ISSUE 76 15th October 1999

Editor Paul Blanchard

Medical Consultant

Conference Report Special (Part 1)

39th ICAAC, San Francisco, USA
September 26-29, 1999






39th Interscience Conference on Antimicrobial Agents & Chemotherapy,
San Francisco, USA.
September 26-29, 1999

Triple nucleoside analogue combinations

Abacavir study CNA3005: Follow-up data on the full cohort at 48 weeks

Paul Blanchard, ATP.

Combination therapy with three nucleoside analogues is a newer strategy being explored as an alternative to the established approach combining 2 nucleosides with either a protease inhibitor or an NNRTI. The potential advantages of a triple nucleoside approach include reduced pill burden compared to PI containing combinations and the possibility of suppressing viral replication for a period before initiating PI or NNRTI drugs. Concerns remain however over the potential for additive longer-term toxicity's when combining three drugs of the nucleoside class. The ability of non-PI containing regimens to foster qualitative immune system reconstitution and reversal of the structural deterioration seen in lymph nodes in more advanced disease also remains to be proven.

Follow-up data to 48 weeks for the full cohort of study CNA3005 was presented at ICAAC by Schlomo Stazewski. This is a phase III, randomised, double-blind, placebo-controlled study of the triple nucleoside combination of abacavir (ABC)/ZDV/3TC compared to the PI based combination of indinavir(IND)/ZDV/3TC in antiretroviral na夫e subjects. The ZDV/3TC used in this study was administered in both arms as the coformulated tablet Combivir(. To maintain blinding adherence to three times daily dosing, indinavir fasting requirements and a daily dosing of 16 pills (including placebo) was maintained across both arms. After 16 weeks subjects with plasma HIV RNA remaining above 400 copies/mL could opt to change to open-label therapy.

Mean HIV RNA and CD4 count at baseline in the 562 subjects enrolled were 4.8 log copies/mL and 360 cells/mm3 respectively. Each treatment arm was prospectively stratified by entry viral load into two groups, stratum A with >10,000 to 100,000 copies/mL and stratum B with >100,000 copies/mL. Unfortunately the discontinuation rate in this trial was high with only 55% of subjects in the ABC arm and 45% in the IDV arm completing week 48. Virologic failure rates were, however, low in both arms. Discontinuation rates were no doubt influenced by the necessities of dosing associated with maintaining blinding. Approximately 19% of patients in both arms discontinued treatment due to adverse events (including a 7% rate of abacavir hypersensitivity in those assigned to the abacavir arm).

The intent-to-treat analysis using an HIV RNA assay with a 50 copies/mL cut-off revealed a superior response for the indinavir arm with 46% of subjects <50 copies/mL versus 40% of abacavir subjects at 48 weeks. An analysis of the subgroups stratified by viral load at entry revealed that this difference was largely accounted for by those subjects in stratum B (>100,000 copies/mL) [see table below].

ITT analysis at 48 weeks % <50 copies/mL:

All subjects 46% 40%
10,000 to 100,000 46% 45%
>100,000 45% 31%

Using a less sensitive HIV RNA assay 51% of subjects in both treatment arms were <400 copies/mL at week 48 (intent-to-treat). Both indinavir and abacavir containing combinations showed similar CD4 cell count rises at 142 and 149 cells/mm3 respectively.

Ref: Staszewski S, Keiser P, Gathe J, et al. Comparison of response with abacavir/Combivir to indinavir/Combivir in therapy-na夫e adults at 48 weeks. 39th ICAAC; September 26-29, 1999; San Francisco, CA. Abstract 505.

Return to Contents

Atlantic Study: 48-week data

Steven G Deeks, MD. HIV InSite.

The question of whether to select a protease-based or protease-sparing initial regimen has received a great deal of attention because of the desire to minimise side effects, improve adherence, and maximise future therapeutic options.

