size and timing, by stable urea and cystatin C, and by the absence of other features of kidney injury. It still lowers the calculated creatinine clearance used for dosing, which can lead to unnecessary dose reduction.

Cytochrome P450 enzymes and drug transporters in cardiovascular prescribing

CYP3A4 metabolizes the largest number of cardiovascular drugs: simvastatin, lovastatin, and atorvastatin; amlodipine, felodipine, nifedipine, diltiazem, and verapamil; amiodarone, dronedarone, quinidine, disopyramide, and ranolazine; ivabradine, eplerenone, and finerenone. Other substrates are ticagrelor, cilostazol, apixaban and rivaroxaban (in part), and the calcineurin and mammalian target of rapamycin (mTOR) inhibitors used after heart transplantation. Strong CYP3A inhibitors on the FDA list include clarithromycin, itraconazole, ketoconazole, posaconazole, voriconazole, ritonavir, cobicistat, and nefazodone. Moderate inhibitors include diltiazem, verapamil, dronedarone, erythromycin, fluconazole, and grapefruit juice. Strong inducers include rifampin, carbamazepine, phenytoin, enzalutamide, apalutamide, and St John's wort; bosentan, efavirenz, and phenobarbital are moderate inducers.

CYP2C9 clears S-warfarin, the more potent enantiomer, as well as fluvastatin, losartan and irbesartan in part, and most nonsteroidal anti-inflammatory drugs (NSAIDs). The FDA lists amiodarone, fluconazole, and miconazole as moderate CYP2C9 inhibitors; metronidazole and sulfamethoxazole raise the international normalized ratio (INR) through the same pathway.

CYP2C19 converts clopidogrel to its active metabolite. Strong CYP2C19 inhibitors are fluconazole, fluoxetine, fluvoxamine, and ticlopidine; omeprazole and esomeprazole inhibit the enzyme enough to lower clopidogrel's antiplatelet effect, and US labeling advises against combining them with clopidogrel.

CYP2D6 metabolizes metoprolol, carvedilol, propranolol, nebivolol, timolol, flecainide, propafenone, and mexiletine in part. Strong CYP2D6 inhibitors are bupropion, fluoxetine, paroxetine, quinidine, and terbinafine; duloxetine, cinacalcet, mirabegron, and abiraterone are moderate inhibitors. CYP2D6 is not inducible, so variation comes from inhibition and from genotype.

CYP2C8 contributes to the metabolism of amiodarone, repaglinide, and pioglitazone. Gemfibrozil is the strong CYP2C8 inhibitor on the FDA list, and the acyl glucuronide of clopidogrel inhibits CYP2C8 and raises repaglinide exposure.

P-glycoprotein is an efflux transporter in the intestine, the renal tubule, the biliary canaliculus, and the blood-brain barrier. Its sensitive substrates include digoxin, dabigatran etexilate, and edoxaban; apixaban and rivaroxaban are substrates of both P-glycoprotein and CYP3A4. Amiodarone, dronedarone, verapamil, quinidine, propafenone, ranolazine, clarithromycin, erythromycin, itraconazole, ketoconazole, cyclosporine, and ritonavir inhibit it. Rifampin, carbamazepine, phenytoin, and St John's wort induce it.

Organic anion transporting polypeptides 1B1 and 1B3 (OATP1B1 and OATP1B3) take statins up into hepatocytes. Inhibition raises systemic statin exposure, the concentration that drives myopathy. Cyclosporine, gemfibrozil, clarithromycin, rifampin (acutely), darolutamide, teriflunomide, and the boosted protease inhibitors atazanavir and lopinavir inhibit OATP1B1. Every statin is a substrate, so switching statins avoids a CYP3A4-mediated interaction but not an OATP-mediated one. Rosuvastatin is also a substrate of the efflux transporter breast cancer resistance protein (BCRP), inhibited by cyclosporine, febuxostat, ticagrelor, and several kinase inhibitors.

Mechanisms of cardiovascular drug interactions by enzyme, transporter, and additive effect. Inhibitor and inducer categories follow the FDA table of substrates, inhibitors, and inducers. The table summarizes mechanisms; dose limits appear in the drug sections below.

Pharmacodynamic interactions: QT, potassium, bleeding, blood pressure, and conduction

Pharmacodynamic interactions occur when two drugs act on the same physiological system, with no change in either drug's concentration. They are predictable from pharmacology and are the most common interactions in cardiovascular practice.

Additive QT prolongation. Drugs that block the rapid component of the delayed rectifier potassium current (I_Kr) prolong repolarization, and their effects add. Risk of torsades de pointes rises with female sex, hypokalemia, hypomagnesemia, bradycardia, heart failure, recent conversion from atrial fibrillation, a long baseline QT interval, and a pharmacokinetic interaction that raises the concentration of one of the drugs. In a cohort study, erythromycin doubled the rate of sudden cardiac death, and the rate was about five times higher when erythromycin was combined with a CYP3A inhibitor such as diltiazem, verapamil, or an azole antifungal. The American Heart Association (AHA) statement on drug-induced arrhythmias reviews these risk factors. QT measurement, drug categories, monitoring thresholds, and the treatment of torsades are covered on the long QT syndrome and QT prolongation page.

Additive potassium retention. ACE inhibitors, ARBs, sacubitril-valsartan, mineralocorticoid receptor antagonists, potassium-sparing diuretics, trimethoprim, NSAIDs, heparin, and calcineurin inhibitors each lower renal potassium excretion. After publication of the Randomized Aldactone Evaluation Study (RALES) in 1999, spironolactone prescriptions for heart failure rose sharply in Ontario. Among older adults recently hospitalized for heart failure and taking ACE inhibitors, the rate of hospitalization for hyperkalemia rose from 2.4 to 11.0 per 1,000 patients and associated in-hospital mortality rose from 0.3 to 2.0 per 1,000. Heart failure readmissions and all-cause mortality did not fall significantly. Trimethoprim blocks the epithelial sodium channel in the collecting duct in the same way as amiloride. In older adults taking an ACE inhibitor or ARB, trimethoprim-sulfamethoxazole was associated with about a 7-fold higher risk of hospitalization for hyperkalemia than amoxicillin. The 2023 Beers Criteria advise caution with trimethoprim-sulfamethoxazole in patients taking an ACE inhibitor, ARB, or angiotensin receptor-neprilysin inhibitor (ARNI) with reduced creatinine clearance. They also advise against routine use of two or more renin-angiotensin system inhibitors or potassium-sparing diuretics together in chronic kidney disease (CKD) stage 3a or worse.

Bleeding. An oral anticoagulant combined with aspirin, a P2Y12 inhibitor, an NSAID, or a selective serotonin reuptake inhibitor (SSRI) raises bleeding risk without any change in anticoagulant concentration. SSRIs deplete platelet serotonin, and the Beers Criteria list warfarin with an SSRI among combinations to avoid when possible. Ibrutinib and other Bruton tyrosine kinase inhibitors impair platelet signaling and add to anticoagulant bleeding. The duration of combined antithrombotic therapy after percutaneous coronary intervention in patients with atrial fibrillation is covered on the atrial fibrillation page.

Blood pressure. Organic nitrates with a phosphodiesterase type 5 (PDE5) inhibitor produce severe hypotension by sequential amplification of cyclic guanosine monophosphate signaling. Nitrates are not given within 24 hours of sildenafil or within 48 hours of tadalafil, and riociguat is contraindicated with both nitrates and PDE5 inhibitors. Alpha-1 blockers given for prostatic symptoms add orthostatic hypotension to antihypertensive therapy; the Beers Criteria advise against nonselective peripheral alpha-1 blockers as antihypertensives and against combining them with loop diuretics in older women because of urinary incontinence.

Conduction and heart rate. Beta-blockers, verapamil, diltiazem, digoxin, amiodarone, dronedarone, ivabradine, and clonidine slow the sinus rate or atrioventricular (AV) conduction, and their effects add. Verapamil or diltiazem with a beta-blocker can cause high-grade AV block or asystole, particularly in older adults, with intravenous dosing, or with conduction disease. Sofosbuvir-containing hepatitis C regimens with amiodarone have caused symptomatic bradycardia requiring pacemaker implantation. Fentanyl with amiodarone can cause hypotension and bradycardia.

Heart failure and fluid retention. In heart failure with reduced ejection fraction, the 2022 AHA/American College of Cardiology (ACC)/Heart Failure Society of America (HFSA) heart failure guideline assigns a Class 3: Harm recommendation to nondihydropyridine calcium channel blockers, class IC antiarrhythmic drugs and dronedarone, thiazolidinediones, the dipeptidyl peptidase-4 (DPP-4) inhibitors saxagliptin and alogliptin in type 2 diabetes, and NSAIDs. Dronedarone labeling also contraindicates the drug in recently decompensated or New York Heart Association (NYHA) class IV heart failure. NSAIDs blunt the natriuretic effect of loop diuretics and the antihypertensive effect of most drug classes. A diuretic, a renin-angiotensin system inhibitor, and an NSAID together remove the prostaglandin- and angiotensin-dependent defenses of glomerular filtration. In a nested case-control study, this triple combination was associated with a 31% higher risk of acute kidney injury, with the highest risk in the first 30 days. Drugs that cause or worsen heart failure are listed in the AHA statement on that topic.

Pharmacodynamic interactions of cardiovascular drugs. eGFR, estimated glomerular filtration rate; ARNI, angiotensin receptor-neprilysin inhibitor; PDE5, phosphodiesterase type 5; QTc, corrected QT interval.

