




Lopinavir/r + Atazanavir for Salvage; Genotypic Inhibitory Quotient; pglycoprotein Inhibitor may limit resistance
The 20 Faces of HIV
European Resistance Workshop–Part 6. Pharmacology and Resistance
Mark Mascolini markmascolini@earthlink.net
Research often addresses antiretroviral resistance and pharmacology separately, although drug levels and emergence of resistant virus are immutably bound. Several studies at the Third European HIV Drug Resistance Workshop focused on this union, especially as it affects the protease inhibitors atazanavir and lopinavir. Another study plumbed the impact of Pglycoprotein, the transmembrane drug transporter, on resistance to protease inhibitors (PIs).
The genotypic inhibitory quotient (GIQ) factors drug levels and resistance into a single formula, defined in a study by Ana Rendón (Carlos III Hospital, Madrid) as minimum PI concentration divided by the number of major protease mutations [1]. She studied the effect of ritonavir boosting–and the prophetic power of GIQ–in 63 people who began unboosted atazanavir then switched to a ritonavirboosted dose of 300/100 mg daily.
This PIexperienced group had taken antiretrovirals for a median 75 months, and during that time a median 7.5 protease mutations piled up in each person. While taking unboosted atazanavir for a median 11 months, 35 (55%) reached a viral load below 50 copies/mL. The other 28 had a median load of 3.36 logs (about 2300 copies/mL) when they switched to the boosted regimen.
The median atazanavir minimum concentration measured 138 ng/mL without ritonavir and 316 ng/mL with the boost (P < 0.001). After 6 months of boosting the median CD4 count rose 69 cells/無 (interquartile range [IQR] 51 to +169 cells/無) (P = 0.009). All but one person with a viral load below 50 copies/mL on unboosted atazanavir maintained that level, while seven with a detectable load went under the 50copy mark with boosting. In a 6month ontreatment analysis, 24 of 25 people had a viral load under 50 copies/mL.
Thirteen people (21%) could not stomach 100 mg of ritonavir daily and retreated to unboosted atazanavir within 6 months.
Defining virologic response as at least a 1log (10fold) drop in viral load or a reading below 50 copies/mL after 6 months of atazanavir/ritonavir, Rendón found that neither atazanavir’s minimum concentration nor the number of protease mutations before atazanavir predicted response (Table 6.1). But GIQ did.
Table 6.1. GIQ predicts 6month response to atazanavir/ritonavir

Responders 
Nonresponders 
P 
Atazanavir Cmin (ng/mL) 
372 
224 
0.4 
Protease mutations (n) 
4 
8 
0.3 
GIQ 
113 
35 
0.04 
Cmin = minimum concentration; GIQ = genotypic inhibitory quotient (Cmin/number of protease mutations).
The failure to correlate number of protease mutations with response, suggested session chair Jonathan Schapiro, may reflect the small size of this study.
Genotypic inhibitory quotient also predicted the 12month virologic response to lopinavir/ritonavir in a group of 85 people who had tried a median of two PIs (IQR 1 to 3) for a median 43 months (IQR 25 to 65) [2]. Jolanda Hoefnagel (Radboud University Nijmegen Medical Centre, Nijmegen) found that which mutation set you put in the GIQ denominator affects the equation’s predictive power.
The group began lopinavir with a median load of 4.5 logs (IQR 3.9 to 5.1 logs) and a median CD4 count of 280 cells/無 (IQR 163 to 365 cells/無). After a year of treatment, 63 (74%) had a viral load below 500 copies/mL and 22 (26%) did not.
Hoefnagel figured GIQs by dividing lopinavir minimum concentration by the number of protease mutations derived from two sets–an International AIDS Society (IAS)USA list of mutations at 21 sites [3] and mutations at 11 sites Abbott researcher Dale Kempf identified as critical to lopinavir’s performance [4]:
 IASUSA list: mutations at positions 10, 20, 24, 30, 32, 33, 36, 46, 47, 48, 50, 53, 54, 63, 71, 73, 77, 82, 84, 88, and 90
 Kempf list: mutations at positions 10, 20, 24, 46, 53, 54, 63, 71, 82, 84, and 90
Mirroring Rendón’s atazanavir finding, Hoefnagel determined that lopinavir minimum concentration did not predict virologic response. But the number of mutations from either set and the GIQs figured with either set did correlate with 12month response (Table 6.2).
Table 6.2. Predictors of 12month salvage response to lopinavir