The Atlantic study, a multisite, international study conducted by Murphy and colleagues, compared 3 initial antiretroviral regimens, each based on a dual nucleoside d4T/ddI combination. Regimens included either a protease inhibitor (indinavir), a non-nucleoside (nevirapine) or a third nucleoside (3TC). The 298 participants entering the study were antiretroviral-naive, asymptomatic, and had CD4 counts above 200 cells/mm3 and HIV RNA above 500 copies/mL.

In the as-treated analysis at 48 weeks, the percentage of patients with HIV RNA below 50 copies/mL in the indinavir, nevirapine, and triple-nucleoside arms was 90, 82, and 78%, respectively. In the more conservative intent-to-treat analysis, the percentage was 57, 51, and 49%, respectively. The response among treatment arms, at least at 48 weeks, was similar, although patients with baseline HIV RNA above 4.36 log10 copies/mL were less likely to go below 50 copies/mL when randomised to the triple nucleoside arm

Ref: Murphy RL, Katlama C, Johnson V, et al. The Atlantic Study: A randomised, open-label trial comparing two protease inhibitor (PI)-sparing antiretroviral strategies versus a standard PI-containing regimen. 39th ICAAC; September 26-29, 1999; San Francisco, CA.

Abstract LB-22. Report source: HIV InSite, UCSF Positive Health Program. See:

Return to Contents

Durable suppression of viral load and continuing rise in CD4 cell count may be possible with triple nucleoside therapy

Paul Blanchard, ATP.

Hittinger and colleagues from Toulon, France identified 67 antiretroviral na夫e patients treated in their hospital with triple nucleoside therapy for longer than 48 weeks (in 13 patients for 144 weeks). 49 patients received a combination of ZDV/ddI/3TC and 18 received d4T/ddI/3TC. Mean baseline plasma HIV RNA and CD4 counts of all 67 patients were 4.96 log copies/mL and 356 cells/mm3 respectively.

At week 48, 45 out of 51 evaluable patients (88.2%) had HIV RNA <200 copies/mL and experienced a mean gain of 236 CD4 cell/mm3. Thirteen patients had reached 144 weeks and 11/13 of these patients were <20 copies/mL HIV RNA, 2/13 were around 250 copies/mL. These thirteen subjects also experienced continued and significant gains in CD4 count with a mean increase of 129 cells/mm3 from week 48 to week 144. Tolerability of both the ZDV and the d4T containing combinations was reported to be good.

Ref: Hittinger G. Sustained suppression of HIV-1 replication using a combination of three nucleoside reverse transcriptase inhibitors (NRTI) in na夫e patients. 39th ICAAC, September 26-29, 1999. Abstract 1972.

CNA3005 hints at the possible utility of triple nucleoside analogue combination therapy based on abacavir. However, the reduction in performance over PI based therapy at moderate viral loads suggests that such regimens may not be optimally suppressive.

Additionally, pill burden of this triple NA regimen is no different from both nevirapine or efavirenz NNRTI combinations when these are combined with Combivir.

The Atlantic study provides further support for non-PI based regimens (in treatment-naive patients with a low viral load). However, the study is not complete, and it may not be adequately powered to demonstrate true equivalence among the three treatment approaches. Considering the low viral load at baseline, differences may become evident only after long-term follow-up. The current trend suggests that virologic failure on the triple nucleoside analogue regimen (d4T, ddI and 3TC) may be more frequent.

In conclusion caution should be exercised in the use of this strategy in those with higher viral loads, symptomatic disease and lower CD4 counts.

Return to Contents

Protease Inhibitors

Alternative dosing schedules for protease inhibitors explored

Paul Blanchard, ATP.

Three times daily dosing of antiretrovirals intended for long-term administration is being perceived as increasingly problematic by both patients and doctors. The development of once daily nonnucleoside reverse transcriptase inhibitors (NNRTIs) and the recent European licensing of ddI in a once-daily formulation has further highlighted these issues. In Europe, the only protease inhibitor (PI) currently licensed for less than three times daily administration is ritonavir which is dosed every 12 hours. There are, however, significant tolerability problems associated with the 600mg dose needed for 12 hourly administration.

