|Home | About | Journals | Submit | Contact Us | Français|
Combination antiretroviral therapy (cART) has improved survival of HIV-infected patients, but patients now experience co-morbidities that require pharmacological intervention, increasing the risk of drug-drug interactions. HIV protease inhibitors (PIs), non-nucleoside reverse transcriptase inhibitors (NNRTIs), and the CCR5 antagonist maraviroc are primarily metabolized via the cytochrome P450 (CYP) system and are prone to pharmacokinetic interactions.(1, 2) This article will address some key challenges prescribers face when using available drug interaction resources in making day-to-day clinical decisions.
Drug-drug interactions are considered clinically significant if they result in alteration of the desired outcomes leading to reduced clinical effectiveness or increase in the risk and/or magnitude of toxicities. Drugs are most vulnerable to the consequences of significant pharmacokinetic interactions when they are substrates of major metabolic or drug transport pathways and also have a narrow therapeutic window. Many cART regimens contain an NNRTI, a PI (usually boosted with ritonavir, a strong CYP3A4/5 inhibitor), or both, together with two nucleoside reverse transcriptase inhibitors (NRTIs). As HIV patients live longer, polypharmacy is unavoidable. Many patients may require additional medications to manage antiretroviral-associated adverse effects (such as hyperlipidemia) or other co-morbidities (such as cardiovascular diseases, erectile dysfunction, and depression). There are also patients who enter into care with advanced HIV, requiring treatment for AIDS-defining conditions (such as tuberculosis, invasive fungal infections, or malignancies). Anticipation and recognition of potential pharmacokinetic interactions should start when one intends to add, switch, or discontinue one or more medication(s).
The approval of the PIs in the mid 1990s has dramatically increased the awareness of pharmacokinetic-associated drug-drug interactions and their potential impact on drug therapy. Pharmaceutical companies and HIV experts increased their efforts in raising awareness and educating prescribers on the significance of interactions between PIs, in particular ritonavir, and other drugs metabolized via CYP3A4/5. However, CYP450-mediated interactions are not the sole mechanism by which drug interactions occur. An interaction that received much attention is the effect of acid suppression by proton pump inhibitors or histamine-2 (H2) antagonists on atazanavir absorption. This led to various pharmacokinetic studies that had different study designs and yielded some conflicting results.(3–5) Uridine diphosphate glucuronosyltransferase (UGT)1A1 is an enzyme responsible for the conjugation of many pharmaceutical products, but understanding of the impact of UGT1A1-mediated interactions is still in its infancy.(6) Raltegravir, a UGT1A1 substrate, has an advantage over PIs and NNRTIs because of the lack of CYP-associated interactions. When co-administered with rifampin, a potent UGT1A1 inducer, however, the pre-dose raltegravir concentration is reduced by 60%, necessitating a doubling of the raltegravir dose.(7) Other unexpected drug interactions continue to emerge, recognized only when studies or case reports identify such problems. Noted examples include: 1) reports of some serious and even fatal didanosine-associated toxicities(8, 9) when co-administered with tenofovir, as a result of increased didanosine concentrations (10); and 2) ritonavir-boosted PIs causing higher than expected systemic steroid exposure leading to adrenal insufficiency or Cushing’s Syndrome after coadministration with inhaled (11) or locally administered corticosteroids (12). There are yet other drugs used in HIV patients, including some herbal and complementary products where the ingredients and metabolic pathways are not well described; interactions with these agents cannot be easily predicted or anticipated.(13)
Although most HIV practitioners are aware of potential drug-drug interactions, many find themselves lacking the knowledge and tools needed to recognize and manage such interactions with confidence. To this end, detecting and managing drug interactions should not be the sole responsibility of prescribers. Pharmacists, patients, and other healthcare providers should also be aware of potential drug interactions, and be vigilant when drugs are added or removed from a regimen. Prescribers and pharmacists should be aware of key drugs or drug classes with the highest potential for significant interactions, such as PIs, NNRTIs, rifamycins, macrolides, azoles, anticonvulsants, psychotropic agents, etc., and consult appropriate resources for guidance, but note that there are limitations to all the current available references, as discussed below.
