Therapeutic Implications of Nonadherence
Antiretroviral Drug Exposure
Compliance with Antiretroviral Regimens
Antiretroviral Exposure Versus Response Relationships
Prescribing with Adherence in Mind
Compared with single-drug regimens, combination regimens of 2 or more antiretroviral drugs provide better virologic and immunologic benefit and improve clinical outcome in individuals infected with HIV. Several combination regimens with superior therapeutic efficacy have been recommended for beginning drug therapy in individuals with established HIV infection.1,2 As with drug regimens for any condition, however, success depends on both the prescriber's understanding of pharmacokinetics and pharmacodynamics and the patient's ability to take the drugs as prescribed.
The optimal dosing regimen for each drug should maintain therapeutic efficacy (as demonstrated by suppression of viral burden) and at the same time minimize the risk of side effects. Pharmacokinetic processes, including absorption of drug into the systemic circulation, distribution among tissues, and elimination by metabolic or excretory pathways, influence access of the drug to target tissues. Since these processes vary among individuals, plasma drug concentrations and subsequent exposure of target tissues differ among individuals given the same dose. Hence, for many drugs, therapeutic efficacy correlates better with systemic drug exposure than with dose.3
A summary indicator of systemic drug exposure, such as the area under the plasma concentration-time curve (AUC), is therefore a useful measure of drug delivery to target tissues where the drug elicits its effect(s). Adequate systemic drug exposure is required to maintain therapeutic efficacy and achieve a favorable clinical outcome.
Systemic exposure to the drug may be affected by conditions that limit the availability of drug to the bloodstream and ultimately to target tissues. Many factors, including genetic variability, age, race, gender, renal or hepatic disease, and interactions with concomitant drug therapy, influence systemic exposure to drugs and contribute to pharmacokinetic variability among individuals. A less tangible but nonetheless important factor that affects systemic exposure to drug therapy is failure to adhere to prescribed antiretroviral drug regimens.
Monitoring antiretroviral drug-taking behavior has revealed substantial deviations in the amount and timing of doses taken relative to prescribed dosing regimens.4 In addition, and notwithstanding the complexity of antiretroviral drug regimens, published data suggest that HIV-infected individuals demonstrate adherence patterns similar to those observed in a wide spectrum of other ambulatory diseases.4
Commonly, patterns of underdosing are observed (Figure 1). For example, the daily dosing frequency may be reduced such that a drug prescribed twice daily is actually taken as a once-a-day regimen. As a result, exposure to the drug over a certain period will be less than expected if the drug had been taken as prescribed.
|Figure 1. These graphs illustrate patterns of drug taking by 6 individuals receiving a 3-times-daily zidovudine regimen in ACTG 175. Three doses per day (y axis) are expected to be taken for each day of the monitoring period (x axis) between 2 consecutive clinic visits. Unlike protease inhibitors, nucleoside analogues such as zidovudine, didanosine, and stavudine can be "forgiving" of missed doses. Even missing half the doses, for example, may decrease the average therapeutic response by only about 20%.|
Even when patients comply fully with the number of prescribed doses per day, dosing intervals are likely to vary considerably. Because of the short half-lives of most antiretroviral drugs, such variation can result in low plasma concentrations before the next dose.5 A common phenomenon, even among individuals considered to be "good" adherers, is to observe periods of 2 or more consecutively missed doses (or "drug holidays").
The duration and frequency of drug holidays are critical since there appears to be a relationship between the time course of drug exposure and the dynamics of viral replication and, potentially, even with the emergence of resistant virus. In one study involving a small number of patients, drug holidays as short as 3 days during saquinavir monotherapy were associated with a rapid rise in HIV RNA copy number and possibly with the emergence of mutations resistant to saquinavir.6
A basic principle of pharmacotherapeutics is that therapeutic efficacy is a function of systemic drug exposure (Figure 2), which, in turn, is a function of each individual's pharmacokinetics and drug-taking behavior. Data from several studies suggest that the relationship between antiretroviral drug exposure and surrogate markers of antiviral response may be described using a saturable exposure-versus-response relationship (a so-called Emax model).7-9 In this model, when little or no drug is present, no response is observed. Once a minimally effective plasma drug concentration is achieved, there is an approximately linear increase in therapeutic response with increasing systemic drug exposure. At the plateau of the exposure-versus-response curve, the maximum therapeutic response is achieved, and any further increase in systemic drug exposure produces no additional response.