Combinations of protease inhibitors exploit interactions in metabolism and pharmacokinetics allowing the exploration of alternative dosing schedules. Indeed many physicians are now exclusively prescribing indinavir, for instance, only as part of the double PI combination of indinavir/ritonavir together with nucleoside analogues (NA's). As originally reported by Dr Cassy Workman, this PI combination gives prolongs the half-life of indinavir allowing both twice daily dosing and the relaxation of dietary intake restrictions around the dosing time. Various 12 hourly dose combinations of indinavir/ritonavir have been explored ranging from 400mg/400mg of each to 800mg indinavir administered with 100mg of ritonavir. The lower dose of ritonavir in this combination appears to improve tolerance and may be associated with less elevation of plasma triglycerides as reported at this meeting by Dr Mike Youle and colleagues (1).

Similarly, combinations of ritonavir/saquinavir and saquinavir/nelfinavir in 12 hourly dosing have displayed both pharmacokinetic and virological outcomes suggesting less than 3 times daily dosing in combination is possible. A newer protease inhibitor, amprenavir, has recently received approval for marketing in the US with 12 hourly dosing, but this agents remains unlicensed in Europe at the time of writing. Indeed, advantages might also be obtained even for amprenavir by combination with another protease inhibitor to reduce the high pill burden of this compound.

Once daily indinavir/ritonavir?

David M Burger and colleagues from the Netherlands presented a poster evaluating indinavir levels in 12 non-HIV infected healthy volunteers across a range of once daily indinavir/ritonavir doses (2). Dr Burger concluded that in these subjects a dose of 1,200mg indinavir (IDV) with 200mg ritonavir (RTV) was optimal for once daily administration. An initial induction dosage of 1,200mg IDV/400mg RTV may be necessary however to ensure adequate initial trough values of indinavir. Of some concern, the median IDV Cmax of this once daily combination was 18,241 nM, approximately 1.6 times the mean Cmax seen with 800mg q8h in the Merck 021 study. There is therefore a theoretical concern that once daily IND/RTV combinations at this dosage may have the potential for increased adverse effects such as indinavir nephrolithiasis. Alfred Saah and colleagues from Merck presented similar data from the Merck 089 study, again in healthy subjects (3). IDV/RTV once daily at doses of 800/100, 800/200 and 1200/100 all produced IDV exposure (AUC24h) comparable to 800mg q8h seen in the Merck 021 study. Cmax was, however, elevated above the 800mg q8h levels in all IND/RTV combination dosages, 1200/100 producing a 1.8-fold increase.

Once daily saquinavir/ritonavir?

Similarly to IND/RTV, saquinavir/ritonavir has become a common twice daily PI combination. Michael Saag and colleagues at the University of Alabama presented their results from a study of SQV/RTV once daily in 41 healthy non-HIV infected subjects (4). This study used the soft gel capsule formulation of saquinavir (Fortovase). The dose combinations studied were 1,200/100, 1,600/100, 1800/100 and 1,200/200mg of SQV/RTV. Compared to SQV-sgc 1,200 mg tid (licensed dose) all once-daily combinations showed favourable SQV concentration profiles. Based on these data a once-daily dose of 1,600/100 was selected for further evaluation in HIV-infected patients.

Twice daily administration of SQV-sgc has already been explored in combination with nelfinavir in the TIDBID study. Further data from TIDBID was also presented at ICAAC (see this issue of DocFax).