As it is generally considered unethical to commit HIV patients to antiretroviral monotherapy even for a short duration, most pharmacokinetic drug interaction studies are conducted in healthy volunteers using stringent dosing schedules, which allow for accurate assessment of the impact of one drug on another. Extrapolation of the data from these studies to generate recommendations for HIV-infected patients does have some limitations. Altered CYP metabolic activities (14) and reduced drug exposure have been reported with some drugs in HIV-infected patients when compared to HIV (−) volunteers (15, 16). The small number of participants typically studied in pharmacokinetic trials, wide inter-participant variability, and short study duration are other limitations. Moreover, healthy volunteer data may not be generalizable to populations with altered pharmacokinetics such as pregnant women, children, patients with renal or hepatic impairment, or patients with specific genetic polymorphisms (17).
Product prescribing information is a readily available source for clinicians to use. Results from pharmacokinetic studies conducted by the manufacturers are presented in table format. Significant interactions requiring prescribers’ attention are explicitly outlined. As listed in Table 1, some drugs are listed as “contraindicated” when concomitant use with an antiretroviral agent may result in significant harm; such combinations should be avoided. In less serious cases, a warning may be added to suggest that a drug combination should not be co-administered, unless no alternative drug is available and the therapeutic benefit outweighs the risk. In recent years, the manufacturers of antiretroviral drugs have been more diligent in conducting healthy volunteer pharmacokinetic studies, particularly with drugs that are frequently prescribed together for HIV-infected patients and where interactions may occur. These efforts have been successful in alerting clinicians to avoid certain drugs (for example, concomitant use of lovastatin or simvastatin and PIs) or to use them with dosage adjustment (for example, sildenafil and PIs). In addition, the results from some studies have led to more sophisticated but sometimes complicated dosage recommendations, particularly for newer agents such as atazanavir, maraviroc, and etravirine. As outlined in Table 2, although thorough, some dosing guidelines can be confusing and may result in both prescribing and administration errors due to misinterpretation of the intended instructions. To illustrate, when atazanavir is used in combination with acid suppressive agents, the atazanavir dose and administration time depends on all of the following factors (1) whether the patient is “treatment-naïve” or “treatment-experienced”; (2) whether or not co-administered with ritonavir, efavirenz, and/or tenofovir; (3) dose of PPI (omeprazole equivalent) or H2-antagonist (famotidine equivalent) used; and (4) timing of the PPI or H2-antagonist.(18) These recommendations are well intended and based on the design of time-limited pharmacokinetic studies; however, it is not practical and almost impossible to counsel patients and expect them to apply this rigid dosing schedule daily for the duration of the combined therapy.
Another challenge arises when the manufacturer of one drug performs a pharmacokinetic study whereby dosage recommendations are made in one drug label but not revised in the label(s) of the interacting drug(s). Recent examples include the interaction between voriconazole with ritonavir and voriconazole with efavirenz. Where the voriconazole label provided warning and recommendations for concomitant use of these combinations, this information did not appear in the antiretroviral drug labels until a couple of years later. The lack of coordination between the manufacturers of these products makes it difficult for clinicians to track interactions and prescribe correct dosages accordingly.
The awareness of the importance of antiretroviral-associated DDI and the wider usage of the internet have led to emergence of various electronic databases for clinicians to assess clinically significant drug interactions (Table 3). Most of these sites provide free access to the public and some can be downloaded into personal digital assistance (PDA) devices. The information and recommendations are based on data compiled from drug product labels, pharmacokinetic studies published or presented at major conferences, and key case reports. These databases are updated periodically when new data become available. Some sites provide interactive tools, where clinicians can enter the drugs in a patient’s medication profile, and a list of interactions will be displayed. They are an excellent starting point when new drugs are prescribed. However, the pharmacokinetic data only represent interactions between two drugs, and the extent of changes in pharmacokinetic parameters in the presence of three or more drugs is complex and almost impossible to predict.