|Figure 2. Relationships among systemic drug exposure, therapeutic response (blue curve), and development of resistance (green curve) are illustrated here. If the maintenance dose of a drug results in systemic drug exposures in the flat part of the exposure-versus-response curve (A), a maximum therapeutic response is achieved, and emergence of viral resistance is minimized. Missing half the doses (B) results in an approximate 20% reduction in therapeutic response and a small risk of resistant virus emerging. If the maintenance dose of the drug produces systemic drug exposures at B on the curve, a 50% reduction in adherence results in both a significant reduction in therapeutic response and a substantial risk of emergence of resistant virus (C).|
In general, if there is enough leeway between the plasma concentration that produces acceptable therapeutic efficacy and the concentration associated with toxicity, dosing regimens are designed to ensure that a maximum response (Emax = 100%) is attained with the systemic exposure achieved at the usual maintenance dose. Thus, treated patients who adhere to the prescribed regimens of these drugs will attain systemic drug exposure in the flat part of the exposure-versus-response curve (A in Figure 2). For these drugs, even missing half of the doses (resulting in an average systemic exposure equivalent to B in Figure 2) decreases the average observed response by only approximately 20%.
Currently accepted standard dosing regimens for several nucleoside analogues that inhibit viral reverse transcriptase appear to fall in this category. In AIDS Clinical Trials Group (ACTG) protocol 019, a constant increase in CD4 T-cell count was observed over the range of systemic exposures to oral zidovudine (ZDV) after daily doses of 500 mg and 1500 mg in asymptomatic patients.10 Similarly, Collier and coworkers11 found no difference in either immunologic (increased CD4 count) or virologic (decreased p24 antigen) markers of disease progression resulting from systemic zidovudine exposure after 300-, 600-, and 1500-mg daily doses. In addition, results of dose-escalation studies for didanosine (ddI)12,13 and stavudine (d4T)14 suggest clinically tolerated doses of these drugs are also on the plateau of the dose-versus-response relationship. Hence, missed doses may have only a minor influence on therapeutic outcome.
An important consideration, however, is that the dideoxynucleosides are prodrugs that are anabolized intracellularly to the respective active triphosphate moiety. Compared with dose and with plasma concentration, intracellular accumulation of the active moiety is a more sensitive measure of drug exposure and correlates of response,15 although, currently, limited availability of sensitive analytic methods has hindered the ability to study intracellular kinetics of these drugs.
The HIV protease inhibitor drugs are clinically more efficacious than the dideoxynucleoside drugs. In pharmacologic terms, that means they have a greater absolute Emax. However, therapeutic effects elicited by standard, clinically tolerated doses of several of the protease inhibitors appear to fall within the graded region of the exposure-versus-response curve (Figure 2).9,16 In general, for a drug that produces a submaximal response (eg, 80% Emax, B in Figure 2), missing 50% of the prescribed doses results in a substantial reduction in response, such that the observed response may be reduced by approximately 75% (C in Figure 2, where the observed response is approximately 20% Emax).
In a safety and efficacy study of 2 dosages of saquinavir (3600 mg or 7200 mg daily), the reduction in viral load correlated with systemic exposure.9 These data indicate that inadequate antiretroviral exposure as a result of failure to adhere to prescribed dosing regimens of protease inhibitors is associated with an increased risk of therapeutic failure. Although this was a study of saquinavir monotherapy, the principle may similarly apply to the components of combination-drug regimens.
