Can Monitoring Drug Levels Help Treat Drug-Resistant HIV-AIDS?

Nancy Sheehan, Pharmacy Department, McGill University Health Center

Nancy Sheehan, B.Pharm, M.Sc

The possibility that antiretroviral drug serum levels can guide "salvage" regimens for treatment-resistant HIV-infected patients was explored in a recently published case series.

"Although prospective studies have shown the utility of therapeutic drug monitoring with unboosted protease inhibitors in the context of antiretroviral-naїve patients, little evidence is available in treatment-experienced patients and with more potent and better tolerated antiretroviral regimens," wrote Nancy Sheehan, BPharm, MSc, Pharmacy Department, McGill University Health Center, Montréal, Québec, Canada and colleagues.

The investigators identified 14 HIV-infected patients with multi-resistant virus who were to receive a darunavir (Prezista, Janssen)-based "salvage" regimen. In the open-label, prospective study conducted from 2007 to 2010, drug levels were monitored and resistance-associated mutations (RAM) identified to ascertain whether any pharmacokinetic/pharmacodynamic (PK/PD) parameters predicted therapeutic response.

The patients were treated with darunavir (DRV) 600mg and ritonavir (Norvir, Abbvie) 100mg twice daily and an optimized background regimen selected by a specialist provider which could include etravirine (Intelence, Janssen) 200mg twice daily and/or raltegravir (Isentress, Merck).

The primary measure of efficacy was virologic response with viral load of <50 copies/ml at week 48. Secondary efficacy endpoints included percentage of patients achieving virologic response at week 24 and a viral load decrease of ≥1 log at week 12. Treatment failure corresponded to a failure to achieve or maintain a viral load of ≤200 copies/ml.

PK/PD included the Genotypic Inhibitory Quotient (gIQ) for darunavir, calculated as the ratio between the median of all DRV trough serum concentration levels and the number of RAM; and the Instantaneous Inhibitory Potential (IIP) of the 3-drug regimen, an index which integrates the dose-response curve slope that is used principally to predict in vitro antiviral activity.

"Antiviral activity is greatly influenced by the slope, a parameter which is not considered in inhibitory quotients," Sheehan and colleagues indicated. "There is still, however, no consensus on the relevance of the IIP for the prediction of virologic outcome in vivo."

Thirteen patients were maintained on treatment through 48 weeks and all had controlled viremia, with 2 (15.4%) having viral loads of between 50-200 copies/ml. At 48 weeks, 79% had a viral load <50 copies/ml. The only patient with 2 consecutive viral loads >50 copies/ml at the end of the study period had the lowest IIP. Neither the IIP, nor any other PK/PD parameter was associated with virologic response in the univariate logistic regression analysis, however.

"Remarkably, according to the published cut-off values, none of the previously proposed PK/PD parameters appeared to be predictive of virologic outcome in our cohort," Sheehan and colleagues remarked.

Based on their observations, however, the investigators suggest that the utility of IIP for the prediction of virologic outcome warrants further study in a larger population. They also encourage additional research into prediction algorithms, noting that others have considered integrating factors such as drug distribution in lymphoid tissues and treatment adherence.

The study of the utility of therapeutic drug monitoring in a series of treatment-experienced HIV-infected patients was published in the April issue of Antiviral Research.