Tipranavir + ritonavir: Twice daily dosing and reduced pill burden

Tipranavir is a newer protease inhibitor under development at Pharmacia & Upjohn. It is the first non-peptidomimetic PI to reach clinical studies and is generating interest due to its in vitro activity against HIV highly resistant to multiple PI's. Tipranavir is an inducer of CYP3A and like saquinavir its development has been dogged by poor pharmacokinetics. The potential for pharmacokinetic enhancement by combination with the CYP3A inhibitor ritonavir was therefore investigated by J R Baldwin and colleagues at Pharmacia & Upjohn and their data presented at ICAAC (5). Two studies were conducted in healthy non-HIV infected subjects of both tipranavir alone and various dosage combinations of tipranavir + ritonavir twice daily. A significant two-way interaction occurred with tipranavir levels being elevated and ritonavir levels reduced by coadministration. Tipranavir/ritonavir twice daily at 600/100 for example increased tipranavir Cmax 5-fold and Cmin 9-fold compared to TPV alone (1,350mg BID). The combination was generally well tolerated and further studies of this combination in HIV-infected subjects are underway although the chosen dosages were not reported.

TIDBID: Preliminary 24 week data on full cohort

The TIDBID study is a randomised, open-label study comparing saquinavir-sgc (SQV- and three times daily regimens or in combination with nelfinavir (NFV) and one NA twice daily. Both na夫e and nucleoside reverse transcriptase inhibitor experienced subjects were enrolled and for the experienced subjects nucleoside analogues new nucleoside analogues were initiated along with PI at baseline. The three regimens under study are SQV-sgc 1200mg TID + 2NA's, SQV-sgc 1600mg BID + 2NA's, or SQV-sgc 1200mg BID + NFV 1250mg BID + 1 NA. Data on the full cohort at 24 weeks were presented at ICAAC by Cal Cohen of the Community Research Initiative (6).

838 subjects had completed 24 weeks of study 213 of whom were NA experienced (25%). These 838 subjects baseline demographics revealed a mean HIV RNA of 4.7 log copies/mL and mean CD4 count of 312 cells/mm3. At 24 weeks percentages of subjects with plasma HIV RNA below the limit of quantification was similar across the arms ranging from 51 - 58% for a cut off of <400 copies/mL and 40 - 42% for <50 copies/mL (both Intent to Treat analysis). CD4 rises were also comparable across the arms at 143 - 166 cells/mm3. Results were also similar at 24 weeks for those entering with viral load either above or below 30,000 copies/mL. An on-treatment presentation of results at 48 weeks for 268 subjects showed these virological data maintained. However, no intent to treat analysis at 48 weeks was available for comparison. Diarrhoea was the most common adverse event at 12-15% in the single PI arms and 23% in the SQV/NFV arm. In a virology substudy the saquinavir related resistance mutation L90M was detected in 1/29 subjects experiencing virological relapse before the 24 weeks. The 48V mutation was not detected in any of these 29 subjects. The 184V reverse transcriptase gene mutation was the most frequent genotypic change associated with virologic failure prior to week 24 in this study.

References 1. Youle M. Lipid profiles in patients on ritonavir/indinavir containing salvage regimens. 39th ICAAC, September 26-29, 1999. Abstract 1290. 2. Burger DM. Dose-finding study of a once daily indinavir/ritonavir regimen in healthy volunteers. 39th ICAAC, September 26-29, 1999. Abstract 321. 3. Saah A. Multiple dose pharmacokinetics and tolerability of indinavir and ritonavir combinations in a once daily regimen in healthy volunteers (Merck 089). 39th ICAAC, September 26-29, 1999. Abstract 329. 4. Saag MS. Saquinavir systemic exposure and safety of once daily administration of FortovaseTM in combination with low dose ritonavir. 39th ICAAC, September 26-29, 1999. Abstract 330. 5. Baldwin JR. Pharmacokinetic interaction between the HIV protease inhibitors tipranavir and ritonavir. 39th ICAAC, September 26-29, 1999. Abstract 657. 6. Cohen C. TIDBID study: FortovaseTM (FTV) TID regimen compared to FTV BID or FTV+NFV BID regimens in HIV-1 infected patients. 39th ICAAC, September 26-29, 1999. Abstract 508.

