Should we now adopt the HIV-RNA < 50 copy endpoint for clinical trials of antiretroviral-experienced as well as naive patients?

A randomized controlled trial to assess safety, tolerability, and antepartum viral load with increased lopinavir/ritonavir dosage in pregnancy

HIV mother-to-child transmission (MTCT) is significantly reduced if antepartum viral load (apVL) is<50 copies/mL. Pharmacokinetic studies suggest increasing the dosage of lopinavir/ritonavir (LPV/r) in pregnancy. It is important to assess tolerance, safety, and rate of patients presenting a apVL<50 copies/mL when treating with increased dose of LPV/r during pregnancy. Confirmed HIV-infected pregnant women with a fetus at a gestational age of 14-33 weeks were randomly assigned to receive LPV/r 400/100 or 600/150 mg b.i.d. plus two nucleoside analogues (NRTIs). Treatment was discontinued in the case of alanine transaminase (ALT) of grade III elevation or higher, glucose, or triglycerides. Thirty-two women were randomized to the LPV/r 400/100 mg dose, and 31 women were randomized to the 600/150 mg dose. Overall, 9.4% of the women receiving the conventional dose, and 17.2% receiving the increased dose, discontinued treatment because of adverse events (p=0.29). The rates of gastrointestinal (GI) symptoms, laboratory abnormalities, preterm delivery, and low birth weight were similar in both groups. There were no cases of HIV MTCT. Among the women with a baseline VL>50 copies/mL assigned to the conventional dose group, 45% (95% confidence interval [CI] 62.5-27.5%) had a apVL>50 copies/mL compared with 10.5% (95% CI 21.6-0.6%) of those assigned to the increased dose group (p=0.01). There was no significant difference found for the patients with a baseline VL<50 copies/mL. In pregnant women with a baseline VL>50 copies/mL, it may be warranted to initiate LPV/r dosing at 600/150 mg, whereas the conventional dose is sufficient for pregnant women with a baseline VL<50 copies/mL.

Dimension reduced kernel estimation for distribution function with incomplete data

This work focuses on the estimation of distribution functions with incomplete data, where the variable of interest Y has ignorable missingness but the covariate X is always observed. When X is high dimensional, parametric approaches to incorporate X - information is encumbered by the risk of model misspecification and nonparametric approaches by the curse of dimensionality. We propose a semiparametric approach, which is developed under a nonparametric kernel regression framework, but with a parametric working index to condense the high dimensional X - information for reduced dimension. This kernel dimension reduction estimator has double robustness to model misspecification and is most efficient if the working index adequately conveys the X - information about the distribution of Y. Numerical studies indicate better performance of the semiparametric estimator over its parametric and nonparametric counterparts. We apply the kernel dimension reduction estimation to an HIV study for the effect of antiretroviral therapy on HIV virologic suppression.

Methodological issues in the use of composite endpoints in clinical trials: examples from the HIV field

BACKGROUND

In many fields, the choice of a primary endpoint for a trial is not always the ultimate clinical endpoint of interest, but rather some surrogate endpoint believed to be relevant for predicting the effect of the intervention on the clinical endpoint. The classic example of such a field is clinical HIV treatment research, where a variety of primary endpoints are used to evaluate the efficacy of new antiretroviral drugs or new combinations of existing drugs. The choice of endpoint reflects either the goal of therapy as recommended by treatment guidelines (e.g. rapid virological suppression) or the licensing requirements of official drug approval organizations (e.g. time to loss of virological response [TLOVR]).

PURPOSE

To review the diversity of endpoints used in recent clinical trials in HIV infection and highlight the methodological issues.

METHODS

We identified articles relating to antiretroviral therapy by searching PubMed and through hand searches of relevant conference abstracts. We restricted the search to randomized controlled trials conducted in HIV-infected adults published/presented from January 2005 until March 2008.

