link-up with 32 week data report in combination with AZT/3TC

At the Chicago Retrovirus Conference, data from several studies of efavirenz were reported: 60 weeks IDV+EFZ; 24 weeks EFZ+AZT/3TC; EFZ resistance; EFZ CSF concentration in monkeys; potentially key interaction data for NFV+EFV and EFV+141W94.

EFV is metabolized through the same process as other NNRTIs and protease inhibitors--through the P450 liver enzyme system. Part of the p450 system is the CYP3A4. EFV possesses enzyme inducing activity on CYP3A4. Enzyme inducing activity leads to an increase in metabolism. Increasing metabolism results in less drug remaining in your blood; therefore, you may have to raise the dose of the accompanying drug.

For example, EFV reduces IDV blood levels by 35%, so when EFV is used in combination with IDV the normal IDV dose of 800 mg every 8 hrs is increased to 1000 mg every 8 hrs.

In a EFZ+141W94 PK study presented at Chicago meeting, researchers looked at 6 HIV infected individuals (CD4 5-311 cells; bDNA 590-370,100 HIV copies/ml) received 141W94 1200 mg bid for 1 week in combination with 2 or 3 NRTIs (4 individuals received 1592U89). Following collection of samples on day 7, EFV 600 mg once daily was added for 1 week. After which samples were collected again.

The mean 141W94 AUC was decreased 36%, the mean Cmax was decreased 39%, and the mean Cmin was decreased 43%, after the addition of EFV. There was wide variability among 141W94 PK measures in this small subset of patients (CV%=71). For an explanation of Cmin and Cmax see the NATAP pharmacokinetics report.

The investigators said that the clinical effects of the decreased 141 drug levels is unknown although the 141 Cmin achieved with EFV is similar to concentrations shown to be effective using 900 mg bid 141 with NRTIs. The AUC, Cmin, and Cmax values for EFV in combination with 141 were slightly higher than those observed in another study of healthy volunteers after 7 days of EFV 600 mg qd. It remains to be seen if these higher values for EFV are due to differences in intersubject metabolism or whether there is a PK interaction with 141.

The Tmax represents the amount of time it takes to reach the Cmax which is the peak drug level. See PK Primer on web site (LINK-UP) for explanation of PK terminology. As you can see in the following table the 141W94 AUC, Cmax and Cmin are reduced when EFV is added. And, there is a good deal of variability. The number after the ( is the standard deviation and represents the observed variability.

During week 1, individuals received 141W94; during week 2 they received 141W94+EFV. CL/F (L/h) is the amount of clearance of drug as measured by liter per hour. You will note that as AUC declines the rate of clearance increases.

PK Interaction Between EFV and NFV in Healthy Volunteers

WD Fiske and others from DuPont Merck reported the data from this study and concluded that when taken in combination, both drugs should be taken at their regular doses. Dose modification is not necessary.

Investigators reported that coadministration of EFV 600 mg qd with NFV 750 mg tid in this study resulted in an increase of 20%, a statistically significant difference, in NFV AUC compared to NFV taken alone. The NFV Cmax was also increased by about 20%, also a statistically significant difference . The abstract did not specifically report the Cmin change but based upon visual observation of a PK line graph in the abstract the Cmin was higher at hour 8, and it was reported to me that the NFV Cmin increased by 22% in this study. The authors reported there is no significant difference between the effect on EFV drug levels when coadministered with NFV compared to levels of EFV taken alone.

After taking NFV, a hydroxylated metabolite of nelfinavir called AG-1402 or M-8 forms. Two entities are present- NFV and AG-1402. The abstract said that AG-1402 has comparable in vitro antiviral activity to NFV, and that EFV reduced AG-1402 AUC by 37% (statistically significant difference) as compared to NFV taken alone. It was reported to me that:

• EFV effect on NFV may or may not be clinically relevant

• It is not known to what degree M-8 contributes to the overall NFV in vivo antiviral response, if at all.

• There is 3 to 4 times more NFV present than AG-1402, so after factoring in the changes in blood levels for NFV and M-8, the sum of NFV + M-8 in circulation when coadministered with EFV is slightly higher than without EFV. Therefore, the overall clinical response shouldn’t be effected.