Interactions with amiodarone and dronedarone

Amiodarone is the most interaction-prone drug in cardiology. It is metabolized to its active metabolite desethylamiodarone by CYP3A and CYP2C8, and in vitro the parent drug and metabolite inhibit CYP2C9, CYP2C19, CYP2D6, CYP3A, CYP2A6, CYP2B6, and CYP2C8, as well as P-glycoprotein and OCT2. Its large volume of distribution and slow tissue release give a mean terminal half-life of about 58 days (range 15 to 142 days) for amiodarone and about 36 days for desethylamiodarone in the US label. Three consequences follow. First, interactions develop gradually over weeks of loading, so an INR or digoxin concentration that is acceptable in the first week can be toxic by the fourth. Second, interactions persist for months after amiodarone is stopped, so a warfarin dose reduced at the start of amiodarone is increased gradually after amiodarone is withdrawn, guided by the INR. Finally, the drug's own effects (bradycardia, QT prolongation, thyroid dysfunction) interact pharmacodynamically with other drugs.

The amiodarone label specifies these adjustments. The warfarin dose is reduced by one-third to one-half with prothrombin time monitoring, because coadministration can roughly double the prothrombin time within 3 to 4 days. The digoxin dose is reduced by half, or digoxin is stopped. Simvastatin is limited to 20 mg daily and lovastatin to 40 mg daily. Cyclosporine concentrations and renal function are monitored. Lidocaine is started at a lower dose. Other class I and III antiarrhythmics are started at lower than usual doses, and previously administered antiarrhythmics are reduced by 30% to 50%. Other QT-prolonging drugs and grapefruit juice, which raised amiodarone AUC by 50%, are avoided. Cholestyramine and St John's wort lower amiodarone concentrations. Oral amiodarone raises dabigatran exposure through P-glycoprotein inhibition without a labeled dose change in atrial fibrillation with preserved kidney function.

Dronedarone is a moderate CYP3A inhibitor and a P-glycoprotein inhibitor, and it is itself a CYP3A substrate. Its label contraindicates strong CYP3A inhibitors (ketoconazole, itraconazole, voriconazole, cyclosporine, clarithromycin, nefazodone, ritonavir, and telithromycin), erythromycin, and other drugs that prolong the QT interval, and it recommends avoiding rifampin and other CYP3A inducers. Digoxin is stopped or its dose halved. Simvastatin is limited to 10 mg daily. Beta-blockers and calcium channel blockers are started at low doses and increased only after ECG confirmation of tolerance. Tacrolimus, sirolimus, and other narrow-index CYP3A substrates are monitored and adjusted. Dabigatran is reduced to 75 mg twice daily with dronedarone in patients with moderate kidney dysfunction (creatinine clearance 30 to 50 mL/min). The 2023 Beers Criteria advise avoiding dronedarone in permanent atrial fibrillation and in severe or recently decompensated heart failure and caution in reduced ejection fraction with class I or II symptoms.

Dose adjustments of coadministered drugs with amiodarone and dronedarone, from the US prescribing information for amiodarone, dronedarone, and lovastatin. CrCl, creatinine clearance. Amiodarone recommendations apply for months after the drug is stopped.

Digoxin interactions and dosing

Digoxin has a narrow therapeutic range, a large apparent volume of distribution (about 475 to 500 L) bound mainly to skeletal muscle, and renal elimination of 50% to 70% of a dose as unchanged drug. The half-life is 1.5 to 2 days with normal kidney function and 3.5 to 5 days in anuric patients, so a change in dose or in an interacting drug reaches a new steady state only after 1 to 3 weeks. Digoxin is not removed effectively by dialysis.

The US digoxin label groups interacting drugs by the size of the effect. Drugs that raise digoxin concentrations by more than 50% include amiodarone (about 70%), dronedarone (AUC about 150%), propafenone (AUC 60% to 270%), verapamil (50% to 75%), clarithromycin, captopril, quinidine, and ranolazine. For these, the serum digoxin concentration is measured before the interacting drug is started, and the digoxin dose is reduced by about 30% to 50% or the dosing interval extended. Drugs that raise concentrations by less than 50%, including atorvastatin, carvedilol, diltiazem, nifedipine, spironolactone, ticagrelor, and trimethoprim, call for a reduction of about 15% to 30% with monitoring. Azithromycin, cyclosporine, omeprazole, ibuprofen, and metformin raise concentrations by an undefined amount. Antacids, cholestyramine, metoclopramide, phenytoin, rifampin, St John's wort, sucralfate, and sulfasalazine lower concentrations, sometimes enough to require a 20% to 40% dose increase.

Pharmacodynamic interactions also contribute to toxicity. Hypokalemia, hypomagnesemia, and hypercalcemia increase digoxin toxicity at any concentration, so a loop or thiazide diuretic is a common cause of toxicity without a change in digoxin dose. Rapid intravenous calcium in a patient taking digoxin can provoke serious arrhythmias. ACE inhibitors, ARBs, and NSAIDs can reduce digoxin excretion by lowering glomerular filtration. Beta-blockers, nondihydropyridine calcium channel blockers, and ivabradine add to bradycardia and AV block. Dofetilide combined with digoxin was associated with more torsades de pointes in the dofetilide trials, and dronedarone with digoxin with more sudden death.

The digoxin label provides maintenance doses by lean body weight and creatinine clearance. A patient with a creatinine clearance of 10 mL/min and a lean weight of 40 kg receives 62.5 mcg daily, and steady state takes about 19 days at that clearance. In obesity, lean or ideal body weight is used instead of total weight because digoxin does not distribute into fat. Concentrations are drawn at least 6 hours after a dose. The label notes that concentrations above 2 ng/mL are associated with toxicity without added benefit. In a post hoc analysis of the Digitalis Investigation Group trial, men with heart failure and concentrations of 0.5 to 0.8 ng/mL had lower mortality than those assigned placebo. Concentrations of 1.2 ng/mL or higher were associated with higher mortality. This is the basis for targeting the low end of the range. The 2023 Beers Criteria advise against digoxin as first-line therapy for atrial fibrillation or heart failure and, when it is used, against doses above 0.125 mg daily in older adults. They also caution against abrupt withdrawal in long-term users with reduced ejection fraction.

Warfarin interactions

Warfarin is a racemate. S-warfarin, three to five times more potent than R-warfarin, is cleared by CYP2C9; R-warfarin by CYP1A2 and CYP3A4. Inhibition of CYP2C9 produces the largest INR changes. Amiodarone, fluconazole, miconazole (including oral gel and vaginal preparations), metronidazole, and trimethoprim-sulfamethoxazole raise the INR substantially. Rifampin, carbamazepine, phenytoin, and St John's wort lower it, and the warfarin dose may need to double or more during rifampin therapy, then fall again over the weeks after rifampin is stopped. The 2023 Beers Criteria list warfarin with amiodarone, ciprofloxacin, macrolides other than azithromycin, trimethoprim-sulfamethoxazole, or SSRIs as combinations to avoid when possible, with close INR monitoring when they cannot be avoided.

Diet and nonprescription drugs also change the INR. Vitamin K intake from leafy green vegetables, nutritional supplements, and enteral feeding opposes warfarin; the aim is consistent intake without avoidance. Acetaminophen raises the INR in a dose-dependent way: in a case-control study, regular intake above about 9 g per week was associated with a markedly higher risk of an INR above 6. Acute alcohol excess inhibits warfarin metabolism, whereas chronic heavy use induces it and adds liver disease. Antibiotics as a class raise the INR through reduced vitamin K-producing gut flora, decreased intake during illness, and fever-related increases in clotting factor catabolism, in addition to specific CYP2C9 effects. Thyroid hormone replacement increases warfarin response by accelerating the catabolism of clotting factors; hyperthyroidism, including amiodarone-induced thyrotoxicosis, raises the INR. Statins (rosuvastatin, simvastatin) can raise the INR modestly, and their labels advise an INR after initiation or dose change.

Interactions are managed by anticipation. The INR is checked within 3 to 7 days of starting or stopping an interacting drug, and an empiric dose change is made at initiation when the interaction is large and predictable, as with amiodarone, fluconazole, or trimethoprim-sulfamethoxazole. The 2023 Beers Criteria advise against starting warfarin for nonvalvular atrial fibrillation or venous thromboembolism unless DOACs are contraindicated or there are substantial barriers to their use.

Interactions with direct oral anticoagulants

All four DOACs are substrates of P-glycoprotein. Apixaban and rivaroxaban are also CYP3A4 substrates; edoxaban undergoes minimal CYP3A4 metabolism; dabigatran etexilate is a prodrug hydrolyzed by esterases and is not metabolized by CYP enzymes. Renal elimination is about 80% for dabigatran, about 50% for edoxaban, about one-third for rivaroxaban as unchanged drug, and about 27% for apixaban, which determines how strongly kidney function amplifies an interaction.

Dose adjustment of direct oral anticoagulants with interacting drugs, from US prescribing information. P-gp, P-glycoprotein; CrCl, Cockcroft-Gault creatinine clearance with actual body weight. Dose selection by kidney function, age, and weight for atrial fibrillation is calculated on the atrial fibrillation page.

Observational data confirm that these interactions change outcomes. In a Taiwanese cohort of patients with atrial fibrillation taking DOACs, concurrent amiodarone, fluconazole, rifampin, or phenytoin was associated with a higher rate of major bleeding than DOAC use alone. Concurrent atorvastatin, digoxin, and erythromycin or clarithromycin were associated with lower rates. Interpretation is limited by confounding, but the signal for azoles and inducers is consistent with pharmacology. The 2023 Beers Criteria advise avoiding rivaroxaban for long-term treatment of atrial fibrillation or venous thromboembolism in older adults. They also advise caution in choosing dabigatran over apixaban, based on higher bleeding rates in comparative observational studies.

European guidance differs mainly for apixaban and strong dual inhibitors: the 2021 European Heart Rhythm Association practical guide and the European label advise against these combinations, whereas US labeling allows apixaban at half dose. When a strong inducer such as rifampin or carbamazepine is unavoidable, warfarin dosed to the INR or a parenteral anticoagulant is the usual alternative to a DOAC.

