Real-World Evidence of the Top 100 Prescribed Drugs in the USA and Their Potential for Drug Interactions with Nirmatrelvir; Ritonavir

Nirmatrelvir (coadministered with ritonavir as PAXLOVIDTM) reduces the risk of COVID-19-related hospitalizations and all-cause death in individuals with mild-to-moderate COVID-19 at high risk of progression to severe disease. Ritonavir is coadministered as a pharmacokinetic enhancer. However, ritonavir may cause drug-drug interactions (DDIs) due to its interactions with various drug-metabolizing enzymes and transporters, including cytochrome P450 (CYP) 3A, CYP2D6, and P-glycoprotein transporters. To better understand the extent of DDIs (or lack thereof) of nirmatrelvir; ritonavir in a clinical setting, this study used real-world evidence (RWE) from the Optum Clinformatics Data Mart database to identify the top 100 drugs most commonly prescribed to US patients at high risk of progression to severe COVID-19 disease. The top 100 drugs were identified based on total counts associated with drugs prescribed to high-risk patients (i.e.,  ≥ 1 medical condition associated with an increased risk of severe COVID-19) who were continuously enrolled in the database throughout 2019 and had  ≥ 1 prescription claim. Each of the 100 drugs was then assessed for DDI risk based on their metabolism, excretion, and transport pathways identified from available US prescribing and medical literature sources. Seventy drugs identified were not expected to have DDIs with nirmatrelvir; ritonavir, including many cardiovascular agents, anti-infectives, antidiabetic agents, and antidepressants. Conversely, 30 drugs, including corticosteroids, narcotic analgesics, anticoagulants, statins, and sedatives/hypnotics, were expected to cause DDIs with nirmatrelvir; ritonavir. This RWE analysis is complementary to the prescribing information and other DDI management tools for guiding healthcare providers in managing DDIs.


Introduction
Nirmatrelvir (coadministered with ritonavir [nirmatrelvir; ritonavir] as PAXLOVID TM ; Pfizer Inc, New York, NY, USA) is a potent and selective oral COVID-19 treatment that inhibits the main 3-chymotrypsin-like cysteine protease (3CL pro ; also known as M Pro ) of SARS-CoV-2 (1, 2).In the phase 2/3 EPIC-HR (Evaluation of Protease Inhibition for COVID-19 in High-Risk Patients) clinical trial in individuals with mild-to-moderate COVID-19 at high risk of progression to severe disease, nirmatrelvir; ritonavir treatment reduced the risk of progression to severe disease by 86% versus placebo when treatment was administered within 5 days of symptom onset (3).The US Food and Drug Administration (FDA) granted Emergency Use Authorization (EUA) for nirmatrelvir; ritonavir in the USA in adults and pediatric patients (12 years and older weighing ≥ 40 kg) with mildto-moderate COVID-19 at high risk of progression to severe 73 Page 2 of 23 disease (2).The standard dose presentation, nirmatrelvir 300 mg and ritonavir 100 mg, is administered twice daily for 5 days as soon as possible after diagnosis of COVID-19 and within 5 days of symptom onset (2) and is considered a preferred therapy in the USA for non-hospitalized adults by the COVID-19 Treatment Guidelines Panel (4).
In a clinical DDI study of nirmatrelvir; ritonavir as a victim drug of CYP3A4 inhibition (NCT04962022), nirmatrelvir mean area under the concentration versus time curve (AUC) and maximum observed concentration (C max ) increased by an additional 39% and 19%, respectively, when nirmatrelvir; ritonavir was coadministered with itraconazole, a potent CYP3A4 inhibitor, compared with nirmatrelvir; ritonavir administered alone (15).This suggests that additional CYP3A4 inhibitors do not substantially increase the degree of nirmatrelvir CYP3A4 inhibition beyond the effect of ritonavir when coadministered with nirmatrelvir; ritonavir.Conversely, in another clinical DDI study (NCT04962230), nirmatrelvir mean AUC and C max decreased by 55% and 43%, respectively, following coadministration with carbamazepine, a potent CYP3A4 inducer, compared with nirmatrelvir; ritonavir alone (15).Therefore, DDIs that cause induction of nirmatrelvir as a victim drug should also be considered when prescribing nirmatrelvir; ritonavir.