 

J Kahn and others reported a 60 week update (of cohort 4) for the drug combination of IDV+EFZ. 101 protease and NNRTI naive individuals were initially randomized to EFZ 200 mg once a day (qd) + IDV 800 mg tid (every 8 hours) or IDV alone. At 12 weeks those taking IDV alone added EFZ+d4T. After about 36 weeks doses of EFZ were increased to 600 mg once a day. In our January ‘98 newsletter, NATAP Reports, the 48 week data was reported with a discussion of dosing, pediatrics plans, and rash. The results observed at week 48 are sustained at week 60. DuPont Merck, the developer of Efavirenz, expects to submit their application (NDA-New Drug Application) to the FDA for accelerated approval in June ‘98. In an early 14 day monotherapy study where the dose was 200 mg the peak reduction in viral load was 1.7 log when using the Roche Amplicor 400 copy test.

In this table you will see reference to viral load of "<1 copy/ml" using the standard Roche Amplicor (<400 copies/ml) test. This was explained in our January ‘98 issue. The explanation is more extensive in the EFV report in the New Drug Section (LINK-UP). If a person’s Amplicor test result was "<1 copy/ml", that means the lab could not detect virus. It does NOT mean virus is not present, just that it wasn’t detectable. The Amplicor test is not as sensitive as the Roche Ultrasensitive test, so study investigators reported that at week 24,71% of participants with "<1 copy/ml" had Ultrasenstive results <40 copies/ml. And also at week 24, 87% of individuals with Ultrasenstive <40 copies/ml had "<1 copy/ml" by standard Amplicor test.

In other words, some individuals with <1 copy/ml had >40 copies/ml. And some individuals with >40 copies/ml had <1 copy/ml. It reflects, among other things, variability in testing results. It’s a little complicated but I hope this explanation is clear enough.

59 patients were initially randomized to EFV+IDV but 54 actually initiated treatment with EFV+IDV. 42 were initially randomized to the IDV arm and 42 actually initiated treatment in that arm. Leading up to week 60, investigators reported the % <400 copies/ml was 74% at week 12, 87% at week 16, and 94% at week 24; and, the % "<1 copy/ml" was 64% at week 12, 71% at week 16, 88% at week 24.

Premature Discontinuations (24 weeks)

One patient was discontinued from EFV due to rash

129 patients’ data from several similar cohorts ( a cohort is a group of study participants) were examined to conduct this analysis. Two factors were found to be significant predictors of treatment success or failure. The first predictor is baseline HIV RNA. Failure is defined as a rebound to above 400 copies/ml after having achieved consecutive readings below 400 copies/ml. Investigators found that for every 1 log increase in baseline viral load there was a 4.3 fold increased risk for treatment failure (rebound to >400 copies/ml). For example, 10,000 copies/ml to 100,000 copies/ml is a 1 log difference.

The second significant predictor in this study was whether or not a person lowered viral load to "<1 copy/ml". Investigators found that if viral load, in this study, was suppressed to between 1 and 400 copies/ml there was a 7.59 greater chance of rebounding to >400 copies/ml than if your viral load was "<1 copy/ml". Therefore, suggesting that lowering viral load to <400 copies/ml may not be as durable as lowering viral load to "1 copy/ml" or <40 copies/ml. In other words, the investigators concluded more durability may be expected from lowering your viral load to a low level. Other studies have also suggested that lowering viral load to <40 copies/ml will create more durability. Many leading authorities and treating physicians have accepted the idea that lowering viral load to <50 copies/ml will be more durable, but we don’t yet have much substantive long term data supporting this thinking.

Additionally, in certain circumstances it is difficult to know exactly what to do with this information. For example, if a person reduces their viral load to 300 copies/ml with a 3 drug regimen consisting of a protease inhibitor plus 2 NRTIs, should they change the regimen or add another drug to further reduce viral load to <50 copies/ml (lower limit of detection using the Ultrasenstive test)? That would be an individual decision, although the Abbott study #462 (ritonavir+saquinavir) suggested that a well timed treatment intensification was successful in lowering viral load to undetectable for those who had not yet reached undetectable with ritonavir+saquinavir alone. For more details on the 462 study see the current Double Protease Report on the NATAP web site(LINK-UP).