Although once daily dosing may offer advantages in certain populations (i.e. Directly observed therapy), there is little evidence that it would otherwise hold any advantages for adherence over twice-daily dosing. Unless the pharmacokinetics are especially forgiving, missing a once-daily dose (no drug for 48 hours) may have more adverse impact than lapsing once on a twice-daily dose (no drug for 24 hours).

Return to Contents

Resistance and Resistance Testing

Simon Collins, ATP.

In an interactive session looking at controversies in HIV management, Charles Boucher looked at the role of resistance tests for patients changing therapy because of a lack of complete suppression of viral replication or by a rise in viral RNA after a period of undetectability.

The 'interactive' nature of this presentation showed that over 50% of the audience (made up of predominantly HIV-treating physicians) regularly used resistance testing in the above situation. It was also interesting to see that the main reason that the tests were not used routinely by other doctors was linked to funding and cost (42%) rather than because of difficulties of interpretation (15%) or because the benefit had not been proven (17%).

It was highlighted that two prospective studies using genotyping (the Viradapt study and the GART study) have shown that the HIV-RNA response to a change in therapy based on genotypic together with expert advice is better than a switch based on standards of care and treatment history. Dr Boucher illustrated this benefit to members of the audience by using a case study with three separate possible resistance test results to demonstrate how knowledge of results from genotype testing would change up to 85% of their prescription advice compared to their recommendations before having genotypic test results. Given the role played by costs in accessing resistance assays, he also provided an algorithm for using this technology in practice which recommended using genotype testing for cases of PEP, primary infection, failing treatments and pregnancy. Where the results of a genotype test where inconclusive, the use of a following phenotype assay was recommended for a more accurate prediction of the success of a subsequent regimen.[1]

The added value of phenotyping

The predictive use of phenotypic drug susceptibility testing and sustained virological suppression in experienced patients changing treatment, was addressed in a late-breaker presentation by Michael Saag from the University of Alabama. [2]

Patients were drawn from an ongoing, prospective cohort study of ART-experienced patients changing treatment who had plasma collected at baseline that was tested retrospectively for phenotypic drug susceptibility (ViroLogic PhenoSense(TM) assay). Two different definitions of drug sensitivity were used: IC50 (2.5 or (4.0 times the drug-sensitive reference IC50. Successful viral suppression was defined as (0.5 log reduction from baseline viral load at all follow-up visits. Results were presented for seventy-one patients (85% male, 63% white, 76% homosexual) with median duration of follow-up of 8 months. At enrolment, median viral load was 70,644 copies/ml RNA, median CD4 count was 142/mm3, and median number of prior ART regimens was 5. 72% were PI experienced and 98% used a PI in their subsequent regimen. 90% patients were NNRTI naive. Univariate proportional hazards analyses indicated a significant association between time to virologic failure and:

Log of baseline viral load, baseline CD4, number of past regimens, PI use in the new regimen, and number of medications in the new regimen were not found to be significantly associated with time to virologic failure. Multivariate analyses revealed that the number of drugs in the new regimen to which baseline virus was sensitive (using a susceptibility cut-off of either (2.5 or (4.0) was the best independent factor associated with time to virologic failure (RR .59 per susceptible drug in treatment [.46, .77]). ART history did not provide additional predictive value when added to the model and only had a predictive value when phenotypic susceptibility was removed from the equation.

Dr Saag concluded that sustained virologic suppression is best predicted by phenotypic susceptibility to drugs in the new regimen and that this suggested a role for phenotypic testing in suggested in selecting salvage regimens for ART experienced patients.

