Integrating TDM with Phenotypic Assay Tests

The development of lopinavir and its subsequent combination with “mini-dose” ritonavir (lopinavir/r) to reduce its hepatic metabolism and increase the area under the plasma concentration versus time curve (AUC) provides a clinical example of integrating trough concentration data with HIV resistance testing.

Hsu et al reported on study M98-957 examining lopinavir/r that was conducted in patients with multiple prior PI experience but no NNRTI experience. The initial dose of 400/100 was given with efavirenz for either two weeks or 24 weeks before increasing to 533/133 bid. In this study the pharmacokinetics of lopinavir/r were influenced by the inclusion of the NNRTI, efavirenz. The hepatic induction effects of efavirenz lowered lopinavir concentrations requiring an increase in the dose. The subjects were stratified by their baseline inhibitory quotient into those with a value of <1, 1-4, 4-10 and >10 had a response rate at 24 weeks that was predicted to a gfreater degree by the IQ rather than the lopinavir trough or phenotype alone (Hsu).

Kempf et al reported on M98-985 in which a group of subjects who had a plasma HIV RNA > 50 copies/ml and received low-dose ritonavir to increase their indinavir plasma concentrations. This study noted a good correlation between the number of PI mutations and the percentage of subjects achieving a virologic response (>0.5 log copies/ml from baseline or a viral load < 50 copies/ml). While the correlation between the phenotype and the virtual phenotype in this small sample of subjects was not as high as expected, overall in a larger database, Virco Inc. has established a good correlation between these two parameters. These investigators calculated a virtual “inhibitory quotient” (VIQ) by determining the quotient: trough PI plasma concentration/ virtual phenotype (fold change) x the wildtype IC 50 corrected for protein binding. The VIQ was a better predictor of virologic response than either the phenotype or the drug concentrations alone. Subsequently the subjects were separated into groups with a virtual IQ of < 0.32, 0.32-1, 1-3.2, 3.2-10, 10-32 and 32-100. Virologic response was better predicted by VIQ at weeks 3, 24 and 48 as compared to correlations with virtual phenotype, number of PI mutations or the number of RT mutations.

The IQ is a ratio and changes in either the numerator (AUC, Cmin) or the denominator (IC 50) can have an important influence on the antiviral activity contributed by that antiviral to the overall effect of the regimen. It is also noteworthy that the IQ can only be determined for each drug in a combination regimen, and the optimal manner in which three or four IQ calculations can be combined is unknown. Interestingly, the development of the plasma HIV RNA as a marker of potency of a combination regimen may allow for a systematic approach to overall IQ calculations to be developed and tested in vivo to determine the optimal IQ.

PI-Boosting” an attempt to enhance the IQ- In general, single PI-containing regimens approach low trough plasma concentrations at the end of their recommended dosing interval. Low trough concentrations, leave little room for variations in individual patient medication administration times and in general are probably not well-tolerated in patients with erratic adherence patterns. The cycle of low trough concentrations, reduced periods of sustained viral suppression probably contribute to low-level replication that after a period of time lead to loss of viral suppression and increases in plasma HIV RNA. Genotypic assays conducted at this time often yield virus with certain mutations in the protease region. Therefore, the cycle leads to reduced virus susceptibility which even though quantitated by the phenotypic assay is likely to yield a lower antiviral response in “salvage therapy regimens”.

References/Abstracts

Therapeutic Drug Monitoring