Another important cause of therapeutic failure with all antiretroviral drugs is the emergence of resistant quasispecies of the virus. Optimal dosing regimens provide systemic drug exposures that suppress viral replication and minimize the probability that resistance mutations will emerge. In clinical practice, drug-resistant organisms seem more likely to emerge when drug exposure is suboptimal (Figure 2).17 Presumably, individual variations in pharmacokinetics and adherence behavior can combine to produce drug concentrations in the graded region of the exposure-versus-response curve. Concentrations in that region permit viral replication (and thus mutation) to continue and place selection pressure on the virus favoring replication of resistant strains that may later become the predominant form of HIV in the body. Thus, individuals with frequent drug holidays or partial adherence that leads to exposures in the middle of the curve increase their risk of resistance and treatment failure.
Contrary to a common opinion, health care professionals usually cannot predict adherence behavior or identify individuals likely to take drug holidays.4,18 Research does
show that factors related to the complexity of the drug regimen and the perception of severity and progression of HIV disease may influence how well an individual adheres to a prescribed regimen.19,20 And other factors, including depression and psychological stress, may influence adherence. But socioeconomic factors such as level of education, employment, and religious support do not predict adherence.21 Strategies to improve adherence with antiretroviral drugs require active participation by both the health care provider and the patient (Table).20,22
|Recommendations for physicians
It is essential for the prescribing physician to understand and attempt to minimize the difficulties of complying with complex regimens for HIV-infected individuals. Clinicians must remember that a dosing regimen may be further complicated by other requirements, such as rigid dosing instructions. For example, some drugs, such as didanosine, must be taken on an empty stomach; others, such as saquinavir, are better absorbed from the gastrointestinal tract when taken with food. Concomitant treatment or prophylaxis of opportunistic infections and administration of other drugs for symptomatic relief can also make the antiretroviral regimen more complex.
Simplifying the dosing regimen is an important consideration when trying to improve adherence to prescribed regimens. A reduced dosing frequency (once or twice a day) almost certainly will improve compliance,4 but longer dosing intervals for drugs with short half-lives pose a different problem for nonadherent patients. Consider, for example, a drug with a half-life of about 10 hours. While it may be more convenient for a patient to take this drug every 12 hours rather than every 8 hours, a single missed dose on a twice-a-day schedule will result in relatively longer subtherapeutic concentrations. As a result, the risk of resistance and virologic breakthrough will be greater.
If one assumes minimal delay between therapeutic drug concentrations and onset of suppression of viral replication (as suggested by Vanhove and colleagues6), the length of the dosing interval relative to the drug's half-life and the steepness of the linear portion of the exposure-versus-response relationship will be important determinants of the impact of missed doses on viral replication.23
A steep exposure-versus-response relationship for lamivudine may explain in part why this drug is more prone to rapid emergence of resistance than other nucleoside drugs. A steep relationship implies that below a threshold exposure there is virtually no response, while at exposures greater than the threshold the maximum response is observed. Hence, although clinically tolerated therapeutic doses of lamivudine may be at the plateau (maximum) of the exposure-versus-response curve, short periods of missed dosing result in rapid diminution of response and rapid emergence of resistance.24
The relationship between dosing frequency and drug half-life may explain the results of ACTG 175, which showed (among other things) that didanosine monotherapy is superior to zidovudine monotherapy. Since similar nonadherence behaviors were observed with both drugs, the superiority of didanosine may have resulted from the longer intracellular half-life of didanosine triphosphate compared with zidovudine triphosphate, relative to 12- hour and 8-hour dosing intervals for didanosine and zidovudine, respectively. Hence, in this instance, twice-daily didanosine may be more "forgiving" of missed doses compared with zidovudine in a 3-times-daily regimen.25
Our data indicate that patients often demonstrate very similar patterns of adherence for drugs with similar therapeutic and side-effect profiles prescribed at the same daily dosing frequency, such as 3 times daily. This characteristic may be advantageous for dual protease inhibitor regimens, such as ritonavir plus saquinavir, which take advantage of a drug-drug interaction to minimize first-pass metabolism of saquinavir as well as the number of daily doses and dosing frequencies of the drug. As a result, both drugs may be administered in a twice-a-day regimen.26
On the other hand, the increased potential for drug-drug interactions associated with multiple-drug regimens can complicate exposure-versus-response relationships and increase risk of side effects.26 For example, an increased incidence or perception of the severity of side effects can influence adherence to drug regimens and affect outcome.27
Patient education and follow-up are especially important to ensure that patients adhere to prescribed regimens and understand the consequences of nonadherence to antiretroviral drug therapy. Health care providers should be perceived as sensitive and supportive of the patient's position.22 Patients should not be made to feel that they are being criticized or blamed for not complying with prescribed drug therapy; rather, they should be informed of the possible consequences of poor adherence and should be a part of the decision-making team regarding their drug therapy. Furthermore, dose-taking behaviors may become more erratic when routine schedules are disrupted (for example, on weekends or during vacations).28 Hence, linking dose taking to routine daily activities may improve adherence to dosing schedules.
