Poor Differential Adherence of Ritonavir Tablets Used as Protease Inhibitor Booster

Dear Editor:

Protease inhibitors (PIs) are potent antiretroviral agents that in combination regimens have dramatically changed the natural history of HIV disease. However, therapy with many PIs at standard doses may be limited by inconvenient dosing regimens, high tablet volumes, and variable drug exposure. Booster agents, such as low doses of ritonavir (RTV) are increasingly used to maintain high PI trough exposures. ¹ The advantages of the boosted PI approach include raising trough drug concentrations, diminishing interpatient variability, prolonging drug half-life to allow twice-daily and possibly once-daily dosing, and diminishing food requirements and tablet volume. In addition, increases in drug exposure may potentially enable inhibition of the virus in the presence of reduced sensitivity to PIs. The administration of RTV in independent tablets is associated with disadvantages, including lack of achievement of optimal efficacy of the PI regimen, increased number of tablets, and the fact that RTV tablets have size, flavor, and storage characteristics that are poorly accepted by patients, which may adversely affect adherence and compromise the efficacy of highly active antiretroviral therapy (HAART). ² Moreover, the number of tablets in the RTV container complicates the coordination of drug collection with the remaining antiretroviral agents, a circumstance that can also favor missed or skipped doses. Although RTV-booster regimens do not require the strict >95% adherence of traditional PI-based regimens, less than that percentage is strongly associated with increased mortality over time. ³

Medication event monitoring systems (MEMS) and self-reported questionnaires are common methods to measure adherence. MEMS can be time-consuming and expensive, whereas self-reported questionnaires are subject to measurement bias, such as recall and response bias. ⁴ Another method to assess compliance is the use of pharmacy records. The refill records of computerized pharmacy systems are used increasingly as a source of compliance information. Because refill compliance data only give information about whether or not the medication is collected by the patient, it provides an upper bound for medication consumption, and allows identification of those patients that cannot be compliant simply because they do not obtain enough medication. ^5,6 Moreover, incorporating refill-based measures of adherence into clinical practice facilitates early identification of subjects who may experience virologic failure because of poor adherence, and may be useful in real time to determine whether an individual is exhibiting incomplete adherence. ⁷ We compared refill data for RTV and the PI boosted with RTV tablets in patients with good adherence to PI medication (pharmacy refill rate >80%) and examined the possible effect of poor selective adherence to RTV on the efficacy of antiretroviral treatment.

Between April 1 and April 30, 2009, a cross-sectional study of all HIV-1–infected adult patients treated with antiretroviral drugs that included a PI boosted with RTV as an independent drug was conducted. To be included in the study good adherence to PI medication (pharmacy refill rate >80%) and maintenance of the same regimen for at least the previous 6 months were required. Poor differential adherence to RTV was defined as a difference in the percentage of obtaining PI and RTV tablets greater than 10%. The patient's medical record was used to assess viral loads >100 copies per milliliter before and over 1 month after the study, identification of relevant resistances in the reverse transcriptase or protease, and the need of treatment changes due to virological failure in the subsequent 6 months. The SPSS (version 15.1, SPSS Inc., Chicago, IL) computer program was used for data analysis. Statistical significance was set at p < 0.05.

Of the 244 patients on antiretroviral therapy, 56 (30%) had maintained the same PI boosted with RTV tablets for at least 6 months. PIs included atanazavir in 19 subjects, darunavir in 19, fosamprenavir in 16, tripanavir in 1, and saquinavir in 1. Of these 56 patients, 45 with a pharmacy refill rate greater than 80% were included in the study. The mean (standard deviation) drug collection period was 273 (57) days. Poor differential adherence to RTV was documented in 8 (18%) patients, viremia (>100 copies per milliliter) in at least one occasion in some of the tests performed during the study period in 16 (35.5%), relevant resistances in 6 (13.3%), and need to change antiretroviral therapy due to treatment failure in 9 (20%). Two further patients with treatment failure were lost to follow-up. Poor differential adherence to RTV was significantly associated with the presence of detectable viremia but did not show any relation with detection of resistances or need to modify HAART in the subsequent months.

The present results indicate that the use of RTV tablets to booster PIs favors poor selective adherence. This poor adherence might explain the high frequency of viremia in patients who otherwise were compliant with the PI regimen. The fact that poor differential adherence to RTV was not associated with detection of resistances or the need to modify HAART may be explained by explicit recommendations to patients to improve adherence to RTV tablets as well as by the small sample size and the two cases lost to follow-up. In summary, clinicians should be particularly alert in HIV-1–infected patients receiving RTV tablets as a booster because poor adherence is frequent and may be a risk factor for detectable viremia. The refill records of pharmacy registry systems offer a simple and objective tool for assessing differential adherence to antiretroviral drugs. The combination RTV/PI pills could overcome this issue.