The greater value of determining phenotypic resistance to protease inhibitors over genotypic mutations was also shown in a study of 32 anti-retroviral experienced patients using an open-label 5-drug rescue therapy of RTV (100mg bid), SQV (1000mg bid), efavirenz (600mg QD) and recycled RTIs.[3] Patients exhibiting phenotypic resistance to SQV at baseline in this study experienced a median decrease in HIV RNA of 0.82 log copies/ml at week 36 compared to a 1.52 log copies/ml decrease in those exhibiting sensitive viral strains (p=0.01). 42 % of patients exhibiting phenotypic resistance to SQV at baseline experienced a viral load below 500 copies/ml at week 36 of therapy, as compared with 84 % in those exhibiting sensitive viral strains (p=0.02). Genotypic resistance to SQV was not predictive of virologic failure.

PhenoSense - an accurate, fast-turnaround phenotypic assay?

One of the arguments for the use of genotype over phenotype testing (apart from cost) is the longer turnaround time for phenotype tests which could allow substantial mutations to continue to develop between drawing a blood sample and receiving the results in order to chose treatments for a subsequent regimen. Two abstracts looked at the reproducibility and validation of a new phenotypic test (PhenoSense, Virologic) which works on a single viral replication cycle which is performed in 8-10 days and which is potentially sensitive at 500 copies/mL. [4,5]

Over 1500 plasma, serum, or virus samples were tested against 12-15 anti-retroviral drugs. Data analysis was based on IC50 and fold change in drug susceptibility compared to a drug-sensitive reference virus. Results were presented from sensitivity experiments with 249 samples from a virus dilution series, and with 189 consecutive patient samples, and showed >90% successful amplification of the PR/RT coding region at 500 HIV RNA copies/mL. The assay provided accurate drug susceptibility results, consistent with published values, for wild-type virus and multiple viruses with well-characterised resistance mutations. Virus drug susceptibility in 150 treatment naive patients was <2-fold the drug sensitive reference IC50 in 90% patients for all drugs except DLV (90% < 3.6 fold) and NFV (90% < 2.5 fold). Repeated testing of patient samples (27 assays per sample) by multiple operators over multiple weeks demonstrated <2.5 fold interassay and interoperator variability for all drugs. Virus mixing studies successfully detected minor populations of drug resistant viruses at concentrations as low as 10-20% of the total virus population. Specificity studies demonstrated no false positive results from sero-negative plasma samples, and interfering substances such as haemoglobin, triglycerides, bilirubin, anticoagulants, and micro-organisms did not significantly affect assay results.

Given the unreliability of genotypic assays for patients with non-clade B virus noted in a separate presentation [6] it is also important to note that HIV strains from clades A-F were successfully tested in this phenotypic assay.

References: 1. Boucher C. The Value of Resistance Assays When Changing Therapy. 39th ICAAC, September 26-29, 1999. Abstract 1368. 2. Saag M. Predictive Value of HIV Phenotypic Susceptibility Testing in a Clinical Cohort. 39th ICAAC, September 26-29, 1999. Abstract LB-17. 3. Haubrich H. Reproducibility of an HIV Phenotype Resistance Assay in Clinical Practice. 39th ICAAC, September 26-29, 1999. Abstract 417. 4. Piketty C. Phenotypic Resistance to Protease Inhibitors Predicts Outcome of a Five Drug Combination Including Ritonavir, Saquinavir and Efavirenz in Patients Who Failed on HAART. 39th ICAAC, September 26-29, 1999. Abstract 2068. 5. Hellmann N. Validation of the Performance Characteristics of a Novel, Rapid Phenotypic Drug Susceptibility Assay, PhenoSense(tm) HIV. 39th ICAAC, September 26-29, 1999. Abstract 418. 6. Harris J. Incorrect Determination of HIV-1 Genotypic Drug Resistance by Hybridization-Based Sequencing Methods. 39th ICAAC, September 26-29, 1999. Abstract 1174.

ZDV/d4T cross resistance?

The following ICAAC presentation looked at cross-resistance between RTIs.