Preliminary week 24 results were reported on this non-protease inhibitor regimen.137 treatment naive individuals were randomized to receive AZT/3TC (twice daily) plus one of three doses of efavirenz (200 mg qd, 400 mg qd or 600 mg qd--qd means once a day), or the placebo group who received AZT/3TC alone. At week 16, individuals randomized to AZT+3TC alone added IDV+EFV. 24 weeks is too short a period of time to conclude that EFV+AZT/3TC is an equal substitute for a 3-drug potent protease containing regimen. Data from protease containing 3 drug regimens have shown suppression of viral load out to 1-2 years. But, this study is being closely followed to observe the durability of the data.

Patient demographics at enrollment were: 26/104 (25%) white, 66/104 (63%) African-American,12% other ethnicity, 87/104 (84%) male, 17/104 (16%) female.

Premature Discontinuations (24 weeks)

Changes in CD4 and Viral Load from Baseline

At week 16, several therapy changes were implemented for study participants when it was determined that all the efavirenz dose arms were statistically significant superior to the AZT/3TC arm:

• EFV+IDV was added to AZT+3TC

• The EFV 200mg and 400mg qd doses increased to 600mg qd and if person was not undetectable (400 copies/ml), IDV+d4T (d4T was substituted for AZT) was added

• Those who started treatment with EFV 600mg qd+AZT/3TC added IDV+d4T (d4T was substituted for AZT) if their viral load was not undetectable.

The tables below contains changes at both weeks 16 and 24.

CD4 increases are approximations based upon visual observations of line graph charts. The number of evaluable patients (n) at week 16 ranged from 27 to 30 for all 4 arms. At week 24, the n ranged from 20 to 24 individuals for all 4 arms. Again, at week 16, therapy changes were made for all arms. Individuals in all 4 arms received EFV 600 mg once daily after 16 weeks. So, it’s possible the sudden increases in the 200 and 400 mg arms were due to the improvements in the regimen at week 16. The therapy changes are described in the paragraph above.

The test with a lower limit of detection of 400 copies/ml was used. The peak reduction in viral load for the AZT/3TC group was 1.7 log at week 2. Remember that at week 16 participants made therapy changes. Although maybe not statistically significant, the mean viral load reeducation of the AZT/3TC arm was about 1.9 log at week 20 and decreased to 1.65 at week 24. There were 27-31 evaluable patients at week 16, and 21-24 at week 24. The viral load reductions in the table are approximations based on visual observation of a line graph. Again, at week 16 treatment regimens were improved.

For the individuals initially receiving AZT/3TC alone for 16 weeks, about 45% were <400 copies/ml at week 16; after adding EFV+IDV 65% were <400 copies/ml. This is significantly inferior to the 3 EFV treatment arms.

At week 24:

• 100% in the 600mg qd dose were <400 copies/ml

• 91% in the 400mg qd dose arm were <400 copies/ml

• 96% in the 200mg qd arm were <400 copies/ml.

DuPont Merck officials have said they believe the 600 mg qd dose has better antiviral activity and will create more durable viral load suppression than the other doses. As well, they believe the 600mg qd dose may be more effective suppressing resistant virus in the peripheral blood and CNS. How well the 600mg qd dose regimen can consistently suppress resistant virus remains to be confirmed with data. The K103N mutation is key to EFV resistance. Its presence can create 19 fold resistance to EFV (see Resistance Section below). At week 16, there were 27-31 evaluable patients, and at week 24, there were 21-24 evaluable patients in each of the 4 study arms.

* Patient reported multiple reasons for discontinuation.