Antiplatelet drug interactions: aspirin, clopidogrel, prasugrel, and ticagrelor

Clopidogrel and CYP2C19. Clopidogrel is a prodrug converted to its active thiol metabolite in two oxidative steps in which CYP2C19 has a major role. Omeprazole and esomeprazole lower active metabolite exposure and platelet inhibition, and US labeling advises against combining them with clopidogrel; pantoprazole and other proton pump inhibitors with weaker CYP2C19 inhibition are used when gastroprotection is indicated. The Clopidogrel and the Optimization of Gastrointestinal Events Trial (COGENT) randomized patients taking clopidogrel and aspirin to omeprazole or placebo. Omeprazole reduced gastrointestinal events (1.1% versus 2.9% at 180 days) without a significant difference in cardiovascular events, although the trial ended early and was underpowered for cardiovascular outcomes. Strong CYP2C19 inhibitors such as fluconazole, fluoxetine, and fluvoxamine are avoided with clopidogrel when an alternative exists. The clopidogrel acyl glucuronide inhibits CYP2C8 and raises repaglinide concentrations, causing hypoglycemia.

Ticagrelor. Ticagrelor is a CYP3A4 substrate and is avoided with strong CYP3A inhibitors and strong CYP3A inducers. Aspirin maintenance doses above 100 mg daily reduce its effectiveness and are avoided, a finding from the North American subgroup of the PLATO trial. Ticagrelor raises simvastatin and lovastatin exposure, and doses of these statins above 40 mg daily are avoided; the rosuvastatin label limits the dose to 20 mg daily with ticagrelor because of BCRP inhibition. Digoxin concentrations are monitored when ticagrelor is started or changed.

Opioids and oral P2Y12 inhibitors. Opioids slow gastric emptying and delay absorption of all oral P2Y12 inhibitors. In the randomized IMPRESSION trial in patients with myocardial infarction, intravenous morphine lowered ticagrelor exposure by about one-third and delayed platelet inhibition. The ticagrelor label notes reduced exposure with opioids and lists a parenteral antiplatelet agent as an option in acute coronary syndromes when opioids are required.

Aspirin and NSAIDs. Ibuprofen taken before aspirin occupies the cyclooxygenase-1 (COX-1) channel and prevents aspirin's irreversible acetylation of platelet COX-1. Immediate-release aspirin is therefore taken at least 30 minutes before, or more than 8 hours after, ibuprofen when the two must be combined. Celecoxib does not block aspirin's platelet effect. NSAIDs also add gastrointestinal bleeding risk to any antiplatelet or anticoagulant regimen.

Prasugrel is activated mainly by CYP3A and CYP2B6, with little contribution from CYP2C19, and has few clinically important pharmacokinetic interactions. The 2023 Beers Criteria advise caution with prasugrel and ticagrelor in adults aged 75 years and older and suggest a 5-mg prasugrel maintenance dose when prasugrel is used in that age group.

Statin interactions and dose limits

Statin myopathy is concentration dependent, and most serious cases involve an interacting drug. The 2016 AHA scientific statement on statin interactions organizes management by mechanism: CYP3A4 inhibition affects simvastatin and lovastatin most and atorvastatin to a lesser degree; OATP1B1 inhibition affects all statins; CYP2C9 inhibition affects fluvastatin. Pravastatin, rosuvastatin, and pitavastatin are minimally metabolized by CYP3A4 and are the usual substitutes when a CYP3A4 inhibitor is required, but they remain sensitive to OATP1B1 inhibitors such as cyclosporine and gemfibrozil.

Daily dose limits from current US prescribing information for each statin and from the 2016 AHA statement. Simvastatin 80 mg is restricted to patients who have taken it long term without myopathy. Limits for simvastatin, lovastatin, and atorvastatin with macrolides, azoles, and protease inhibitors reflect CYP3A4 inhibition; limits for rosuvastatin, pravastatin, and pitavastatin reflect OATP1B1 inhibition and, for rosuvastatin, BCRP inhibition. Current fluvastatin labeling advises avoiding fluvastatin with cyclosporine, fluconazole, or gemfibrozil.

Fibrates, niacin, and colchicine. Gemfibrozil inhibits statin glucuronidation and OATP1B1 and is contraindicated or avoided with every statin; fenofibrate has a much smaller pharmacokinetic effect and is the fibrate used with statins, with attention to kidney function. Lipid-modifying doses of niacin (1 g or more daily) add myopathy risk, and simvastatin labeling advises against this combination in Chinese patients, who have a higher incidence of myopathy. Colchicine with a statin has caused myopathy and rhabdomyolysis, a combination made common by colchicine's use in chronic coronary disease and pericarditis. Daptomycin is a separate cause of myopathy, and the simvastatin label suggests suspending the statin during daptomycin therapy.

Grapefruit juice. Furanocoumarins irreversibly inactivate intestinal CYP3A4, so the effect of a single glass lasts about a day and regular intake produces continuous inhibition. In a crossover study, one 200-mL glass of regular-strength juice daily raised simvastatin AUC by about 3.6-fold. Simvastatin and lovastatin labeling advise avoiding grapefruit juice; atorvastatin labeling advises against more than 1.2 L daily.

Population and kidney factors. Rosuvastatin is started at 5 mg in patients of Asian ancestry because exposure is about twice as high, and is started at 5 mg with a 10-mg maximum in severe kidney impairment not treated with hemodialysis. Simvastatin is started at 5 mg in severe kidney impairment. Statins are contraindicated in acute liver failure and decompensated cirrhosis. Statin intensity, intolerance, and alternatives are covered on the dyslipidemia and LDL lowering page.

Calcium channel blocker, beta-blocker, ranolazine, ivabradine, and nitrate interactions

Verapamil and diltiazem are moderate CYP3A inhibitors and CYP3A substrates; verapamil also inhibits P-glycoprotein. They raise concentrations of simvastatin, lovastatin, cyclosporine, tacrolimus, sirolimus, carbamazepine, digoxin (verapamil more than diltiazem), dofetilide (verapamil is contraindicated with it), and ivabradine (avoided). Their own concentrations rise with clarithromycin, erythromycin, and azole antifungals and fall with rifampin and other inducers. Pharmacodynamically, they add to beta-blockers in slowing AV conduction and depressing contractility.

Dihydropyridine calcium channel blockers are CYP3A substrates with high first-pass extraction. Felodipine and nifedipine exposure rises several-fold with grapefruit juice and strong CYP3A inhibitors; amlodipine, with high oral bioavailability, is affected less. Simvastatin labeling limits simvastatin to 20 mg daily with amlodipine. In older adults in Ontario taking a calcium channel blocker, clarithromycin was associated with about twice the risk of hospitalization with acute kidney injury (odds ratio 1.98) and a higher risk of hospitalization with hypotension (odds ratio 1.60) compared with azithromycin, which does not inhibit CYP3A4. The absolute increase in kidney injury was 0.22 percentage points, and the association was strongest with nifedipine.

Beta-blockers. Metoprolol, carvedilol, propranolol, timolol, and nebivolol are cleared in part by CYP2D6. Fluoxetine, paroxetine, bupropion, quinidine, and terbinafine (strong inhibitors) and duloxetine, cinacalcet, and mirabegron (moderate inhibitors) raise metoprolol exposure several-fold, producing bradycardia, hypotension, and fatigue; a more renally cleared beta-blocker such as bisoprolol (about half renal) or atenolol avoids the interaction. Beta-blockers blunt the response to epinephrine during anaphylaxis, mask the adrenergic symptoms of hypoglycemia (sweating is preserved), and, combined with clonidine, can cause severe rebound hypertension if clonidine is withdrawn first. Unopposed alpha-stimulation follows nonselective beta-blockade in pheochromocytoma and in acute cocaine or amphetamine toxicity.

Ranolazine is a CYP3A and P-glycoprotein substrate and a weak inhibitor of CYP3A, CYP2D6, P-glycoprotein, and OCT2. It is contraindicated with strong CYP3A inhibitors, with CYP3A inducers, and in liver cirrhosis. With moderate CYP3A inhibitors such as diltiazem, verapamil, and erythromycin, the dose is limited to 500 mg twice daily. Simvastatin is limited to 20 mg daily with any ranolazine dose, and metformin to 1,700 mg daily with ranolazine 1,000 mg twice daily. Ranolazine prolongs the QTc in a dose-related manner without the torsades risk of pure I_Kr blockers. Acute kidney injury has been reported in patients with severe kidney impairment, and kidney function is monitored after initiation and periodically in moderate or severe impairment.

Ivabradine is a sensitive CYP3A substrate. It is contraindicated with strong CYP3A inhibitors and avoided with moderate inhibitors (diltiazem, verapamil, grapefruit juice) and with CYP3A inducers. Diltiazem and verapamil are excluded for a pharmacodynamic reason as well, because they add to ivabradine's slowing of the sinus rate.

Nitrates. The hypotensive interaction with PDE5 inhibitors and riociguat is described above. Sildenafil for pulmonary arterial hypertension and tadalafil are also CYP3A substrates whose exposure rises with ritonavir and other strong inhibitors.

Renin-angiotensin, neprilysin, mineralocorticoid, SGLT2, and diuretic interactions

Renin-angiotensin system inhibitors. Dual blockade with an ACE inhibitor and an ARB, or either with aliskiren, increased hyperkalemia, hypotension, and kidney dysfunction without cardiovascular benefit in the ONTARGET and ALTITUDE trials and is avoided. Sacubitril-valsartan is not combined with an ACE inhibitor and is started at least 36 hours after the last ACE inhibitor dose, because combined neprilysin and ACE inhibition raises bradykinin and the risk of angioedema; no washout is needed after an ARB. Sacubitril inhibits OATP1B1 and OATP1B3 and raises atorvastatin peak concentrations. Lithium clearance falls with ACE inhibitors, ARBs, and ARNIs, and the 2023 Beers Criteria advise avoiding these combinations or monitoring lithium concentrations.