Ritonavir is commonly used as a pharmacokinetic (PK) enhancer in human immunodeficiency (HIV) and hepatitis C virus therapies (16).While infectious disease prescribers may have more experience managing DDIs with ritonavir, general practitioners who see patients with COVID-19 at risk of severe outcomes may not be as familiar with ritonavir DDIs, and they may need additional resources to manage them.Regulatory labels like the US PAXLOVID EUA Fact Sheet for Healthcare Providers (HCPs) remain the authoritative resource for drugs with potential interactions with nirmatrelvir; ritonavir (2).In addition, numerous nirmatrelvir; ritonavir DDI management resources developed by various institutions to manage DDIs are available, including the National Institutes of Health (NIH) DDI list (17), Ontario COVID-19 Science Advisory Table DDI resources (18), and University of Liverpool COVID-19 Drug Interactions Checker (19), as well as a number of articles that have highlighted both specific drug classes and commonly prescribed drugs with DDIs associated with nirmatrelvir; ritonavir (13,(20)(21)(22)(23)(24)(25)(26).
In this study, we used real-world evidence (RWE) to identify the 100 drugs most commonly prescribed in the USA to patients at high-risk of developing COVID-19 complications and assessed the potential for DDIs with nirmatrelvir; ritonavir.Our analysis aims to provide an understanding of drugs that are commonly encountered by HCPs and their potential for interaction with nirmatrelvir; ritonavir.As such, it differs from prior reviews on nirmatrelvir; ritonavir DDIs, which focus on identifying interactions from a mechanismbased perspective and primarily on drugs that interact with nirmatrelvir; ritonavir.The information presented herein on the top 100 drugs prescribed to the high-risk population (for which nirmatrelvir; ritonavir is indicated) identifies drugs that are likely, as well as unlikely, to interact with nirmatrelvir; ritonavir.This can help HCPs gain familiarity with potential DDIs and address questions from patients on common concomitant medications when prescribing nirmatrelvir; ritonavir.

Drug Data Source
This descriptive retrospective analysis used the Optum Clinformatics Data Mart (CDM) database to evaluate drug use among adults with confirmed COVID-19.The database includes deidentified patient-level data from an administrative health claims database for large commercial and Medicare Advantage health plans submitted for medical and pharmacy healthcare services, and includes information related to healthcare costs and resource utilization.The population includes individuals from all 50 states.

Top 100 Drug Selection
The initial data selection was performed using the full year of the 2019 CDM database.The 2019 analysis period was chosen to avoid capture of anomalous drug utilization patterns resulting from altered healthcare-seeking behaviors during the early COVID-19 pandemic period.The criteria for high-risk patients included ≥ 1 characteristic or underlying medical condition associated with an increased risk of developing severe illness from COVID-19 as defined by the US Centers for Disease Control and Prevention (CDC) (27).The condition must have appeared in the claim records during the 12 months prior for inclusion in the dataset.
Medical conditions attributed to high-risk patients based on CDC guidelines as of October 2021 included the following factors: ≥ 65 years of age (where age was assessed on the index date) and/or a history of cancer, chronic kidney disease, chronic lung diseases, chronic liver disease, dementia or other neurologic condition, diabetes (type 1 or type 2), Down syndrome, cardiovascular disease, HIV infection, hypertension, immune deficiencies/immunocompromised state, mental health conditions (i.e., mood disorders, including depression, schizophrenia, or spectrum disorder), obesity, pregnancy or recent pregnancy, sickle cell disease or thalassemia, smoking (current or former), solid organ or blood stem cell transplant, stroke or cerebrovascular disease, tuberculosis, or substance use disorders (27).The patient cohort was limited to those with confirmed continuous insurance enrollment from January 1, 2019, through December 31, 2019, and who had ≥ 1 prescription claim from a pharmacy or medical administration.