NATAP reported 16 week Adverse Events data in a previous report following Hamburg. The same 16 week data was reported in Chicago. Rather than duplicating the effort, you can view the full Adverse Event Report in the other NATAP report on this web site (LINK UP). The main increase in incidence of a side effect between the dose groups occurs in the 600mg qd dose compared to the other two dose regimens with regards to dizziness. At the 600mg dose there is a higher reported incidence of dizziness (44%) as opposed to 29% in the 400mg arm and 19% in the 200mg arm. Company officials say generally that this side effect lasts for only several hours and seems to disappear within a couple of weeks. If this side effect continues to be a concern, it is recommended that you can consider several options--take the 600 mg dose before bedtime, take 300 mg twice per day, or if that’s not tolerable you can lower the dose to 400 mg once daily and you can take it at bedtime. As stated earlier company officials have said they think the 600mg dose has more antiviral activity.

Rash has been seen when treated with EFV but the company says that in most cases it resolves without requiring a change in EFV treatment.

Investigators in the animal CSF study found that EFV penetrates into the CSF of these monkeys. NATAP’s previous efavirenz report which appeared in the January newsletter, NATAP Reports, and on the web site contains data on EFV CSF penetration in 3 humans. The 200 mg dose was used and the CSF/Plasma ratios of EFV drug levels were 1.19%, 1.04% and 0.65% for the 3 persons. See the report for more information. At the Chicago conference, data was reported from this monkey study whose purpose was to determine the CSF concentrations and long term safety of chronic administration of EFV after 57 weeks of oral dosing in cynomolgus monkeys; and, to determine if there is if the CSF drug level is dose dependent. Three groups of monkeys received 15, 45, or 75 mg/kg of EFV for 1 week and then twice daily (8 hours apart) for 56 weeks. After 57 weeks plasma and CSF samples were obtained 3 hours after dosing and 16 hours or later after dosing. If you want to skip to the conclusions and avoid the data see the table below.

Here are a few considerations in interpreting and understanding and the measuring of CSF drug levels. The amount of time after dosing when you take a sample from the CSF can be important in assessing drug levels. Drug levels should be higher when sampling is close after the dosing. As time passes, the drug levels or concentration in the CSF usually decrease as you get closer to the next dose. Or, if the CSF drug level remain the same, the CSF/plasma ratio can increase because the plasma drug level declined as it nears the end of the dosing period.

It appears as though some antiretroviral drugs for treating HIV are less effective in suppressing virus in the CSF then in plasma. As well, it appears as if some drugs penetrate the CSF more effectively than others. Many drugs do not penetrate the CSF at levels which would adequately inhibit HIV. It may be important to keep a certain minimum level of drug concentration in the CSF, as it is necessary in the plasma, throughout the entire dosing period in order to maintain adequate viral suppression. If a treatment does not reach the CSF or brain at adequate levels to suppress HIV, continuing replication and consequent resistance may develop as it would in the blood. It appears crucial that research continue to identify drugs that penetrate the CSF well and to better understand their effect in the CSF as well as in the brain.

The brain may be a privileged or sanctuary site for HIV. Virus replication in the brain may have different characteristics than replication in the peripheral blood or in lymph tissue. Although researchers can measure drug concentrations in the CSF, they are unable to effectively directly assess the effect of therapy in the brain. Research in these areas is getting increasing attention, but currently many questions remained unanswered.

NATAP held a full day community conference on January 17, 1998 at NYU Medical Center where a key speaker was Dr Justin McArthur. He is a leading authority on HIV and the brain. He discussed what we know and what we don’t know about HIV and the brain, and HIV treatment and their effect on the brain. A summary of his talk is available on this web site (LINK UP).

The investigators in this CSF monkey study concluded that EFV in cynomolgus monkeys at clinically relevant plasma levels, penetrates the CSF achieving drug concentrations that may suppress HIV replication. CSF levels appear to approximate the unbound plasma (blood) concentrations of EFV. EFV is highly protein bound (99.46-99.76%) in the blood. Only the non-protein bound drug in plasma is available for inhibition of viral replication. There is no protein in the CSF so a 1% drug concentration in the CSF may be equal in effectiveness to the actual non-protein bound drug level in plasma.