Regimens should be tailored, if possible, to the patient's lifestyle.22 For example, if the exposure-versus-response relationship permits, a busy individual may have improved adherence and a better outcome with a twice-daily regimen of drugs taken at the same time each day.
Pharmacoeconomic evaluation of the impact of nonadherent behaviors on morbidity and mortality reveals the high cost of nonadherence in clinical practice.29 Nonadherence incurs direct costs resulting from virologic breakthrough, clinical relapse, hospitalization, and poor disease control, as well as ensuing costs associated with loss of quality of life and human productivity. Adverse treatment outcomes resulting from nonadherence to prescribed antiretroviral drug therapy make it an essential factor to be considered in prescribing appropriate regimens for patients infected with HIV. While therapeutically effective regimens are available, physicians must be judicious in prescribing them and diligent in educating and monitoring patients to ensure that they adhere to prescribed courses and so receive optimal therapeutic benefits.
1. Panel on Clinical Practices for Treatment of HIV Infection. Guidelines for the use of antiretroviral agents in HIV-infected adults and adolescents. Ann Intern Med. 1998;128(12, pt 1); MMWR. 1998;47(RR-5). (June 17, 1998 update available on line at http://www.hivatis.org/698glaa.pdf or http://www.healthcg.com/hiv/nihreport/guide/.)
2. Rosenberg ES, Hirsch MS. Starting antiretroviral therapy: one, two, or three drugs? HIV Adv Res Ther. 1997;7(2):18-23.
3. Rowland M, Tozer TN, eds. Clinical Pharmacokinetics. Philadelphia: Lea and Febiger; 1995:53-65.
4. Kastrissios H, Blaschke TF. Medication compliance as a feature in drug development. Annu Rev Pharmacol Toxicol. 1997;37:451-475.
5. Kastrissios H, Suarez JR, Flowers NT, et al. Could decreased compliance in an AIDS clinical trial affect analysis of outcomes? Clin Pharmacol Ther. 1995;57:121.
6. Vanhove GF, Schapiro JM, Winters MA, et al. Patient compliance and drug failure in protease inhibitor monotherapy. JAMA. 1996;276:1955-1956.
7. Flexner C, van der Horst C, Jacobson MA, et al. Relationship between plasma concentrations of 39-deoxy-39-fluorothymidine (alovudine) and antiretroviral activity in two concentration-controlled trials. J Infect Dis. 1994;170:1394-1403.
8. Balfour HH Jr, Fletcher CV, Erice A, et al. Effect of foscarnet on quantities of cytomegalovirus and human immunodeficiency virus in blood of persons with AIDS. Antimicrob Agents Chemother. 1996;40:2721-2726.
9. Schapiro JM, Winters MA, Stewart F, et al. The effect of high-dose saquinavir on viral load and CD4+ T-cell counts in HIV-infected patients. Ann Intern Med. 1996;124:1039-1050.
10. Sale M, Sheiner LB, Volberding P, et al. Zidovudine response relationships in early human immunodeficiency virus infection. Clin Pharmacol Ther. 1993;54:556-566.