Responders
63 (74%) 
Nonresponders
22 (26%) 
P

Predictive cutoff

Median IASUSA score (IQR) 
3 (25) 
7 (68) 
0.002 
5.5 
Median Kempf score (IQR) 
2 (14) 
6 (47) 
0.001 
3.5 
Trough, mg/L (IQR) 
5.4 (3.76.5) 
5.3 (3.16.4) 
NS 

Median IAS GIQ (IQR) 
1.6 (0.92.4) 
0.7 (0.41.0) 
0.004 
0.9 
Median Kempf GIQ (IQR) 
3.0 (1.43.9) 
0.9 (0.71.8) 
0.001 
1.5 
GIQ = genotypic inhibitory quotient (Cmin/number of protease mutations); IQR = interquartile range.
A multivariate analysis singled out the Kempfbased GIQ as the only independent predictor of a 12month response.
But if the number of protease mutations signals how one will respond to lopinavir/ritonavir, should clinicians even bother measuring drug levels? Hoefnagel argued that the GIQ has clinical value because it incorporates a factor that clinicians can change–lopinavir’s minimum concentration. The number of protease mutations may give a clue to how a person will respond, but you can’t lower that number.
Ritonavir has carved out a niche as a PI booster without parallel. But what happens when it boosts two PIs? Work by Carlos Azuaje (Vall d’Hebron University Hospital, Barcelona) found that at least one PI–atazanavir–can give lopinavir an extra lift when both are combined with ritonavir [5].
The study involved nine men and three women whose treatment histories included at least two failures of a PI regimen. Before starting atazanavir (300 mg once daily), lopinavir/ritonavir (400/100 mg twice daily), and one to three other antiretrovirals, these people had a median viral load of 4.88 logs and a median CD4 count of 275 cells/無. They averaged five protease mutations (range two to seven) before starting the double boosted PIs and had tried an average of four earlier potent regimens.
No one had to stop their PIs because of side effects, although many had mild and transient diarrhea in the first few weeks of treatment. Three people had mild jaundice, a side effect of atazanavir.
After 6 months of treatment only one person suffered a virologic failure. That person had low atazanavir and lopinavir troughs (0.28 and 2.50 痢/mL) despite good adherence. Everyone else reached a viral load below 50 copies/mL, while the median CD4 count climbed from about 275 to 350 cells/無 (P = 0.0028).
Azuaje compared lopinavir levels with those recorded in an earlier study of 15 people taking standarddose lopinavir/ritonavir and 25 taking lopinavir/ritonavir with 1000 mg of saquinavir twice daily [6]. Lopinavir levels with atazanavir proved substantially higher than in either of these two comparison groups (Table 6.3).
Table 6.3. Higher lopinavir levels with atazanavir