Mineralocorticoid receptor antagonists. Eplerenone is a sensitive CYP3A substrate: it is contraindicated with strong CYP3A inhibitors, and with moderate inhibitors (erythromycin, fluconazole, verapamil, saquinavir) it is limited to 25 mg daily in heart failure after myocardial infarction. Finerenone is contraindicated with strong CYP3A inhibitors, is avoided with strong or moderate inducers, and requires potassium monitoring with moderate or weak inhibitors. Spironolactone is not CYP3A-dependent but has the largest antiandrogenic effects. All three add to the potassium-raising drugs listed earlier; potassium and creatinine are checked within the first week after initiation or dose increase and again after any interacting drug is added.

Sodium-glucose cotransporter 2 (SGLT2) inhibitors. Empagliflozin, dapagliflozin, canagliflozin, ertugliflozin, and sotagliflozin have few pharmacokinetic interactions, but several pharmacodynamic ones. They add to the volume effect of loop and thiazide diuretics, and a loop diuretic dose is sometimes reduced when an SGLT2 inhibitor is started in a euvolemic patient. With insulin or a sulfonylurea, they increase the risk of hypoglycemia, and the insulin or sulfonylurea dose is often lowered when glycated hemoglobin is near target. They lower serum lithium concentrations, which are monitored. Canagliflozin exposure falls with UDP-glucuronosyltransferase inducers such as rifampin and phenytoin, and its labeling suggests a higher dose when an inducer is added. Ketoacidosis, which can occur with normal or modestly raised glucose, is most likely during acute illness, reduced oral intake, insulin dose reduction, alcohol excess, or surgery. US labeling advises stopping canagliflozin, dapagliflozin, and empagliflozin at least 3 days and ertugliflozin at least 4 days before scheduled surgery. The 2023 Beers Criteria add SGLT2 inhibitors to the drugs used with caution in older adults, with monitoring for urogenital infection and ketoacidosis.

Sick-day guidance. Kidney Disease: Improving Global Outcomes (KDIGO) 2024 Practice Point 4.3.2 states that when medications are stopped during an acute illness, a clear plan for restarting them is communicated to the patient and to other clinicians and documented. Sick-day guidance during a dehydrating illness typically covers ACE inhibitors, ARBs, diuretics, mineralocorticoid antagonists, SGLT2 inhibitors, metformin, sulfonylureas, and NSAIDs. KDIGO notes that randomized evidence that sick-day rules prevent kidney injury is lacking and that the most common error is failure to restart the drugs. Practice Point 4.3.3 allows planned discontinuation of metformin, ACE inhibitors, ARBs, and SGLT2 inhibitors 48 to 72 hours before elective surgery, again with a restart plan.

Diuretics. Thiazides and loop diuretics cause hypokalemia and hypomagnesemia, which increase digoxin toxicity and the risk of torsades with QT-prolonging drugs; dofetilide is contraindicated with hydrochlorothiazide alone or combined with triamterene. Thiazides lower lithium clearance more than loop diuretics, and both are on the Beers list of combinations to avoid with lithium. Thiazides and, less often, loop diuretics cause hyponatremia, particularly in older women and with SSRIs, and the Beers Criteria list diuretics among drugs that can cause or worsen hyponatremia. Loop diuretics in high intravenous doses add to the ototoxicity of aminoglycosides. NSAIDs blunt the natriuretic response to loop diuretics, and probenecid reduces their tubular secretion. Thiazides raise uric acid and blood glucose modestly.

Interactions of other antiarrhythmic drugs

Dofetilide is cleared about 80% by the kidney through filtration and OCT2-mediated secretion, and in part by CYP3A4. It is contraindicated with verapamil, cimetidine, trimethoprim (alone or with sulfamethoxazole), ketoconazole, prochlorperazine, dolutegravir, megestrol, and hydrochlorothiazide (alone or with triamterene). CYP3A4 inhibitors, including macrolides, azole antifungals, protease inhibitors, amiodarone, diltiazem, and grapefruit juice, are used with caution, and other QT-prolonging drugs are not recommended. Class I and class III antiarrhythmics are stopped for at least 3 half-lives before dofetilide; amiodarone is stopped for at least 3 months or until its plasma concentration is below 0.3 mcg/mL. Renally secreted drugs such as metformin, amiloride, and triamterene are combined with care. Initiation requires at least 3 days of continuous ECG monitoring in a facility, with QTc measured 2 to 3 hours after each dose.

Sotalol is not metabolized and is cleared by the kidney, so its interactions are pharmacodynamic: other QT-prolonging drugs, diuretic-induced hypokalemia and hypomagnesemia, and other drugs that slow the sinus rate or AV conduction. Antacids containing aluminum or magnesium lower its absorption if taken within 2 hours.

Flecainide and propafenone. Flecainide is metabolized by CYP2D6 and is about one-third renally cleared; strong CYP2D6 inhibitors and amiodarone raise its concentration, and amiodarone labeling advises reducing other antiarrhythmics by 30% to 50%. Propafenone is metabolized by CYP2D6, CYP3A4, and CYP1A2; its label advises avoiding simultaneous use of a CYP2D6 inhibitor and a CYP3A4 inhibitor, because inhibition of both pathways raises concentrations to proarrhythmic levels. Propafenone is itself a P-glycoprotein inhibitor that raises digoxin by 60% to 270% and a CYP2D6 inhibitor that raises metoprolol and propranolol exposure. Both drugs are avoided in structural heart disease and are combined with an AV nodal blocker in atrial fibrillation to reduce the risk of 1:1 conduction of atrial flutter.

Mexiletine is metabolized by CYP2D6 and CYP1A2; smoking induces CYP1A2 and lowers mexiletine concentrations, and ciprofloxacin and fluvoxamine (CYP1A2 inhibitors) raise them. Quinidine is a strong CYP2D6 inhibitor and a P-glycoprotein inhibitor; it raises digoxin and the exposure of CYP2D6 substrates. Disopyramide is a CYP3A4 substrate and adds anticholinergic effects (urinary retention, constipation, glaucoma) that are particularly troublesome in older adults. Procainamide is acetylated to N-acetylprocainamide, a class III metabolite that accumulates in kidney disease; cimetidine and trimethoprim reduce renal clearance of both compounds.

Interactions with transplant immunosuppressants

Cyclosporine, tacrolimus, sirolimus, and everolimus are narrow-therapeutic-index CYP3A4 and P-glycoprotein substrates monitored by trough concentrations. Diltiazem, verapamil, amiodarone, dronedarone, clarithromycin, erythromycin, and azole antifungals raise their concentrations; rifampin, carbamazepine, phenytoin, and St John's wort lower them and can precipitate rejection. Diltiazem has been used deliberately after transplantation to lower cyclosporine dose requirements, which illustrates the magnitude of the effect. Whenever one of these drugs is started, stopped, or changed in dose in a transplant recipient, immunosuppressant trough concentrations are rechecked within days.

Cyclosporine is also a potent inhibitor of OATP1B1, P-glycoprotein, BCRP, and CYP3A4. It is contraindicated with simvastatin and pitavastatin, avoided with lovastatin, atorvastatin, and fluvastatin, and limits rosuvastatin to 5 mg and pravastatin to 20 mg. It raises colchicine, digoxin, and dabigatran exposure. Tacrolimus does not share cyclosporine's OATP inhibition, and its effect on statin exposure is much smaller. Calcineurin inhibitors cause hyperkalemia and nephrotoxicity that add to renin-angiotensin system inhibitors, mineralocorticoid antagonists, trimethoprim, and NSAIDs.

Interactions with antiviral drugs: HIV, hepatitis C, and nirmatrelvir-ritonavir

HIV therapy. Ritonavir and cobicistat are used as pharmacokinetic boosters because they are strong CYP3A inhibitors; boosted protease inhibitor and boosted elvitegravir regimens therefore carry the full range of CYP3A interactions. These include contraindication with simvastatin and lovastatin, dose limits for atorvastatin and rosuvastatin, avoidance of rivaroxaban and ticagrelor, reduced apixaban doses, contraindication of ranolazine, ivabradine, eplerenone, dronedarone, and several antiarrhythmics, and higher concentrations of calcium channel blockers and PDE5 inhibitors. Ritonavir also induces CYP2C9 and glucuronidation, so the INR can fall with warfarin. Efavirenz and etravirine induce CYP3A and lower concentrations of CYP3A substrates. Unboosted integrase inhibitors (dolutegravir, bictegravir, raltegravir) have few CYP interactions, but dolutegravir and bictegravir inhibit OCT2 and MATE1, raising metformin and dofetilide concentrations and serum creatinine; dofetilide is contraindicated with both.

Hepatitis C therapy. Sofosbuvir-containing regimens with amiodarone have caused severe symptomatic bradycardia, including deaths and cases requiring pacing, typically within days to weeks of starting treatment; the mechanism is not fully defined, and the combination is not recommended. Glecaprevir-pibrentasvir, sofosbuvir-velpatasvir-voxilaprevir, and ledipasvir-sofosbuvir inhibit OATP1B1, BCRP, or P-glycoprotein and raise statin and dabigatran exposure; rosuvastatin labeling advises avoiding sofosbuvir-velpatasvir-voxilaprevir and ledipasvir-sofosbuvir.

Nirmatrelvir-ritonavir. This 5-day COVID-19 treatment produces strong CYP3A and P-glycoprotein inhibition that resolves within several days of the last dose. Its US prescribing information contraindicates amiodarone, dronedarone, flecainide, propafenone, quinidine, ranolazine, eplerenone, finerenone, ivabradine, lovastatin, simvastatin, and sildenafil for pulmonary arterial hypertension.