11. Collier AC, Tartaglione T, Corey L. A reduced dose of zidovudine in patients with AIDS. N Engl J Med. 1991;324:996.
12. Cooley TP, Kunches LM, Saunders CA, et al. Once-daily administration of 29,39-dideoxyinosine (ddI) in patients with the acquired immunodeficiency syndrome or AIDS-related complex: results of a phase I trial. N Engl J Med. 1990;322:1340-1345.
13. Lambert JS, Seidlin M, Reichman RC, et al. 29,39-Dideoxyinosine (ddI) in patients with the acquired immunodeficiency syndrome or AIDS-related complex: results of a phase I trial. N Engl J Med. 1990;322:1333-1340.
14. Browne MJ, Mayer KH, Chafee SBD, et al. 29,39-Didehydro-39-deoxythymidine (d4T) in patients with AIDS or AIDS-related complex: a phase I trial. J Infect Dis. 1993;167:21-29.
15. Fletcher CV, Kawle SP, Page LM, et al. Intracellular triphosphate concentrations of antiretroviral nucleosides as a determinant of clinical response in HIV-infected patients. Presented at: 4th Conference on Retroviruses and Opportunistic Infections; January 22-26, 1997; Washington, DC. Abstract 13.
16. Danner SA, Carr A, Leonard JM, et al. A short-term study of the safety, pharmacokinetics, and efficacy of ritonavir, an inhibitor of HIV-1 protease. N Engl J Med. 1995;333:1528-1533.
17. Blaschke TF. Noncompliance and resistance to protease inhibitors. Presented at: 4th Conference on Retroviruses and Opportunistic Infections; January 22-26, 1997; Washington, DC. Abstract S43.
18. Geletko SM, Segarra M, Ravin DS, et al. Zidovudine compliance as measured by different methods in an HIV ambulatory clinic. J Pharm Technol. 1996;12:105-108.
19. Muma RD, Ross MW, Parcel GS, et al. Zidovudine adherence among individuals with HIV infection. AIDS Care. 1995;7:439-447.
20. Ickovics JR, Meisler AW. Adherence in AIDS clinical trials: a framework for clinical research and clinical care. J Clin Epidemiol. 1997;50:385-391.
21. Singh N, Squier C, Sivek C, et al. Determinants of compliance with antiretroviral therapy in patients with human immunodeficiency virus: prospective assessment with implications for enhancing compliance. AIDS Care. 1996;8:261-269.
22. Katzenstein DA, Lyons C, Molaghan JP, et al. HIV therapeutics: confronting adherence. J Assoc Nurses AIDS Care. 1997;8(suppl):46-58.
23. Levy G. A pharmacokinetic perspective on medicament noncompliance. Clin Pharmacol Ther. 1993;54:242-244.
24. van Leeuwen R, Katlama C, Kitchen V, et al. Evaluation of safety and efficacy of 3TC (lamivudine) in patients with asymptomatic or mildly symptomatic human immunodeficiency infection: a phase I/II study. J Infect Dis. 1995;171:1166-1171.
25. Urquhart J, De Klerk E. Contending paradigms for the interpretation of data on patient compliance with therapeutic drug regimens. Stat Med. 1998;17:251-267.
26. Pham PA, Flexner C. Antiretroviral drug interactions in the HIV-infected patient. HIV Adv Res Ther. 1997;7(2):10-17.
27. Kruse W, Eggert-Kruse W, Rampmaier J, et al. Compliance and adverse drug reactions: a prospective study with ethinylestradiol using continuous compliance monitoring. Clin Investig. 1993;71:483-487.
28. Girard P, Sheiner L, Kastrissios H, Blaschke T. A Markov model for drug compliance with application to HIV+ patients. Clin Pharmacol Ther. 1996;59:157.
29. Rosenberg MJ, Waugh MS. The economics of compliance in managed healthcare settings. Am J Managed Care. 1996;2:176-80.