The lower limits of quantification were 3.96 nM for the CSF test and 0.317 uM for the plasma test. The authors stated that plasma concentrations in the 30 and 90 mg/kg/day dose groups were similar to peak (13uM) and trough (5 uM) values observed in humans taking 600 mg once per day. The mean CSF drug level is higher at all 3 dosing levels at the 4 hour sampling time than at the (16 hour sampling time point. As well, the same can be said for the EFV levels in the plasma. Over time the levels go down, but the CSF/plasma ratios are higher at the (16 hour time point than the 4 hour time point.

Mean ( standard deviation. The standard deviation is the amount of variability that can occur from the mean. For example, 34.2 is the mean CSF drug level at 3 hours after dosing 15/mg/kg. This group received 15 mg/kg twice daily.

* The other 3 monkeys in this group had undetectable CSF levels (<4.0 uM)

This section was initially intended to be a review of the EFV resistance reports in Chicago which characterize the resistance profile of EFV. But, I tried to gather a little additional information on DLV and NVP resistance to expand the discussion.

NNRTI cross-resistance is not yet well understood, and additional research is necessary to better understand it. This Resistance section discusses mutations and cross-resistance in more detail than many of you may be interested. If you don’t want to read the entire section, here is a brief summary. The 3 NNRTIs, DLV, NVP and EFV, have resistance mutations in common. The K103N mutation appears commonly associated with resistance to any of the 3 NNRTIs. Appearing less frequently than a K103N mutation, some virus mutants without the K103N appear, which EFV may be able to suppress. A single mutation of Y181C can emerge from NVP or DLV resistance, or a P236L single mutation may emerge from DLV treatment, but EFV was still active against these single mutations in vitro (see below). Conversely, the K103N mutation is highly associated with EFV resistance in vivo, but both NVP and DLV resistance is associated with K103N.

At this point in our knowledge about NNRTI resistance, for the purpose of making treatment decisions now, it may be safe to assume there is at least partial cross-resistance between EFV, DLV and NVP. Treatment decisions should be made prudently.

Resistance to a NNRTI develops more quickly than resistance to a protease inhibitor. A single mutation can cause high level resistance to NVP or DLV. DuPont Merck says high level resistance to EFV may develop more slowly, but high level resistance may not be necessary to cause treatment failure. At this time, it appears as if there’s too little information to know if and how this EFV information will bear out in treatment for individuals.

DuPont Merck reported two studies at Chicago. The in vivo study looked at the circulating plasma virus in patients participating in EFV trials whose viral load rebounded after an initial decline. The most common resistance mutation was the K103N. It is key to the development of EFV resistance. It was observed in 86% to 93% of patients failing EFV combination therapy with either indinavir or AZT/3TC. As well, both delavirdine and nevirapine show loss of activity when the K103N mutation is present. Based on reported in vitro data nevirapine can show a high level loss of activity when this mutation is present. One in vitro study suggested that delavirdine had 7 fold loss of activity when the K103N was present and another study suggested delavirdine developed a high level loss of activity when the K103N was present. Investigators on this DuPont Merck in vivo study said that all the mutations they observed had been seen previously as associated with one or more NNRTIs. Other mutations observed but less frequently than the K103N included Y188L, and G190S. Double mutations observed to occur less frequently than the K103N but more frequently than other double mutants included K103N/V108I, L100I/K103N, K103N/P225H (occurred more frequently) or K103N/G190S. Viruses with double mutations were seen in up to 70% of EFV failures. But it appears as if the K103N is the first mutation to develop.

DuPont Merck also reported results from an in vitro study of recombinant viruses carrying defined mutations. A recombinant virus is one that is artificially created or constructed in the lab to contain certain mutation(s). Resistance by a drug to these recombinant viruses does not necessarily predict responses in humans. However, it appears that some of the in vitro resistance studies for protease inhibitors predicted the cross-resistance we’re experiencing in humans. The in vitro data reported by DuPont Merck (see table below) showed loss of activity (resistance) by efavirenz when the following single and multiple mutations were present in a recombinant virus: S48T/G190S, K103N, L100I, K101E, Y188L, S48T/K103N/G190S. A recombinant virus showed that the K103N mutation was associated with18 fold resistance to EFV. The results of this in vitro study showed that delavirdine and nevirapine also showed reduced activity against almost all of these same mutations (see table below). EFV maintained effectiveness against the single mutations: V106A, Y181C, Y188C, G190A, P236L.