From the US prescribing information and manufacturer interaction tables for nirmatrelvir-ritonavir. The list is not exhaustive.

Anti-infective, psychotropic, anticonvulsant, and oncology drug interactions

Antibiotics. Clarithromycin is a strong and erythromycin a moderate CYP3A inhibitor; both inhibit P-glycoprotein and prolong the QT interval. Azithromycin does not inhibit CYP3A4 meaningfully but raises digoxin to an undefined degree and prolongs the QT interval. Fluoroquinolones prolong the QT interval and raise the INR with warfarin; ciprofloxacin inhibits CYP1A2. Trimethoprim raises potassium, raises creatinine without changing glomerular filtration rate (GFR), and is contraindicated with dofetilide; sulfamethoxazole inhibits CYP2C9. Metronidazole inhibits warfarin metabolism. Rifampin is the most potent clinically used inducer of CYP3A4, CYP2C9, CYP2C19, and P-glycoprotein. Linezolid is a weak monoamine oxidase inhibitor that can raise blood pressure with sympathomimetics.

Antifungals. Ketoconazole, itraconazole, posaconazole, and voriconazole are strong CYP3A inhibitors; fluconazole and isavuconazole are moderate, and fluconazole is also a moderate CYP2C9 and strong CYP2C19 inhibitor. Fluconazole's interaction with warfarin is large and appears within days, including with short courses. Itraconazole has negative inotropic effects and is avoided in heart failure. All azoles except isavuconazole prolong the QT interval; isavuconazole shortens it.

Psychotropic drugs. Fluoxetine and paroxetine are strong CYP2D6 inhibitors; fluvoxamine strongly inhibits CYP1A2 and CYP2C19. Citalopram and escitalopram prolong the QT interval in a dose-dependent way. SSRIs add bleeding risk to antithrombotic drugs. Tricyclic antidepressants, antipsychotics, and methadone prolong the QT interval, and tricyclics and antipsychotics cause orthostatic hypotension. Lithium interacts with diuretics, ACE inhibitors, ARBs, ARNIs, and NSAIDs. Nefazodone is a strong CYP3A inhibitor.

Anticonvulsants. Carbamazepine, phenytoin, phenobarbital, and primidone induce CYP3A4, CYP2C9, and P-glycoprotein and lower the effect of DOACs, ticagrelor, statins metabolized by CYP3A4, calcium channel blockers, dronedarone, and warfarin. Phenytoin is also a CYP2C9 substrate whose concentrations rise with amiodarone and fluconazole. Valproate, levetiracetam, and lamotrigine have fewer cardiovascular interactions.

Oncology drugs. Many tyrosine kinase inhibitors are CYP3A4 substrates, inhibitors, or transporter inhibitors, and several prolong the QT interval. Ibrutinib and other Bruton tyrosine kinase inhibitors increase bleeding with anticoagulants and antiplatelet drugs and cause atrial fibrillation. Enzalutamide and apalutamide are strong CYP3A inducers that reduce DOAC and statin exposure. Darolutamide and capmatinib raise rosuvastatin exposure.

Herbal products, supplements, food, alcohol, and tobacco

KDIGO 2024 Practice Point 4.1.3 advises review and limitation of over-the-counter medicines and dietary or herbal remedies in people with kidney disease, and the same review applies to every patient taking a narrow-index cardiovascular drug. Supplements are not regulated for content, and patients often do not report them unless asked.

St John's wort is a strong inducer of CYP3A4 and P-glycoprotein through the pregnane X receptor. It lowers concentrations of cyclosporine and tacrolimus (with rejection reported), DOACs, warfarin, digoxin, statins, ivabradine, ranolazine, and calcium channel blockers. Grapefruit juice and Seville orange juice inhibit intestinal CYP3A4, with the largest effects on simvastatin, lovastatin, felodipine, and nifedipine. Licorice containing glycyrrhizin inhibits 11-beta-hydroxysteroid dehydrogenase type 2 and causes hypertension, hypokalemia, and metabolic alkalosis, which add to diuretic-induced potassium loss and raise digoxin and QT risk. Potassium-enriched salt substitutes and potassium supplements add to potassium-raising drugs and are used cautiously in advanced kidney disease. Supplements marketed for antiplatelet or anticoagulant effects, including high-dose fish oil, ginkgo, garlic, and vitamin E, have inconsistent clinical evidence of added bleeding but are disclosed and reviewed before procedures and when anticoagulants are started.

Vitamin K and alcohol affect warfarin as described in the warfarin section. Alcohol also adds to the hypotensive effect of vasodilators and to the sedative effects of centrally acting drugs. Tobacco smoke induces CYP1A2 through polycyclic aromatic hydrocarbons (nicotine replacement does not), so smoking cessation raises concentrations of CYP1A2 substrates such as mexiletine, clozapine, olanzapine, and theophylline. Caffeine and other methylxanthines antagonize adenosine receptors and are withheld before vasodilator stress testing with adenosine, regadenoson, or dipyridamole.

Estimating kidney function for drug dosing

Most US drug labels written before 2010 based renal dose adjustment on creatinine clearance estimated by the Cockcroft-Gault equation, because that was the estimate used in the pharmacokinetic studies that defined the dose bands. The equation, published in 1976 from 249 hospitalized men, calculates creatinine clearance (mL/min) as (140 − age) × weight (kg) ÷ (72 × serum creatinine in mg/dL), multiplied by 0.85 in women. It predates standardized creatinine assays, is not indexed to body surface area, and incorporates weight, so it estimates absolute clearance in mL/min.

Laboratories report estimated glomerular filtration rate (eGFR) from the 2021 Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) creatinine equation, which removed the race coefficient, in mL/min/1.73 m². The 2021 CKD-EPI creatinine-cystatin C equation is more accurate than either marker alone. KDIGO 2024 addresses drug dosing in five practice points:

1. Practice Point 4.2.1: GFR is considered when dosing medications cleared by the kidneys.
2. Practice Point 4.2.2: for most people and clinical settings, validated eGFR equations using serum creatinine are appropriate for drug dosing.
3. Practice Point 4.2.3: where more accuracy is required, as for drugs with a narrow therapeutic or toxic range, in drug toxicity, or when creatinine-based estimates may be unreliable, equations combining creatinine and cystatin C, or measured GFR, may be indicated.
4. Practice Point 4.2.4: in people with extremes of body weight, eGFR not indexed for body surface area may be indicated, especially for drugs with a narrow therapeutic range or a minimum effective concentration. Nonindexed eGFR (mL/min) equals the indexed value multiplied by body surface area and divided by 1.73.
5. Practice Point 4.2.5: dosing is adapted when GFR, the non-GFR determinants of the filtration markers, or the volume of distribution are not in a steady state.

Two principles reconcile labeling with KDIGO. When a label specifies a renal estimate, as for dofetilide, sotalol, and the DOACs (Cockcroft-Gault creatinine clearance with actual body weight), that estimate defines the labeled dose. The 2023 ACC/AHA/American College of Chest Physicians (ACCP)/Heart Rhythm Society (HRS) atrial fibrillation guideline uses it for DOAC dosing. When the label is silent or uses eGFR, as for spironolactone, sacubitril-valsartan, and SGLT2 inhibitors, the laboratory eGFR is used. The two estimates diverge most in older adults and at extremes of body size. In older, lighter patients, Cockcroft-Gault gives a lower value than eGFR; substituting eGFR for creatinine clearance in DOAC dosing in this group leads to overdosing. In a cross-sectional study of 4,120 patients with atrial fibrillation in primary care, using the Modification of Diet in Renal Disease (MDRD) eGFR instead of Cockcroft-Gault creatinine clearance missed the indicated rivaroxaban dose reduction in 13.5% of patients. Among patients aged 80 years or older, 14.9% had a creatinine clearance below 30 mL/min but an eGFR of 30 mL/min/1.73 m² or more.

Body weight and muscle mass. In obesity, Cockcroft-Gault with actual weight overestimates clearance, and ideal or adjusted body weight (ideal plus 40% of the excess) is often used for drugs whose labels do not specify a weight. In cachexia or very low muscle mass, serum creatinine is low because creatinine production is low, and all creatinine-based estimates overestimate GFR; cystatin C, which does not depend on muscle mass, corrects this error. Rounding a low serum creatinine up to 1.0 mg/dL is a common local practice but is not supported by labeling or guidelines.

Non-steady-state kidney function. In acute kidney injury, serum creatinine lags behind the fall in GFR, and an estimate calculated from a rising creatinine overstates kidney function; the reverse occurs during recovery. Drugs with narrow therapeutic ranges are held or dosed conservatively until creatinine stabilizes. Drugs that inhibit tubular creatinine secretion, listed in the pharmacokinetic mechanisms section, lower creatinine clearance estimates without changing true GFR; cystatin C helps separate the two.

Kidney function calculator for cardiovascular drug dosing

The calculator returns Cockcroft-Gault creatinine clearance with actual, ideal, and adjusted body weight; the 2021 CKD-EPI creatinine eGFR, indexed and nonindexed; and the 2021 CKD-EPI creatinine-cystatin C eGFR when cystatin C is entered. Drug rows apply US labeled thresholds to Cockcroft-Gault clearance with actual body weight.

Cockcroft-Gault: (140 − age) × weight ÷ (72 × creatinine), × 0.85 in women. Ideal body weight (Devine): 50 kg (men) or 45.5 kg (women) + 2.3 kg per inch over 5 feet, with actual weight used when it is lower; adjusted weight = ideal + 0.4 × (actual − ideal), shown when actual weight exceeds ideal by more than 20%. CKD-EPI 2021 creatinine: 142 × min(Scr/κ, 1)^α × max(Scr/κ, 1)^−1.200 × 0.9938^age × 1.012 in women (κ 0.7 in women, 0.9 in men; α −0.241 in women, −0.302 in men). Body surface area by the Mosteller formula. The rows do not apply to dialysis, acute kidney injury, pregnancy, or indications other than those named.