RESULTS

We identified 28 trials in antiretroviral-naive patients (i.e. patients who were starting antiretroviral therapy for the first time at the time of randomization) and 23 trials in antiretroviral-experienced patients. Most trials were performed for purposes of drug licensing, but others were focused on strategies of using approved drugs. Most trials (40 of 51) used a composite primary endpoint (TLOVR in 13). Of note, 22 of these 40 studies reported that they had used a purely virological efficacy endpoint, but the primary endpoint was actually a composite one due to the way in which missing data and treatment switches were considered as failures.

LIMITATIONS

Examples are restricted to HIV clinical trials.

CONCLUSIONS

Whilst most current HIV clinical trials use composite primary endpoints, there are substantial differences in the components that make up these endpoints. In HIV and other fields where precise definitions are variable, guidelines for standardization of definition and reporting would greatly improve the ability to compare trial results.

Costs to achieve undetectable HIV RNA with darunavir-containing highly active antiretroviral therapy in highly pretreated patients: the POWER experience

BACKGROUND

Very few studies have evaluated the cost of highly active antiretroviral therapy (HAART) per successful treatment in HIV-infected patients.

OBJECTIVES

To evaluate the cost of achieving undetectable plasma HIV-RNA levels in highly treatment-experienced, HIV-1-infected adults receiving darunavir/ritonavir (DRV/r 600 mg/100 mg twice a day) or control protease inhibitor (PI)-based HAART.

METHODS

The mean annual per-patient cost of DRV/r and control PI-based HAART was determined from the proportional use of antiretroviral agents in the DRV/r and control PI arms of the pooled POWER 1 and 2 trials, applying drug acquisition costs for 13 healthcare settings. The mean annual cost per patient of achieving undetectable plasma HIV-RNA levels (<50 copies/mL) was calculated by dividing the cost of each treatment by the proportion of patients with undetectable plasma HIV-RNA levels after 48 weeks in the DRV/r (45%) and control PI (10%) arms of the POWER trials.

RESULTS

Whereas absolute costs of treatment were 1-19% higher with DRV/r versus control PI-based HAART depending on the healthcare setting, the mean annual per-patient cost of achieving undetectable plasma HIV-RNA levels was 73-78% lower. These cost savings were maintained in the sensitivity analyses, adjusting for control PI and enfuvirtide use, and the number of active drugs in the background regimen. The incremental annual cost per additional patient achieving undetectable plasma HIV-RNA levels with DRV/r versus control PI-based HAART in POWER 1 and 2 (£4148) compared favourably with that determined for enfuvirtide (£137, 740; TORO trials) and tipranavir/ritonavir (£32,176; RESIST) versus control therapy.

CONCLUSIONS

DRV/r-based HAART provided consistent reductions in the cost of achieving undetectable plasma HIV-RNA levels compared with control PI-based therapy in highly treatment-experienced patients across various healthcare settings. The incremental cost per additional patient achieving undetectable plasma HIV-RNA levels with DRV/r versus control PI-based HAART was also lower than that calculated for other treatment options in this population. These results suggest that DRV/r is an economically viable option for highly treatment-experienced patients.

A nonparametric likelihood test for detecting discordance between two measurements with application to censored viral load determinations

Patient management frequently involves quantitative evaluation of a patient's attributes. For example, in HIV studies, a high viral load can be a trigger to initiate or modify an antiretroviral therapy. At times, a new method of evaluation may substitute for an established one, provided that the new method does not result in different clinical decisions compared with the old method. Traditional measures of agreement between the two methods are inadequate for deciding if a new method can replace the old. Especially, when the data are censored by a detection limit, estimates of agreement can be biased unless the distribution for the censored data is correctly specified; this is usually not feasible in practice. We propose a nonparametric likelihood test that seamlessly handles censored data. We further show that the proposed test is a generalization of the test on nominal measurement concordance to continuous measurement. An exact permutation procedure is proposed for implementing the test. Our application is an HIV study to determine whether one method of processing plasma samples can safely substitute for the other. The plasma samples are used to determine viral load and a large portion of data are left censored due to a lower detection limit.