The top 100 drugs were then selected and ranked based on total patient counts associated with each drug.Drugs that appeared multiple times were reduced to their first occurrence.Vaccines, medical equipment (e.g., diabetes test strips, lancets, glucometers), X-ray contrast agents/dyes, and intravenous hydration fluids were removed to maximize relevance and clinical applicability of results.Of note, some drugs appear multiple times in the top 100 list due to coadministration with other drugs (lisinopril, amoxicillin, albuterol, losartan, hydrochlorothiazide, and oxycodone) or formulation differences (metoprolol extended versus immediate release).

Drug-Drug Interactions
A three-step process was used to assess DDI potential of nirmatrelvir; ritonavir with each of the top 100 drugs: (1) assessment of elimination and disposition pathways for interaction potential with nirmatrelvir; ritonavir, (2) assessment of the magnitude of change with ritonavir or other similar inhibitors, and (3) assessment of the clinical relevance of the interaction (if expected).The DDI assessment was made on a 1:1 evaluation of nirmatrelvir; ritonavir and the concomitant drug from the top 100 drug list.
First, US prescribing information provided by the FDA was used to understand the elimination and disposition pathways for each of the top 100 drugs.In cases where the originator (branded) prescribing information was not available, prescribing information for an approved generic (reference listed drug) was selected.If details on these pathways were not available in the US prescribing information, then the prescribing information was supplemented with information from DailyMed or scientific literature.Information on elimination and disposition pathways is typically provided in the Pharmacokinetics (Metabolism and Excretion) section within the drug's prescribing information.Data from in vitro metabolism studies and human absorption, distribution, metabolism, and excretion (ADME) studies is generally provided to assess the contribution of each elimination pathway to overall elimination.These sections were reviewed to assess if the drug was a substrate for CYP3A4, CYP2D6, or P-gp and thus potentially likely to interact with nirmatrelvir; ritonavir.It is possible that drugs can have multiple elimination pathways affected by ritonavir.DDI potential was further assessed by reviewing information in the Contraindications, Warning and Precautions, and/or Drug Interactions sections of the prescribing information, which also were used to identify if the drug was a strong CYP3A4 inducer.This mimics the process in which an HCP would access publicly available information for each drug.
Second, if there was no evidence that the elimination or disposition pathways were affected by ritonavir, then the drug was not expected to interact with nirmatrelvir; ritonavir and no further assessment was needed.Conversely, if an interaction with an elimination or disposition pathway was possible, the prescribing information was reviewed for available clinical DDI study data describing the magnitude of change in PK parameters with ritonavir or similarly acting inhibitors.As shown in Supplementary Table I, ritonavir-mediated interactions result in an increase in AUC of 1330% for midazolam (a CYP3A4 substrate), 26% for desipramine (a CYP2D6 substrate), and 95% for dabigatran (a P-gp substrate) (11,29).This served as a guide for determining the magnitude of a PK change (i.e., > 30%) likely to result in a clinically relevant DDI with nirmatrelvir; ritonavir.The clinical relevance was assumed to be less consequential for CYP2D6 and P-gp substrates relative to CYP3A4 substrates, especially if other elimination pathways were involved.Especially for older drugs, whose prescribing information is not routinely updated, the magnitude of change in PK parameters with an inhibitor may not be reported.For these drugs, information in the Warning and Precautions section indicative of a drug interaction as well as information in the Pharmacokinetics (Metabolism and Excretion) section indicative of predominant involvement of CYP3A4, CYP2D6 and/or P-gp in the disposition was considered sufficient to conclude that the drug is likely to interact with nirmatrelvir; ritonavir.Additionally, in such cases, contemporary literature was searched to confirm the magnitude of change in PK (i.e., > 30% change) when possible.
Third, for drugs identified as likely to interact with nirmatrelvir; ritonavir, the Contraindications, Warning and Precautions, and Drug Interactions sections of each drug's prescribing information were reviewed to identify recommendations for 73 Page 4 of 23 management of the interaction.DDI management approaches generally involved a contraindication, discontinuation, dose adjustment, or increased monitoring of adverse events.The clinical relevance of the DDI was noted based on what is reported in the prescribing information, as review of this information would be the general practice followed by HCPs.Information from this analysis for all the 100 drugs was compiled into a single table for ease of data presentation.The primary focus of this analysis was the effect of nirmatrelvir; ritonavir on the concomitant drug, and the effect of the concomitant drug on nirmatrelvir; ritonavir, if any, was also examined.