Dose adjustment of cardiovascular drugs in chronic kidney disease

Renal dose adjustment follows the fraction of the drug and its active metabolites excreted unchanged by the kidney and the drug's therapeutic index. Drugs with high renal clearance and narrow therapeutic index (dofetilide, sotalol, digoxin, dabigatran, enoxaparin, fondaparinux) require adjustment by clearance band. Drugs cleared mainly by the liver (amiodarone, warfarin, most statins, metoprolol, carvedilol, most calcium channel blockers) need no renal adjustment, but kidney disease still changes their effect through altered protein binding, uremic platelet dysfunction, hyperkalemia, or susceptibility to myopathy. Adjustment is made either by reducing the dose with the same interval (which preserves steadier concentrations) or by extending the interval with the same dose (which preserves peak concentrations); the labels for dofetilide and sotalol use these two approaches respectively.

CrCl, Cockcroft-Gault creatinine clearance in mL/min; eGFR in mL/min/1.73 m². Doses are those in current US prescribing information for the named indication; other indications have different bands. Amiodarone, warfarin, metoprolol, carvedilol, amlodipine, atorvastatin, and ticagrelor have no labeled renal adjustment.

The 2023 Beers Criteria table of drugs to avoid or reduce in older adults with reduced kidney function includes amiloride, spironolactone, and triamterene (avoid at CrCl under 30), dabigatran (avoid under 30), dofetilide (reduce at 20 to 59), and edoxaban (reduce at 15 to 50; avoid under 15 or above 95). It also includes enoxaparin (reduce under 30), fondaparinux (avoid under 30), rivaroxaban (labeled dosing at 15 to 50; avoid under 15), and trimethoprim-sulfamethoxazole (reduce at 15 to 29; avoid under 15). Apixaban was removed from the table because of evidence of safe use at lower kidney function.

Dosing in dialysis

Removal of a drug by hemodialysis depends on molecular weight, protein binding, and volume of distribution. Small, water-soluble, poorly protein-bound drugs with a small volume of distribution (dabigatran, atenolol, sotalol, procainamide) are removed substantially; drugs with large volumes of distribution or high protein binding (digoxin, amiodarone, warfarin, apixaban, rivaroxaban, most statins, beta-blockers other than atenolol and sotalol) are not. Dialyzable drugs are given after a dialysis session. Digoxin toxicity in a dialysis patient is treated with digoxin immune Fab, because dialysis does not remove digoxin.

For stroke prevention in patients with atrial fibrillation who receive dialysis, US labeling provides an apixaban dose (5 mg twice daily, 2.5 mg if age 80 years or older or weight 60 kg or less) based on pharmacokinetic data. Rivaroxaban labeling allows 15 mg daily. Randomized evidence is limited. In a US registry analysis of patients receiving dialysis, apixaban was associated with less major bleeding than warfarin, and the standard 5-mg dose with lower rates of stroke and death than the reduced dose. The 2023 ACC/AHA/ACCP/HRS atrial fibrillation guideline states that warfarin or an evidence-based dose of apixaban might be reasonable in end-stage kidney disease or dialysis (Class 2b). Dabigatran and edoxaban are not used in dialysis. Enoxaparin is avoided in favor of unfractionated heparin for therapeutic anticoagulation. Dofetilide is contraindicated, and sotalol is used only with specialist dosing. Spironolactone, eplerenone, and potassium-sparing diuretics are generally avoided because of hyperkalemia between sessions.

Dosing in hepatic impairment

Liver disease alters drug handling through reduced CYP activity and through reduced hepatic blood flow and portosystemic shunting (which raise the oral bioavailability of high-extraction drugs). Hypoalbuminemia raises the free fraction of highly bound drugs, ascites enlarges the volume of distribution of water-soluble drugs, and coagulopathy adds to anticoagulant bleeding. No laboratory test predicts hepatic drug clearance the way creatinine clearance predicts renal clearance. Drug labels therefore use the Child-Pugh classification, the basis of FDA hepatic impairment studies.

Child-Pugh class A, 5 to 6 points (mild impairment); class B, 7 to 9 points (moderate); class C, 10 to 15 points (severe).

Hepatic impairment recommendations for cardiovascular drugs, from US prescribing information. Cardiac cirrhosis from long-standing right heart failure produces the same pharmacokinetic changes.

Congestive hepatopathy from right heart failure or Fontan circulation reduces hepatic blood flow and drug clearance, and decongestion can restore them. Warfarin requirements fall during acute heart failure and rise again with diuresis, and amiodarone hepatotoxicity is distinguished from congestion by the pattern of enzyme elevation and the response to decongestion.

Pharmacogenomics in cardiovascular prescribing

Inherited variation in drug-metabolizing enzymes and transporters changes drug exposure in the same direction as an interacting drug. The Clinical Pharmacogenetics Implementation Consortium (CPIC) publishes guidelines that translate a known genotype into a prescribing action; whether to test is a separate clinical decision. Four gene-drug pairs have the strongest cardiovascular evidence.

CYP2C19 and clopidogrel. Carriers of loss-of-function alleles (most commonly 2 and 3) form less active metabolite and have higher on-treatment platelet reactivity and more ischemic events after percutaneous coronary intervention. Intermediate metabolizers carry one loss-of-function allele and poor metabolizers two; poor metabolizer status is substantially more common in people of East Asian ancestry than of European ancestry. The 2022 CPIC update recommends that, after an acute coronary syndrome or percutaneous coronary intervention, intermediate and poor metabolizers avoid standard-dose clopidogrel if possible. Prasugrel or ticagrelor at standard dose is recommended if no contraindication exists (strong recommendation for both). The US label carries a boxed warning on diminished antiplatelet effect in patients with two loss-of-function alleles.

Randomized evidence is mixed. In TAILOR-PCI, genotype-guided therapy in 5,302 patients undergoing percutaneous coronary intervention was followed by a lower rate of the primary ischemic outcome among loss-of-function carriers (4.0% versus 5.9% at 12 months). The difference did not reach statistical significance (hazard ratio 0.66; 95% CI, 0.43 to 1.02). In POPular Genetics, a genotype-guided strategy after primary percutaneous coronary intervention (clopidogrel for noncarriers, ticagrelor or prasugrel for carriers) was noninferior to universal ticagrelor or prasugrel for net adverse clinical events (5.1% versus 5.9%). It also reduced bleeding (9.8% versus 12.5%). The 2024 AHA scientific statement concluded that the evidence supports CYP2C19 testing before oral P2Y12 inhibitors are prescribed in acute coronary syndromes or percutaneous coronary intervention, with implementation dependent on rapid results and electronic decision support. ACC/AHA practice guidelines have not adopted routine genotyping. A strong CYP2C19 inhibitor, including omeprazole, converts a normal metabolizer toward an intermediate metabolizer phenotype, a process termed phenoconversion.

CYP2C9, VKORC1, and warfarin. CYP2C9 2 and 3 reduce S-warfarin clearance, and the VKORC1 −1639G>A variant increases warfarin sensitivity; together with age, body size, and interacting drugs, they explain about half of the variation in maintenance dose. The warfarin label includes a dose range table by genotype, and the 2017 CPIC guideline recommends a validated pharmacogenetic algorithm when genotype is available, with additional CYP2C9 alleles considered in patients of African ancestry. In COAG, a genotype-guided algorithm did not improve time in therapeutic range over a clinical algorithm in the first 4 weeks (45.2% versus 45.4%) and performed worse in Black patients. In EU-PACT, genotype-guided dosing raised time in therapeutic range compared with standard fixed loading (67.4% versus 60.3%). With DOACs now preferred for most indications, warfarin pharmacogenetics is used mainly where warfarin is required.

SLCO1B1 and statins. The SLCO1B1 c.521T>C variant (rs4149056) reduces OATP1B1 function and raises systemic statin exposure, most for simvastatin acid. The 2022 CPIC statin guideline recommends an alternative statin or a simvastatin dose below 20 mg daily for decreased function, and an alternative statin for poor function. Atorvastatin is started at 40 mg or less (decreased function) or 20 mg or less (poor function), and rosuvastatin at 20 mg or less in poor function. The guideline also covers ABCG2 for rosuvastatin and CYP2C9 for fluvastatin, and it states that genotype is not a reason to withhold statin therapy. Genotype adds to interacting drugs: an OATP1B1 inhibitor in a poor-function patient compounds the exposure increase.

CYP2D6 and beta-blockers and antiarrhythmics. Poor metabolizers have several-fold higher metoprolol concentrations and greater heart-rate reduction at the same dose, and they accumulate flecainide and propafenone. No CPIC guideline addresses beta-blockers; the clinical response (heart rate and blood pressure) is a direct measure that guides titration. Strong CYP2D6 inhibitors such as fluoxetine and paroxetine phenoconvert normal metabolizers to poor metabolizers.

Pharmacogenomic gene-drug pairs in cardiovascular prescribing. PCI, percutaneous coronary intervention; ACS, acute coronary syndrome.

Pharmacokinetic and pharmacodynamic changes with aging

Kidney function declines with age in most people, and serum creatinine underestimates the decline because muscle mass falls in parallel; a normal creatinine in an 85-year-old woman of 50 kg corresponds to a Cockcroft-Gault clearance well under 50 mL/min. Body fat rises and total body water and lean mass fall, which raises the volume of distribution of lipophilic drugs (amiodarone, some benzodiazepines), prolonging their half-lives, and lowers that of water-soluble drugs (digoxin, atenolol, lithium), raising their peak concentrations. Hepatic blood flow and liver mass decline, reducing first-pass extraction of high-clearance drugs such as metoprolol, propranolol, and verapamil. Serum albumin falls in frailty and illness.