Medication errors due to prescribing antiretroviral agents with contraindicated drugs or failure to adjust doses as recommended when prescribing interacting drugs are not uncommon.(19, 20) Errors may also occur when patients transfer from one healthcare system to another, such as admission to a hospital.(21, 22) Many institutions and pharmacies have installed drug interaction detection softwares within the electronic order entry and pharmacy dispensing programs, which provide an alert when drug interactions for the same patient are identified.(2, 23, 24) Institutions have the option to turn on or off the alarm system based on the level of severity of the interactions, in order to avoid alarm overload. Such programs serve as a safety net to avoid inadvertent prescribing errors, but studies have shown that prescribers and pharmacists often elect to override the drug interaction warnings without making a true assessment of the interactions due to heavy workload(25) or “alert fatigue”(23).
Expansion of antiretroviral roll-out programs has improved survival for patients in many resource-limited countries. However, drug interactions are becoming a tremendous challenge in those countries with a high burden of tuberculosis where rifampin remains a key component of treatment.(26) Even though efavirenz, given at usual or slightly increased doses in the presence of rifampin appears to result in reasonable virologic responses,(27) the high proportion of women of childbearing age and the increasing reports of NNRTI-failure and resistance necessitate the use of PI-based regimens which have greater potential for clinically significant interactions. The restricted use of fixed-dose combination therapy for both tuberculosis and HIV infections also restrict the ability to individualize therapy when interacting drugs are prescribed. Aside from tuberculosis, many patients in these regions continue to consume traditional herbal medicines, most of which have not been studied, making evaluation of potential drug interactions impossible.(28) HIV clinicians in resource-limited countries see high volume of patients, have little reference resources, no computer system, and less monitoring tools, making it very difficult to avoid, detect, and manage drug interactions.
While most HIV-infected patients in care have benefitted from effective cART and are living longer, many receive additional medications to manage or prevent comorbidities. Drug interactions are common; some can be anticipated and prevented, some cannot be avoided and others cannot be easily recognized or managed appropriately.(2) A recent survey noted that only 36% of clinically significant drug interaction were identified by physicians in an HIV clinic.(29) The first step to avoiding unwanted interaction is prescriber-initiated assessment of drug interaction potential. Prescribers and pharmacists should always review patients’ medication profiles when starting or changing regimens to determine the best therapy with the least interactions and/or toxicities. Patients should be encouraged to disclose all medications they use (including over-the-counter and herbal preparations), especially if they receive care from multiple providers. If possible, providers, especially consultants, should communicate with the primary care team before prescribing new medication(s) for a patient. Although there are some limitations to all available resources, drug interaction tools, including product labels, should be consulted and order entry alert systems should be utilized to identify interacting drugs. Unfortunately, not all drug combinations with similar metabolic pathway and have potential for clinically significant interactions have been formally evaluated, in those case, the clinicians have to use their best judgment and prescribe these drugs with caution. Prescribing of contraindicated drugs should be avoided. When more than two interacting drugs need to be used concurrently, clinicians should be aware of the limitations when using available resources in predicting the true extent of the interactions in these situations. If available, a clinical pharmacologist/pharmacist should be consulted. In selected cases, therapeutic drug monitoring maybe helpful in assessing the adequacy of drug concentrations and the need for dosage adjustment. After prescribing interacting medications, close follow-up is needed to assess therapeutic efficacy and potential toxicities. When unexplained symptoms or responses are seen after initiating or switching therapy, and a drug interaction is the suspected cause, one should consider reporting this to the manufacturer and to the medical community. In doing so, one adds to the drug interaction knowledge base and assists clinicians in safely prescribing antiretroviral drug combinations.
We thank Dr. Scott Penzak for his critical review of this manuscript. This project is supported by the Division of Clinical Research, National Institute of Allergy and Infectious Diseases, US National Institutes of Health. It is also funded in part with federal funds from the National Cancer Institute, National Institutes of Health, under Contract No. HHSN261200800001E. The content of this publication does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products, or organizations imply endorsement by the U.S. Government.
The authors have no financial conflict of interest.
Alice K. Pau, Division of Clinical Research (DCR), National Institute of Allergy and Infectious Diseases (NIAID), National Institutes of Health (NIH)
Sarita D. Boyd, SAIC-Frederick, Inc., NCI-Frederick, Frederick, MD, in support of DCR, NIAID, NIH.