L. Ross and colleagues from Glaxo Wellcome presented similar data to their study previously presented at the Resistance Meeting in San Diego which looked at resistance mutations in 56 d4T-experienced, ZDV naive patients from two studies of NRTI therapy: NZT 4005 and 40012.

In study 4005, patients had been on mono/double NRTIs, primarily d4T/3TC. In study 40012, patients had been on d4T-based therapies, were ZDV naive, but could be receiving other NRTIs/NNRTIs. Plasma HIV-1 RNA and HIV-1 RT baseline genotypes were obtained from a total of 56 ZDV naive, d4T experienced patients in the two studies.

Despite the absence of prior ZDV therapy, 24/56 (43%) of patients had HIV-1 with mutations that have been associated with ZDV resistance (M41L, D67N, K70R, 210W, T215Y/F or K219Q/E). 14/56 patient isolates (25%) had 2 or more of these mutations, with the T215Y/F mutation observed most frequently (14/56 or 25%). The M41L and K70R mutations were observed in 10/56 (18%) of the subjects, L210W in 9/56 (16%), D67N in 5/56 (9%), and K219Q/E in 2/56 (4%). The V75T mutation, associated with in vitro resistance to d4T, was present in one isolate, along with K70N and M184V mutations. The multidrug resistance mutation Q151M was also observed in one patient in combination with the M184V mutation only.

Phenotypic resistance data was obtained for 22/56 patient isolates. These were all sensitive to d4T and ZDV with the exception of the sample with the multidrug resistance Q151M mutation which was phenotypically resistant to both drugs. The authors concluded that 'ZDV -like' resistance mutations can be selected by d4T combination therapy in ZDV -naive adults. The lack of correlation between genotypic changes and phenotype was unexpected. The clinical significance of these mutations requires continued investigation. Steven Deeks in his commentary of this data in his HIV InSite conference summary emphasised that the incidence of these mutations in d4T-treated patients remains uncommon, and probably only occurs after long-term virologic failure.

Ref: Ross L. d4T-Based Combination Therapy Selects for "ZDV-Like" HIV-1 Resistance Mutations in ZDV-Naive Adult Patients. 39th ICAAC, September 26-29, 1999. Abstract 429.

The ability of d4T to produce the typical ZDV genotypic changes associated with ZDV resistance but with no phenotypic resistance to d4T seems rather strange. This phenomenon clearly requires further investigation as to it's impact on response to treatment.

Return to Contents

Therapeutic Drug Monitoring (TDM) of PIs and optimal response in naive and experienced patients

Simon Collins, ATP.

As with last years ICAAC keynote opening speech summarising the best ways to optimise treatment success (in 1998 by Stefano Vella and this year by John Mellors), sub-optimal drug levels were included along with inadequate drug potency, adherence, toxicity and resistance as key factors in failure to achieve or maintain HIV-1 RNA suppression <50 copies/ml. (It was, incidentally, important to hear that viral escape produced at levels above 50 copies/ml means that eventual progression is inevitable and that the burden of proof that it is otherwise should rest with those who argue against maximal suppression). Two further studies at this years conference looked at the direct link between drug concentration levels achieved and efficacy.

An independent NIAID study from University of Minnesota by C.V. Fletcher was particularly interesting. [1] This randomised study showed both an increased proportion of patients reaching undetectable levels and decreased time to achieving undetectable HIV RNA in patients who received individualised treatment adjustment to ensure they achieved levels within the defined therapeutic range.