Pharmacodynamic sensitivity also changes. Baroreflex responsiveness declines, so vasodilators, diuretics, and alpha-blockers cause more orthostatic hypotension. Sinus and AV nodal conduction disease is common, so nodal blockers more often produce symptomatic bradycardia. Sensitivity to warfarin increases, and bleeding risk on any antithrombotic regimen rises with age. Thirst and sodium-handling responses are blunted, so diuretics more easily cause hyponatremia and volume depletion. Central nervous system sensitivity to anticholinergic, sedative, and opioid drugs increases. Older adults were underrepresented in many of the trials that established cardiovascular drug doses, and "start low, go slow" describes the resulting practice: lower initial doses and longer intervals between titrations, with the target dose unchanged where trials support it.

Polypharmacy, the Beers Criteria, and STOPP/START in older adults

Polypharmacy is most often defined as five or more regularly used medications, and the number of drugs is the strongest predictor of potential interactions. Polypharmacy in heart disease is often appropriate: a patient with heart failure with reduced ejection fraction, atrial fibrillation, and diabetes can need eight or more indicated drugs. The aim is appropriate polypharmacy, in which each drug has a current indication, a dose matched to organ function, and benefits that outweigh burden within the patient's goals and life expectancy.

The 2023 American Geriatrics Society (AGS) Beers Criteria list potentially inappropriate medications for adults 65 years and older outside hospice and palliative care. The cardiovascular entries are summarized below; the criteria prompt review and do not forbid use when a drug is clearly indicated.

Summarized from the 2023 AGS Beers Criteria, Tables 2 to 6. HFrEF, heart failure with reduced ejection fraction.

The STOPP/START criteria, version 3 (2023), developed by a European Delphi panel, contain 133 STOPP criteria for potentially inappropriate medications and 57 START criteria for potential prescribing omissions. They are organized by physiological system and complement the Beers Criteria by identifying undertreatment: START criteria include anticoagulation for atrial fibrillation, statins for established vascular disease, and guideline-directed therapy for heart failure with reduced ejection fraction. STOPP criteria include, for example, NSAIDs and nondihydropyridine calcium channel blockers in heart failure, and antiplatelet drugs combined with anticoagulants in stable coronary disease without a specific indication.

Falls, orthostatic hypotension, syncope, and anticholinergic burden

Orthostatic hypotension, defined as a fall in systolic pressure of 20 mm Hg or more or diastolic pressure of 10 mm Hg or more within 3 minutes of standing, is common in older adults and is worsened by diuretics, alpha-blockers, vasodilators, nitrates, tricyclic antidepressants, antipsychotics, and PDE5 inhibitors. Standing blood pressure is measured when antihypertensive therapy is started or intensified in older adults and when a patient reports dizziness or falls. Bradycardia from combined nodal blockers and hypoglycemia from insulin or sulfonylureas are other drug causes of falls and syncope. A fall in a patient taking an anticoagulant is not by itself a reason to stop anticoagulation, because in most patients with atrial fibrillation the risk of intracranial hemorrhage from falls is small relative to the stroke risk. Fall prevention and correction of contributing drugs are addressed first.

Anticholinergic drugs (first-generation antihistamines, tricyclic antidepressants, bladder antimuscarinics, some antipsychotics, and disopyramide among cardiovascular drugs) impair cognition and cause delirium, urinary retention, constipation, and tachycardia. Cumulative anticholinergic burden across several drugs is associated with cognitive decline, and the Beers Criteria list strongly anticholinergic drugs and advise against combinations of three or more central nervous system-active drugs.

Deprescribing cardiovascular drugs in older adults

Deprescribing is the planned, supervised reduction or withdrawal of a drug whose harms or burden now exceed its benefits. The ACC framework for older adults with cardiovascular disease describes a sequence of five steps. The first is a complete medication list, including over-the-counter drugs and supplements. The second identifies each drug's indication and the drugs that are inappropriate, duplicated, interacting, or without a current indication. The third weighs the time to benefit of preventive drugs against life expectancy, frailty, and the patient's goals. The fourth is a plan to stop or taper one drug at a time. The fifth is monitoring for withdrawal effects and recurrence of the treated condition. Prescribing cascades, in which a drug's adverse effect is treated with another drug (amlodipine edema treated with a diuretic, or ACE inhibitor cough treated with an antitussive), are identified and reversed.

Randomized trials support several deprescribing decisions. In ASPREE, aspirin 100 mg daily in adults aged 70 years and older (65 years and older in US minority groups) without cardiovascular disease did not reduce cardiovascular events and increased major hemorrhage (3.8% versus 2.8%). The 2019 ACC/AHA primary prevention guideline states that low-dose aspirin should not be given routinely for primary prevention after age 70 (Class 3: Harm), and the 2023 Beers Criteria advise against initiation and support deprescribing. In OPTIMISE, among patients aged 80 years and older with systolic pressure below 150 mm Hg on two or more antihypertensives, removal of one drug was noninferior to usual care for blood pressure control at 12 weeks. Two-thirds of patients remained off the drug. In a trial of patients with advanced, life-limiting illness and an estimated life expectancy of 1 month to 1 year, statin discontinuation did not significantly change 60-day mortality and was associated with better quality of life.

Deprescribing considerations in older adults with cardiovascular disease. Drugs with a strong current indication, such as guideline-directed therapy for heart failure with reduced ejection fraction or anticoagulation for atrial fibrillation, are usually continued; their doses are matched to kidney function.

Medication reconciliation, interaction screening, and adherence

Most interaction harm occurs at transitions: admission, discharge, a new prescription from another specialist, or an acute illness. KDIGO 2024 Practice Point 4.3.1 calls for a thorough medication review periodically and at transitions of care to assess adherence, continued indication, and interactions, and Practice Point 4.3.1.2 for collaboration with pharmacists. A complete reconciliation compares what the patient takes, including over-the-counter drugs, supplements, and drugs from other prescribers, with what each record lists, and documents why each discrepancy exists.

Electronic interaction checkers flag most clinically important pairs, but different databases disagree on severity, many alerts concern interactions of little consequence, and frequent alerts lead to override of important ones. Checkers also miss interactions that depend on kidney function, timing, or genotype. Pharmacist-led medication review in older adults and in heart failure identifies drug-related problems that automated screening misses. The patient is often the most reliable source for over-the-counter and supplement use, and asking specifically about NSAIDs, sleep aids, antihistamines, and herbal products identifies many interactions.

Adherence declines as regimen complexity rises. Once-daily dosing, fixed-dose combinations, synchronized refills, blister packaging, and removal of drugs without a current indication improve adherence. Unintentional duplication (two renin-angiotensin system inhibitors from different prescribers, or brand and generic forms of the same drug) and incomplete restarts after a hospital stay are common and preventable.

Sources of error in interaction and dose assessment

• Dosing a DOAC, dofetilide, or sotalol by laboratory eGFR in an older, light patient: eGFR exceeds Cockcroft-Gault clearance, and the labeled dose is exceeded.
• Calculating clearance from a creatinine that is still rising in acute kidney injury: the estimate overstates function, and renally cleared drugs accumulate.
• Interpreting a creatinine rise after trimethoprim, dronedarone, cobicistat, or dolutegravir as kidney injury: these drugs block tubular creatinine secretion without changing GFR.
• Assuming an interaction ends when the inhibitor is stopped: amiodarone's effects persist for months, induction offsets over about 2 weeks, and mechanism-based CYP3A inhibition takes days to reverse.
• Forgetting the reverse interaction: stopping rifampin, carbamazepine, or St John's wort raises concentrations of drugs whose doses were increased to overcome induction.
• Switching statins to avoid a cyclosporine or gemfibrozil interaction: OATP1B1 inhibition affects every statin, and dose caps still apply.
• Reading a digoxin concentration drawn within 6 hours of a dose, or ignoring potassium and magnesium when judging toxicity: the early concentration overstates tissue exposure, and low potassium or magnesium causes toxicity at any concentration.
• Treating a total drug concentration as the active concentration in hypoalbuminemia or uremia: the total concentration of a highly bound drug underestimates the free concentration.
• Relying on an interaction checker without regard to kidney function, genotype, timing, or the clinical consequence of the specific pair: checkers disagree on severity and miss interactions that depend on these factors.
• Stopping drugs during an acute illness without a documented plan to restart them: failure to restart is the most common error in sick-day guidance.