Results
Table I presents the top 100 drugs by prescription volume that are prescribed to patients who meet the criteria of an increased risk of progression to severe disease if infected with SARS-CoV-2.As expected, these drugs are generally prescribed to patients with COVID-19 with an increased risk of serious infection with common comorbidities, such as cardiovascular disease, diabetes, and mental health disorders (27).The top 5 represented drug classes include antibiotics/ antibacterial agents (15 drugs), antidepressants (8 drugs), corticosteroids (7 drugs), narcotic analgesics (6 drugs), and nonsteroidal anti-inflammatory drugs (NSAIDs; 5 drugs).
Of the top 100 drugs prescribed to the high-risk patient population, 70 drugs are not expected to have a DDI with nirmatrelvir; ritonavir based on their known mechanism of clearance and involvement of P-gp drug transport (Table I).Some examples among these drugs unlikely to interact with nirmatrelvir; ritonavir include lisinopril, azithromycin, levothyroxine, amoxicillin, and metformin, among many others.Notably, these drugs do not share clearance mechanisms that are impacted by ritonavir.In general, drugs that are unlikely to interact with nirmatrelvir; ritonavir are more likely to be predominantly eliminated unchanged, are cleared by metabolic enzymes other than CYP3A4 and/or CYP2D6, are not significant substrates of P-gp transporters, or have multiple clearance pathways that potentially compensate for the inhibitory effects of nirmatrelvir; ritonavir.For example, lisinopril, azithromycin, and metformin are excreted predominantly unmetabolized in urine or bile and thus are not likely to be impacted by the inhibitory effects of nirmatrelvir; ritonavir (30)(31)(32).Similarly, non-CYP3A4 substrates, such as metoprolol and lorazepam (primarily metabolized by CYP2D6 and glucuronidation, respectively), also do not represent a significant risk of interaction with nirmatrelvir; ritonavir.Albuterol (33), mupirocin (34), and other drugs that do not reach appreciable systemic exposure levels due to route of administration are also unlikely to interact with nirmatrelvir; ritonavir.
The remaining 30 drugs from this list of top 100 drugs are expected to interact with nirmatrelvir; ritonavir (Table I).The top most commonly represented drug classes among these include corticosteroids, narcotic analgesics, β-hydroxy β-methylglutaryl coenzyme A (HMG-CoA) reductase inhibitors (statins), and sedatives/hypnotics.Of the statins observed in this analysis, only simvastatin is contraindicated with nirmatrelvir; ritonavir, which is based on a significant 16-fold increase in the systemic exposure of simvastatin when given with potent CYP3A4 inhibitors, such as ritonavir, and significant risk of myopathy, including rhabdomyolysis (35,36).Based on rank order, the most common drug classes among the 30

Discussion
Regulatory labels, such as the US nirmatrelvir; ritonavir Fact Sheet for HCPs, remain the regulatory authority-approved resource for managing DDIs (2).The nirmatrelvir; ritonavir EUA Fact Sheet for HCPs has been revised to include many additional drugs anticipated to interact with nirmatrelvir; ritonavir that are not included in the ritonavir label but are likely to cause DDIs with nirmatrelvir; ritonavir (2,37).The emphasis of the nirmatrelvir; ritonavir EUA Fact Sheet for HCPs and other DDI management resources has largely been on drugs that are known or mechanistically likely to cause a DDI with nirmatrelvir; ritonavir that could cause adverse event(s) and thus warrant a contraindication or appropriate clinical management of the interaction.This work aimed to provide guidance on DDIs for drugs commonly prescribed to patients in the USA for pharmacologic maintenance of their underlying conditions and who may subsequently receive nirmatrelvir; ritonavir because of their high risk of developing severe COVID-19.