The Y181C mutation can cause NVP resistance, and the P236L mutation can cause DLV resistance. You will note that neither of these two mutations emerged associated with EFV resistance in vivo. Based on in vitro data, it is reasonable to expect that a Y181C resistant virus will remain sensitive to EFV, and it is reasonable to expect a P236L resistant virus will remain sensitive to EFV. But if a double mutant occurs, such as K103N+P236L (which can emerge from DLV) or a Y181C+K103N (which can emerge from NVP), cross resistance may result.

DuPont Merck officials have said they increased the dose of EFV from 200mg once a day to 600mg once daily at least in part because they hoped the increased drug blood levels might suppress virus containing the K103N mutation and other mutations. A major concern in preventing cross-resistance to EFV after prior resistance to DLV or NVP is to be able to suppress the K103N. Individuals may have the K103N mutation from prior NNRTI experience or as pre-existing mutations prior to any NNRTI therapy. DuPont Merck says the K103N has been suppressed for some individuals. The EFV blood levels some individuals achieve may be adequate to suppress the K103N. But, we have not yet seen that for most individuals the 600 mg qd dose can suppress resistant viruses containing the K103N.

A study of in vitro passage of virus in presence of DLV showed the key mutation due to delavirdine in vitro was P236L. But this mutation may emerge only in a small percentage of individuals with DLV resistance. Additionally, a double mutant of K103N+P236L can emerge. In vivo analysis of patient isolates from clinical trials with DLV monotherapy and combination therapy have shown the K103N mutation to be key. Other mutations identified from in vitro analysis of patient isolates from clinical trials include K103T, Y181C, P236L, K103N+Y181C and K103N+P236L. Upjohn says subjects with the K103N and/or the Y181C mutations are likely to be resistant to DLV and NVP. They reported that combination therapy of DLV+AZT was observed to cause resensitization to AZT. In Chicago, a report by Lisa Demeter and the AACTG 261 team suggests that concomitant therapy with DLV and NRTIs may alter the resistance mutation profile that emerges. For example, DLV+ddI may cause the Y181C to emerge while DLV+AZT prevents the Y181C, but the K103N may emerge. Based on the in vivo resistance data reported for EFV, the K103N appears to be the only mutation in common between EFV and DLV.

Genotypic mutations observed in clinical studies with NVP include L100I, K103N, V106A, V108I, Y181C, Y188C/L, and G190A. The Y181C mutation may be the most common one. In vitro studies have shown high level nevirapine resistance when the Y181C or the K103N mutation was present. Multiple mutants observed in an early clinical study include: Y181C+K103N, Y181C+K103N+V108I, Y181C+K101E+G190A. In this study (1), resistance data on 6/7 individuals showed each of the 6 had a multiple mutation, Y181C was present in 6/7 mutants, and a K103N mutation was present in 2/6. Based on both the in vivo and in vitro EFV resistance data reported, NVP and EFV have several mutations in common: K103N, G190A, L100I, Y188L, K101E. Based on the EFV in vitro study data EFV appears to suppress all but the L100I, the Y188L multiple mutation, and the K103N. The same study reported a maximum decrease of 2 log (range 1.96 to 2.43) after 2 weeks in the monotherapy patients, using the Roche Amplicor test.

The EFV in vitro study reported at Chicago and discussed above indicated EFV may suppress virus with a mutation at Y181C or P236L; Y181C is associated with NVP resistance, and P236L is associated with DLV resistance. But as stated above P236L may be present in a small percentage on individuals with DLV resistance and they may have accompanying additional mutations. The same may be said about Y181C and NVP resistance. This brings me full circle to my original statement at the beginning of this section--at this point in research, it is fair to expect at least partial cross-resistance between NVP, DLV and EFV.

Retrovirus abstract #702. The number in parenthesis is the fold resistance. The number outside parenthesis represents the amount of drug (as measured by nM) required to suppress the mutation. Resistance in this table is characterized as the fold increase in IC90 relative to that of the wild type virus for each drug tested. For example, 38 nM of DLV is required for 90% inhibition of wild type virus (NL4-3), and 1000 nM is required to suppress the K103N mutation. That is a 28 fold increase in resistance.