References

1. Wiggins BS, Saseen JJ, Page RL 2nd, et al. Recommendations for management of clinically significant drug-drug interactions with statins and select agents used in patients with cardiovascular disease: a scientific statement from the American Heart Association. Circulation. 2016;134:e468–e495. doi:10.1161/CIR.0000000000000456
2. 2023 American Geriatrics Society Beers Criteria Update Expert Panel. American Geriatrics Society 2023 updated AGS Beers Criteria for potentially inappropriate medication use in older adults. J Am Geriatr Soc. 2023;71:2052–2081. doi:10.1111/jgs.18372
3. Kidney Disease: Improving Global Outcomes (KDIGO) CKD Work Group. KDIGO 2024 clinical practice guideline for the evaluation and management of chronic kidney disease. Kidney Int. 2024;105(4S):S117–S314. doi:10.1016/j.kint.2023.10.018
4. US Food and Drug Administration. Drug development and drug interactions: table of substrates, inhibitors and inducers. Content current as of May 29, 2026. fda.gov
5. O'Mahony D, Cherubini A, Guiteras AR, et al. STOPP/START criteria for potentially inappropriate prescribing in older people: version 3. Eur Geriatr Med. 2023;14:625–632 (correction 2023;14:633). doi:10.1007/s41999-023-00777-y
6. Cockcroft DW, Gault MH. Prediction of creatinine clearance from serum creatinine. Nephron. 1976;16:31–41. doi:10.1159/000180580
7. Inker LA, Eneanya ND, Coresh J, et al. New creatinine- and cystatin C-based equations to estimate GFR without race. N Engl J Med. 2021;385:1737–1749. doi:10.1056/NEJMoa2102953
8. Tisdale JE, Chung MK, Campbell KB, et al. Drug-induced arrhythmias: a scientific statement from the American Heart Association. Circulation. 2020;142:e214–e233. doi:10.1161/CIR.0000000000000905
9. Page RL 2nd, O'Bryant CL, Cheng D, et al. Drugs that may cause or exacerbate heart failure: a scientific statement from the American Heart Association. Circulation. 2016;134:e32–e69. doi:10.1161/CIR.0000000000000426
10. Joglar JA, Chung MK, Armbruster AL, et al. 2023 ACC/AHA/ACCP/HRS guideline for the diagnosis and management of atrial fibrillation. Circulation. 2024;149:e1–e156 (corrections 2024;149:e167, e936, e1413). doi:10.1161/CIR.0000000000001193
11. Heidenreich PA, Bozkurt B, Aguilar D, et al. 2022 AHA/ACC/HFSA guideline for the management of heart failure. Circulation. 2022;145:e895–e1032 (corrections 2022;145:e1033; 2022;146:e185; 2023;147:e674). doi:10.1161/CIR.0000000000001063
12. Krishnaswami A, Steinman MA, Goyal P, et al. Deprescribing in older adults with cardiovascular disease. J Am Coll Cardiol. 2019;73:2584–2595. doi:10.1016/j.jacc.2019.03.467
13. Lee CR, Luzum JA, Sangkuhl K, et al. Clinical Pharmacogenetics Implementation Consortium guideline for CYP2C19 genotype and clopidogrel therapy: 2022 update. Clin Pharmacol Ther. 2022;112:959–967. doi:10.1002/cpt.2526
14. Cooper-DeHoff RM, Niemi M, Ramsey LB, et al. The Clinical Pharmacogenetics Implementation Consortium guideline for SLCO1B1, ABCG2, and CYP2C9 genotypes and statin-associated musculoskeletal symptoms. Clin Pharmacol Ther. 2022;111:1007–1021. doi:10.1002/cpt.2557
15. Johnson JA, Caudle KE, Gong L, et al. Clinical Pharmacogenetics Implementation Consortium (CPIC) guideline for pharmacogenetics-guided warfarin dosing: 2017 update. Clin Pharmacol Ther. 2017;102:397–404. doi:10.1002/cpt.668
16. Pereira NL, Cresci S, Angiolillo DJ, et al. CYP2C19 genetic testing for oral P2Y12 inhibitor therapy: a scientific statement from the American Heart Association. Circulation. 2024;150:e129–e150. doi:10.1161/CIR.0000000000001257
17. Bhatt DL, Cryer BL, Contant CF, et al. Clopidogrel with or without omeprazole in coronary artery disease (COGENT). N Engl J Med. 2010;363:1909–1917. doi:10.1056/NEJMoa1007964
18. Pereira NL, Farkouh ME, So D, et al. Effect of genotype-guided oral P2Y12 inhibitor selection vs conventional clopidogrel therapy on ischemic outcomes after percutaneous coronary intervention: the TAILOR-PCI randomized clinical trial. JAMA. 2020;324:761–771. doi:10.1001/jama.2020.12443
19. Claassens DMF, Vos GJA, Bergmeijer TO, et al. A genotype-guided strategy for oral P2Y12 inhibitors in primary PCI (POPular Genetics). N Engl J Med. 2019;381:1621–1631. doi:10.1056/NEJMoa1907096
20. Kimmel SE, French B, Kasner SE, et al. A pharmacogenetic versus a clinical algorithm for warfarin dosing (COAG). N Engl J Med. 2013;369:2283–2293. doi:10.1056/NEJMoa1310669
21. Pirmohamed M, Burnside G, Eriksson N, et al. A randomized trial of genotype-guided dosing of warfarin (EU-PACT). N Engl J Med. 2013;369:2294–2303. doi:10.1056/NEJMoa1311386
22. Juurlink DN, Mamdani MM, Lee DS, et al. Rates of hyperkalemia after publication of the Randomized Aldactone Evaluation Study. N Engl J Med. 2004;351:543–551. doi:10.1056/NEJMoa040135
23. Juurlink DN, Mamdani M, Kopp A, Laupacis A, Redelmeier DA. Drug-drug interactions among elderly patients hospitalized for drug toxicity. JAMA. 2003;289:1652–1658. doi:10.1001/jama.289.13.1652
24. Antoniou T, Gomes T, Juurlink DN, et al. Trimethoprim-sulfamethoxazole-induced hyperkalemia in patients receiving inhibitors of the renin-angiotensin system: a population-based study. Arch Intern Med. 2010;170:1045–1049. doi:10.1001/archinternmed.2010.142
25. Lapi F, Azoulay L, Yin H, Nessim SJ, Suissa S. Concurrent use of diuretics, angiotensin converting enzyme inhibitors, and angiotensin receptor blockers with non-steroidal anti-inflammatory drugs and risk of acute kidney injury: nested case-control study. BMJ. 2013;346:e8525. doi:10.1136/bmj.e8525
26. Gandhi S, Fleet JL, Bailey DG, et al. Calcium-channel blocker-clarithromycin drug interactions and acute kidney injury. JAMA. 2013;310:2544–2553. doi:10.1001/jama.2013.282426
27. Ray WA, Murray KT, Meredith S, et al. Oral erythromycin and the risk of sudden death from cardiac causes. N Engl J Med. 2004;351:1089–1096. doi:10.1056/NEJMoa040582
28. Chang SH, Chou IJ, Yeh YH, et al. Association between use of non-vitamin K oral anticoagulants with and without concurrent medications and risk of major bleeding in nonvalvular atrial fibrillation. JAMA. 2017;318:1250–1259. doi:10.1001/jama.2017.13883
29. Siontis KC, Zhang X, Eckard A, et al. Outcomes associated with apixaban use in patients with end-stage kidney disease and atrial fibrillation in the United States. Circulation. 2018;138:1519–1529. doi:10.1161/CIRCULATIONAHA.118.035418
30. Steffel J, Collins R, Antz M, et al. 2021 European Heart Rhythm Association practical guide on the use of non-vitamin K antagonist oral anticoagulants in patients with atrial fibrillation. Europace. 2021;23:1612–1676 (correction 2021;23:1676). doi:10.1093/europace/euab065
31. Kubica J, Adamski P, Ostrowska M, et al. Morphine delays and attenuates ticagrelor exposure and action in patients with myocardial infarction: the randomized, double-blind, placebo-controlled IMPRESSION trial. Eur Heart J. 2016;37:245–252. doi:10.1093/eurheartj/ehv547
32. Lilja JJ, Neuvonen M, Neuvonen PJ. Effects of regular consumption of grapefruit juice on the pharmacokinetics of simvastatin. Br J Clin Pharmacol. 2004;58:56–60. doi:10.1111/j.1365-2125.2004.02095.x
33. Hylek EM, Heiman H, Skates SJ, Sheehan MA, Singer DE. Acetaminophen and other risk factors for excessive warfarin anticoagulation. JAMA. 1998;279:657–662. doi:10.1001/jama.279.9.657
34. Rathore SS, Curtis JP, Wang Y, Bristow MR, Krumholz HM. Association of serum digoxin concentration and outcomes in patients with heart failure. JAMA. 2003;289:871–878. doi:10.1001/jama.289.7.871
35. Budnitz DS, Lovegrove MC, Shehab N, Richards CL. Emergency hospitalizations for adverse drug events in older Americans. N Engl J Med. 2011;365:2002–2012. doi:10.1056/NEJMsa1103053
36. Masnoon N, Shakib S, Kalisch-Ellett L, Caughey GE. What is polypharmacy? A systematic review of definitions. BMC Geriatr. 2017;17:230. doi:10.1186/s12877-017-0621-2
37. McNeil JJ, Wolfe R, Woods RL, et al. Effect of aspirin on cardiovascular events and bleeding in the healthy elderly (ASPREE). N Engl J Med. 2018;379:1509–1518. doi:10.1056/NEJMoa1805819
38. Arnett DK, Blumenthal RS, Albert MA, et al. 2019 ACC/AHA guideline on the primary prevention of cardiovascular disease. Circulation. 2019;140:e596–e646 (corrections 2019;140:e649–e650; 2020;141:e60, e774). doi:10.1161/CIR.0000000000000678
39. Sheppard JP, Burt J, Lown M, et al. Effect of antihypertensive medication reduction vs usual care on short-term blood pressure control in patients with hypertension aged 80 years and older: the OPTIMISE randomized clinical trial. JAMA. 2020;323:2039–2051. doi:10.1001/jama.2020.4871
40. Kutner JS, Blatchford PJ, Taylor DH Jr, et al. Safety and benefit of discontinuing statin therapy in the setting of advanced, life-limiting illness: a randomized clinical trial. JAMA Intern Med. 2015;175:691–700. doi:10.1001/jamainternmed.2015.0289
41. Amiodarone hydrochloride tablets (Pacerone) prescribing information, sections 7 and 12.3. Upsher-Smith Laboratories. DailyMed
42. Dronedarone (Multaq) prescribing information. Sanofi-Aventis US. DailyMed
43. Digoxin tablets (Lanoxin) prescribing information, sections 2, 7, and 12. Advanz Pharma. DailyMed
44. Dofetilide (Tikosyn) prescribing information. Pfizer. DailyMed
45. MacCallum PK, Mathur R, Hull SA, et al. Patient safety and estimation of renal function in patients prescribed new oral anticoagulants for stroke prevention in atrial fibrillation: a cross-sectional study. BMJ Open. 2013;3:e003343. doi:10.1136/bmjopen-2013-003343