Of the 100 drugs most commonly prescribed to US patients at high risk of developing severe COVID-19, 70 drugs are not expected to have a clinically meaningful DDI with nirmatrelvir; ritonavir.This assessment was based on examination of the clearance mechanisms for these drugs and the potential for nirmatrelvir; ritonavir to meaningfully change (either increase or decrease) the systemic levels of each drug.HCPs should be aware of drugs that are not likely to cause an interaction, because these drugs are generally not covered in the prescribing information unless specifically examined in a DDI study.This knowledge would allow HCPs to focus on drugs that are likely to cause an interaction with nirmatrelvir; ritonavir and on the management of these DDIs.
In a similar approach of identifying drugs unlikely to cause interactions, the NIH guidance on DDIs with nirmatrelvir; ritonavir outlined 11 drug classes that are unlikely to have clinically relevant interactions with nirmatrelvir; ritonavir (4).This NIH list includes 34 drugs from the top 100 list of drugs identified in the current study that are categorized as unlikely to have DDIs with ritonavir; nirmatrelvir.Similarly, the University of Liverpool COVID-19 Drug Interactions Checker also identifies numerous drugs that are not likely to have DDIs with nirmatrelvir; ritonavir.None of these sources are meant to be an exhaustive list of drugs that are unlikely to be affected by coadministration of nirmatrelvir; ritonavir.Overall, these resources should help HCPs make an informed benefit versus risk decision on DDIs when prescribing and dispensing nirmatrelvir; ritonavir.
Results of this analysis suggest that 30 of the top 100 most commonly prescribed drugs among high-risk patients with COVID-19 in the USA have known DDIs or the potential for DDI with nirmatrelvir; ritonavir.The drug classes represented by the most drugs on the top 100 list with DDI potential are corticosteroids, narcotic analgesics, HMG Co-A reductase inhibitors (statins), and sedatives/hypnotics.Examples of each of these drug classes from the top 100 list are identified in the US Fact Sheet (version dated 09/2022) (2).
Coadministration of corticosteroids by any route of administration with strong CYP3A4 inhibitors, such as ritonavir, results in significant systemic level increases (2,38), which can potentially increase the risk of Cushing's syndrome and adrenal suppression if used for prolonged periods of time.Although there is a potential for a PK interaction based on the mechanism of metabolism for most of these corticosteroids, the risk of a clinically relevant interaction, such as Cushing's syndrome or adrenal suppression, associated with short-term dosing (i.e., 5 days) of nirmatrelvir; ritonavir is low (2).Notable among corticosteroids are beclomethasone, prednisone, and prednisolone, which are recommended in the nirmatrelvir; ritonavir EUA Fact Sheet for HCPs (version dated 09/2022) as alternatives to other corticosteroids, such as betamethasone, budesonide, ciclesonide, dexamethasone, fluticasone, methylprednisolone, mometasone, or triamcinolone (2).
All the narcotic analgesics (i.e., hydrocodone/acetaminophen [#10], tramadol [#24], oxycodone/acetaminophen [#45], codeine/acetaminophen [#62], fentanyl [#66], and oxycodone alone [#68]) commonly prescribed to the patient population at high risk of severe COVID-19 are expected to have a DDI with nirmatrelvir; ritonavir based on the current assessment.These drugs are at least partially metabolized by CYP3A4 and exhibit higher exposure with strong CYP3A inhibition (2,39).Increased narcotic analgesic exposure can place patients at risk of severe adverse events, including potentially fatal respiratory depression (2,40).Careful monitoring of therapeutic and adverse effects is warranted when coadministering with nirmatrelvir; ritonavir.The narcotic analgesic drug class is included in the nirmatrelvir; ritonavir EUA Fact Sheet for HCPs (version dated 09/2022), which notes the risk of overexposure and recommends careful monitoring and consideration of dose reductions (2).
