Day Two-ICAAC, Friday: Pharmacokinetics (PK), New Drugs
The new Bristol-Myers protease inhibitor (BMS-232632) wasdiscussed today. So far studies have only been phase I in uninfected individuals,and in vitro studies. Data suggets it may not be cross-resistant with saquinavirand may have limited cross-resistance wih other PIs. Dr Richard Colonnaof Bristol-Myers suggested the PK profile and halflife suggests the drugmay be once-a-day. In vitro, it appears to be potent. However, it is crucialto remember that this data is very preliminary. It hasn't yet been studiedin HIV-infected humans. But, such studies are expected to begin in 1st quarter'99. Tomorrow is a small communiy meeting with Bristol and I will emphasizeto them that when they begin studies they must initiate salvage trials parallelwith naïve individuals as soon as feasible.
PK and Abacavir and Amprenavir
Dr George Drusano discussed today how amprenavir takenevery 8 hrs had better antiviral activity than when taken twice a day. Hecautioned that because of adherence concerns he is not recommending tidinstead of bid. But a person is more likely to get better antiviral activityfrom amprenavir taking it tid rather than bid. And, more individuals wouldbe likely to reach undetectable if they used it tid. It appears, however,that amprenavir has a good PK profile. Although Glaxo is not prepared todevelop it as a once a day drug, the blood levels remain above the EC95very nicely. There is more flexibility in terms of dosing, in my opinion,than with nelfinavir tid.
There was a discussion about PK parameters and drug efficacy.Some drugs' efficacy are related to their daily AUC (blood levels) and otherdrugs' efficicay are more driven by their trough and its relationship withtheir Cmin. It appears as though abacavir may be a daily AUC drug. Therefore,it may be explored, I think, at 600 mg once-a-day. But, PIs are not dailyAUC driven, they are Cmin/trough driven.
Drusano reported thee is in vitro synergy between amprenavirand abacavir. But that abacavir does not have such a relationship with ptherPIs.
He is trying to develop an assay whereby you can test aperson's viral load several times after beginning therapy to predict ifthat person will reach <50 copies/ml. He and others have expressed herethat early intensification may be a good approach to preventing failure.Preliminary ACTG data suggests if you have not reached <500 copies/mlby week 8, or if you don't have at least a 0.5 log reduction by week 4,after starting therapy, you may not reach <50 copies/ml. Intensificationby adding a new drug may be the answer.
Agouron reported in this PK session that baseline viralload, Viracept blood levels and initial 4-week change in viral load wereearly predictors of long-term virologic response to nelfinavir. This isnot a mystery. This confirms information others have reported. The ACTGdata mentioned in previous paragraph supports the Agouron data.
AG-1549
This is the new Agouron NNRTI.Their preliminary data suggestsit may not be cross-resistant with other NNRTIs but this is very preliminary.This will have to be confirmed in studies and Agouron will have to provethis by taking individuals who've failed Sustiva, nevirapine or delavirdineand testing if they respond to AG-1549. Initial studies showed a 1.3 logreduction in viral load at the highest dose used so far. They said resistancewas slow to develop.
AG-1776
Agouron says it has unique resistance profile but the 82and 84 are positions for mutations other PIs also acquire for resistance.They reported AG-1776 resistant HIV has protease substitutions L10F, R41K,I47V, V82I, or I84V. Agouron said AG-1776 was active against a wide spectrumof viruses containing PI resistance associated mutations. In animals, 1776plasma levels were maintained above the EC95 >12 hrs. Again, Agouronwill have to prove this drug is not cross-resistance in humans. They willhave to test its efficacy in individuals who've failed other PIs.
In the session on resistance today, Dr Richard Harrigandiscussed an abstract he first presented at the Resistance Meeting at LakeMaggiori just prior to the Intl Geneva Conference. At Maggiori, his abstractwas one of several retrospective studies showing that knowing baseline genotypicand/or phenotypic resistance for a person can help predict the responseto drugs in a regimen. He concluded that baseline resistance to drugs predicteda lack of response. And that, genotypic resistance or phenotypic resistancetesting could be relied on equally.
Harrigan looked at individuals who had received RTV+SQV.They determined the patients baseline genotypic and phenotypic resistanceby going back and testing stored blood samples and following the patient'sviral load response after starting theapy. This was a group of patientswho were heavily pre-treated: about 50% were naïve to NRTIs, and 35-56%were protease naïve. The baseline CD4 and viral load were 110-160 andabout 58,000 copies/ml. They used the Virco genotypic and phenotypic resistancetests. Harrigan found a correlation between genotypic and phenotypic reistanceto both SQV and RTV. The higher the phenotypic resistance the higher thegenotypic resistance. Genotypic resistance is called by either sensitive,intermediate or resistant. Both tests gave similar results. In other words,whether you used the genotype or phenotype test both gave similar results.Patients who were naïve to SQV at baseline did not show phenotypicresistance to SQV at baseline, and patients who were RTV naïve didnot show RTV phenotypic resistance. But for those who were RTV or SQV experiencedsome showed resistance to that drug and some did not show phenotypic resistanceto that drug. The genotypic calls were also consistent with prior experience.No one who was naïve to either RTV or SQV showed genotypic resistanceto that drug.
No individuals who had phenotypic baseline resistance >4-foldto RTV and SQV went to undetectable for both protease naïve and experiencedpatients. But some individuals with little or no phenotypic resistance atbaseline did not respond with a viral load reduction. Some individuals withno resistance to RTV or SQV (20-30%) did not achieve a 0.5 log reduction.
He concluded-