LPV/r (n = 15) 
LPV/r/SQV (n = 25) 
LPV/r/ATZ (n = 12) 
AUC12h (痢/mL • h) 
85.1 
85.2 
115.7 
Cmax (痢/mL) 
10.0 
9.5 
12.2 
Cmin (痢/mL) 
5.5 
5.6 
9.1 
AUC12h = 12hour area under the concentrationtime curve; Cmax = maximum concentration; Cmin = minimum concentration.
Atazanavir’s 24hour exposure measured 48.2 痢/mL • h, its maximum concentration 3.75 痢/mL, and its minimum concentration 1.06 痢/mL.
Azuaje concluded that the doses used seem appropriate for this doubleboosted combination and that atazanavir appears to give lopinavir an extra surge. As one might expect, though, total cholesterol rose significantly during 6 months of therapy, from about 170 to 200 mg/dL (P = 0.0028). Triglycerides stayed flat at around 200 mg/dL.
Religious adherence does not guarantee good drug levels, as Azuaje’s report showed. But without faithful pill taking, the risk of low drug levels–and consequent resistance–rises dramatically. To find out how closely adherence (and antiretroviral concentrations) correlate with salvage response, Carlo Torti (University of Brescia) studied 230 people in an ongoing trial [7]. They had tried an average 2.8 regimens and had an average viral load of 3.8 logs and an average 386 CD4 cells/無 when they enrolled in the study.
Torti rated adherence by a combined score derived from a visual analog scale and two behavioral questions. He called a score under 50 poor adherence, 51 to 90 intermediate adherence, and above 90 good adherence. Defining response as a viral load below 400 copies/mL within 24 weeks of starting a new regimen and maintained through followup, Torti isolated three independent response predictors (Table 6.4). Ritonavir boosting predicted a good response, while less than good adherence and PI and nonnucleoside troughs below the 25th percentile versus above the 75th lowered the chance of a sub400copy load.
Table 6.4. Predictors of salvage response in 230 people
Variable 
Odds ratio 
95% confidence interval 
P 
Ritonavirboosted regimen 
4.16 
1.6410.52 
0.003 
Poor or intermediate adherence 
0.39 
0.190.79 
0.009 
Lowest trough quartile 
0.26 
0.120.60 
0.001 
Torti concluded that "monitoring and correction of both adherence and plasma drug concentrations may be crucial" to a good salvage response.
Because Pglycoprotein (Pgp) pumps antiretrovirals and other drugs out of the cells they target, one would expect this drug flusher to promote resistance. And it does, at least in an experimental system devised by Maria Detsika and University of Liverpool colleagues [8]. She found that protease mutations popped up more slowly in cells doused with verapamil, a Pgp inhibitor.
Detsika infected a lab line of lymphocytes (MT4 cells) with HIV then exposed them to escalating concentrations of indinavir, ritonavir, nelfinavir, and amprenavir–a technique used to prod forth resistance mutations. She treated some of the cells with 20 然 of verapamil, while leaving others verapamil free. Then she genotyped virus that evolved during the experiment.
In cells treated with amprenavir, indinavir, or nelfinavir, Detsika found that it took more of the PI to provoke mutations in cells treated with verapamil than in cells that never saw the Pgp inhibitor. Results with ritonavir were less clearcut. The findings suggest that inhibiting Pgp–and thus keeping more drug inside cells–may slow the slide to resistance with some PIs.
References
(To view slides and posters from the Third European HIV Drug Resistance Workshop, go to http://www.hivpresentation.com.)
1. Rendón A, Nóvoa SR, Barreiro P, et al. Predictors of virological response to atazanavirritonavir intensification in protease inhibitor experienced patients. Third European HIV Drug Resistance Workshop. March 30April 1, 2005. Athens. Abstract 48. Poster 8.3.
2. Hoefnagel JGM, Van der Lee MF, Koopmans PP, et al. Predictors of virological response to lopinavirritonavir in protease inhibitorexperienced patients: the genotypic inhibitory quotient. Third European HIV Drug Resistance Workshop. March 30April 1, 2005. Athens. Abstract 46. Poster 8.1.
3. Johnson VA, BrunVézinet F, Clotet B, et al. Update of the drug resistance mutations in HIV1: 2004. Topics HIV Med 2004;12:119124.
4. Kempf DJ, Isaacson JD, King MS, et al. Identification of genotypic changes in human immunodeficiency virus protease that correlate with reduced susceptibility to the protease inhibitor lopinavir among viral isolates from protease inhibitorexperienced patients. J Virol 2001;75:74627469.
5. Azuaje C, Lopez RM, Ribera E, et al. Pharmacokinetics and safety of a double boosting regimen of atazanavir, plus lopinavir, plus minidose ritonavir in multidrugtreated HIVinfected patients. Third European HIV Drug Resistance Workshop. March 30April 1, 2005. Athens. Abstract 51. Poster 8.6.
6. Ribera E, Lopez RM, Diaz M, et al. Steadystate pharmacokinetics of a doubleboosting regimen of saquinavir soft gel plus lopinavir plus minidose ritonavir in human immunodeficiency virusinfected adults. Antimicrob Agents Chemother 2004;48:42564262.
7. Torti C, QuirosRoldan E, Moretti F, et al. The relative prognostic value of adherence and plasma drug concentrations on the virological response after salvage HAART: the RADAR group of the MASTER cohort. Third European HIV Drug Resistance Workshop. March 30April 1, 2005. Athens. Abstract 47. Poster 8.2.
8. Detsika MG, Chandler B, Owen A, et al. Evolution of HIV1 under PI drug pressure in the presence and absence of a Pgp inhibitor. Third European HIV Drug Resistance Workshop. March 30April 1, 2005. Athens. Abstract 41. Poster 7.5