The recommendations presented in this paper differ slightly from those in other DDI management resources, such as the NIH DDI list (17) or the University of Liverpool COVID-19 Drug Interactions Checker (19).For example, the University of Liverpool COVID-19 Drug Interactions Checker does not anticipate DDIs for prednisone, fluticasone, methylprednisolone, estradiol, and hydrocortisone, whereas the present assessment anticipates a risk of DDI with nirmatrelvir; ritonavir.Conversely, the University of Liverpool COVID-19 Drug Interactions Checker anticipates a potential weak interaction for losartan (including losartan/hydrochlorothiazide), cyclobenzaprine, fluoxetine, and venlafaxine and a potential interaction for hydroxyzine, whereas the present assessment does not anticipate a risk of DDI with nirmatrelvir; ritonavir based on known pathways for clearance of these concomitant medications.These recommendation differences can arise because evaluation of the clinical consequences of DDIs can be dependent on the perceived benefit versus risk assessment for patients.Assessment of DDIs with nirmatrelvir; ritonavir is complex, as the assessments are largely based on historical data with ritonavir as a booster with other protease inhibitors.Although data from dedicated DDI studies can help inform DDI assessment with other drugs, most information comes from existing product labels.Therefore, the interpretation of these data and implications for dosing with nirmatrelvir; ritonavir may slightly differ.
The current analysis has some limitations.More than 1,900 drugs are substrates of CYP3A4, and hundreds are substrates of P-gp (45,46).This analysis focused on the top 100 drugs most likely to be coadministered with nirmatrelvir; ritonavir to patients at high risk of severe COVID-19.This analysis leveraged the Optum CDM database, which includes prescription drug utilization.Thus, supplements and over-the-counter drugs, such as Saint John's Wort (Hypericum perforatum), were not captured.The top 100 drugs were queried from a US-based RWE database, and the results may not be fully translatable to other countries.There are likely to be many other drugs that are prescribed in the high-risk COVID-19 patient population beyond these top 100 drugs, some of which may interact with nirmatrelvir; ritonavir.The DDI assessment was made using a 1:1 evaluation of nirmatrelvir; ritonavir and the concomitant drug from the top 100 drug list.It did not consider polypharmacy-related drug interaction (i.e., multiple drugs interacting with each other, regardless of being on this list), which may potentially occur in this patient population.

Conclusions
In conclusion, this RWE analysis found that the majority (70 out of the 100 drugs) of the most prescribed drugs to patients at high risk of severe COVID-19 are not likely to have a DDI with nirmatrelvir; ritonavir based on assessment of their clearance and transport pathways.This analysis serves as a complementary DDI management educational resource, while the PAXLOVID EUA Fact Sheet for HCPs remains the regulatory authority-approved source of DDI information in the USA.No new DDIs were identified as a result of this analysis, and the benefit-risk profile of nirmatrelvir; ritonavir remains unchanged from that noted in the Fact Sheet (version dated 09/2022).A thorough review by HCPs of patients' concomitant drug(s) is essential for managing DDIs to ensure that the right patients receive nirmatrelvir; ritonavir, and that patients are not denied treatment when DDIs can be appropriately managed for nirmatrelvir; ritonavir.

Table I
Top 100 Drugs Prescribed Among US Patients with COVID-19 at High Risk of Progressing to Severe Disease: Potential for a DDI with Nirmatrelvir; Ritonavir a nophen Narcotic analgesic Hydrocodone is primarily metabolized by CYP3A4.Hydrocodone is also metabolized to its primary active metabolite, hydromorphone, via CYP2D6.The hydrocodone label includes a Black Box warning for CYP3A4 inhibitors, stating that coadministration with CYP3A4 inhibitors could prolong adverse reactions and may cause potentially fatal respiratory depression.Coadministration of a single 20 mg dose of hydrocodone with ketoconazole (200 mg BID for 6 days) increased AUC and C max by 135% and 78%, respectively.Acetaminophen is primarily metabolized in the liver by conjugation (to a glucuronide or sulfate) and CYP-dependent oxidation (mainly CYP2E1, with CYP1A2 and CYP3A4 as additional pathways).a Some drugs appear multiple times in the top 100 table due to coadministration with other drugs (lisinopril, amoxicillin, albuterol, losartan, hydrochlorothiazide, and oxycodone) or formulation differences (metoprolol extended versus immediate release).A full list of citations for Table I is provided in Supplementary Table II.