All subjects were antiretroviral naive and received standard therapy with ZDV/3TC/IDV for the first 4 weeks of treatment. An intensive pharmacokinetic study was performed at week 2 and subjects were then randomised to receive controlled (C) or remain on standard (S) therapy. Beginning at wk 4, C subjects received an individualised regimen of all 3 drugs designed to meet plasma concentration targets. S recipients continued to receive standard doses. Plasma HIV RNA was measured (Roche, Ultrasensitive) at baseline, week 2, and every 4 weeks thereafter. The characteristics of viral decay were determined assuming a biexponential model. 24-wk viral load data from all subjects were analysed simultaneously using non-linear mixed effects modelling (NONMEM). The event time of undetectable HIV RNA was defined as the number of days to the first undetectable level, which was confirmed 4 weeks later by a second test. Results were available from 24 subjects who had completed 6 months of therapy (11 C, 13 S). Median baseline HIV RNA in C and S groups was 4.56 and 4.42 log10, respectively (P>0.5). For all subjects, the average half-life of the rapid phase of HIV RNA decline was 3 days the slow phase was 22.8 days. 10 of 11 C (91%) vs. 9 of 13 S subjects (69%) achieved undetectable HIV RNA. Kaplan-Meier estimates of the median times to undetectable HIV RNA were 110 days for C and 176 days for S recipients (P=0.056, Mantel-Cox).The authors concluded that preliminary observation supports the hypothesis that interpatient differences in antiretroviral drug concentrations contribute to heterogeneity in viral suppression.

The second study presented on TDM was from the Viradapt study (and had also been previously presented at the resistance meeting in June). [2] Viradapt had shown a virological benefit of using genotypic resistance tests to guide treatment choice for second-line and salvage combinations and in an attempt to determine the importance of drug levels on therapeutic success or failure, drug levels to all four protease inhibitors used in the study were performed.

85 patients in both arms (49 genotypic, 36 control) had 575 protease inhibitor drug levels performed by HPLC. The two groups were comparable in terms of risk factor, age, sex, previous treatment, CD4 count, mean RNA at baseline and PI plasma concentration. Patients were considered to have optimal concentrations (OC) drug levels if 2 or more measurements were greater than the IC50 of the specific protease inhibitor they were receiving. Patients in whom 2 or greater levels were below the IC50 were considered sub-optimal concentration (SOC).

Patients who had SOC showed a 0.36 log reduction in viral load at week 48 compared with 1.23 log in the patients who achieved OC. When patients were divided in to four groups based on both usage of genotyping and drug level the following week 48 viral load reductions (log copies/mL) from baseline were seen:

no genotype genotype
sub-optimum concentration (SOC) -0.27 ( 0.29 -0.835 ( 0.41
optimum concentration (OC) -0.92 ( 0.28 -1.33 ( 0.19

In a multivariate analysis the following factors were found to independently effect viral load response:

Baseline viral load was found to only have borderline significance RR=1.79 (95% CI 0.07-1.12, p=0.07).

The authors concluded that in this drug experienced patient population the use of genotypic guided changes and the presence of sufficient protease inhibitor drug levels both contributed to favourable virological responses at week 48.

References: 1. Fletcher CV. Viral Dynamics of Concentration-Targeted (C) vs. Standard-Dose (S) Therapy with Zidovudine (ZDV), Lamivudine (3TC), and Indinavir (IDV). 39th ICAAC, September 26-29, 1999. Abstract 322. 2. Garraffo R. Independent Benefit of Sufficient Drug Levels and Genotypic Analysis in Salvage Therapy: Pharmacological Data of the Viradapt Study. 39th ICAAC, September 26-29, 1999. Abstract 1166.

Studies like the Viradapt approach of determining thresholds for in vivo concentrations are crucial. So far all interpretation of in vivo serum levels are based on in vitro concentrations from cell culture assays. The simple transfer of drug levels from in vitro to in vivo remains to be validated.

39th ICAAC Reports (Part 2)

will be published in DocFax 77 next week. Including reports on:
  • NNRTI's
  • Simplification strategies
  • Immunotherapy
  • Stopping HAART
  • Drug toxicity's
  • HIV/HCV coinfection

Return to Contents



9 November 1999
Royal College of Physicians, London




CME ACCREDITATION 51/2 hours Contact ATP for further information or registration.

Return to Contents