The authors concluded the resistance profile partially overlaps with that of the other NNRTIs tested, although potency is maintained against a number of NNRTI resistant mutants. But, a problem may occur when the K103N emerges as part of a multiple mutant containing one of these other single mutations.

A mutant of Y188L including additional mutations emerged in 4.5% to 6.7% of the sequenced patients in two clinical studies (n=51). A mutant of G190S plus other mutations emerged in 0% to 13.3%. These two multiple mutations occurred relatively infrequently. While a multiple mutant containing a K103N mutation occurred 53.3% to 68.2%. While, as stated above, overall 86% to 93% of patients in the study acquired a K103N mutation.

In Chicago, R Kaspar and others reported preliminary results from a small open-label non-randomized study of a twice daily regimen of d4T+3TC+nevirapine. 25 treatment naive individuals were reported to elect this regimen. Participants received prednisone for the first 14 days for the purpose of preventing discontinuation from therapy due to the development of rash, which can result from nevirapine therapy. They received a course of prednisone therapy (40-50 mg once a day), which was described to me by Dr Kaspar as a moderate dose used for poison ivy and asthma, a common dose for anti-inflammatory indications. No cases of rash did occur. Participants followed the recommended dose escalation for nevirapine of 200 mg once a day for 14 days followed by 200 mg twice daily. Standard doses of d4T and 3TC were used.

Dr Kaspar said the participants were a pre-selected group; they chose this treatment regimen because they weren’t comfortable with the compliance demands of protease inhibitors. About 25% of participants are IV drug users. Dr Kaspar said although he hasn’t conducted a formal study, he believes compliance has not been good in this study. Some of the patients may periodically stop taking study medications for a day or more before returning to taking them properly. But, some participants are compliant. Kaspar estimated that maybe 3/4 of the study participants may have missed 20% of their doses of study meds. He thinks NVP may be a forgiving drug, in a compliance sense, due to its 20 hour plasma half-life. He said 3TC’s half-life may also lend itself well to this situation. The plasma half-life for 3TC is 5-7 hours. But, perhaps more important is that the 3TC intracellular (triphosphate) half-life is reported to be10.5 to 15.5 hours.

As you can see, there is a wide range of CD4 and viral load at baseline among the participants.

In a telephone discussion with the author, he reported 20/25 were <400 copies/ml after a mean of about 10.5 months. Of the 25, 2 patients have been noncompliant, and 1 was lost to follow up. Investigators are planning to test all participants using the Roche Ultrasensitive test which has a lower limit of detection of <50 copies/ml.

This study was designed to evaluate the influence of NVP on NFV blood levels, and to describe any changes to NVP levels due to concomitant use of d4T or NFV, compared to historical data for NVP pharmacokinetics. The PK (interaction) study is being followed with a 6 month extension to follow participants for safety and antiviral activity.

The investigators found no statistically significant changes in NFV levels after the addition of NVP (AUC +4%, Cmin -4%, Cmax +14%). D4T levels were also unchanged, compared to historical controls, NVP levels were characterized by authors as unchanged.

25 individuals were enrolled in the study; 22 completed the PK study at day 36; 19 individuals completed week 9, and 9 completed week 21. There were 9 discontinuations-

3 due to adverse events (described below), 5 withdrew consent and 1 was lost to follow-up. 4 patients interrupted therapy due to rash (2) and hepatitis (2).

• CD4: mean- 372, median 281, range 35-937

• Viral Load: mean- 4.5 log (about 31,600 copies/ml), median 4.5 log, range 3.2 log to 5.83 log (1580 copies/ml- 676,080 copies/ml)

• 52% were treatment naive

• AZT/3TC: 6 individuals (24%)

• d4T/3TC: 2 (8%), participants were not permitted to have >6 months prior d4T experience

• AZT: 1 (4%)

• AZT/ddI 1 (4%)

• d4T, AZT, ddI: 1 (4%)

• ddI, hydroxyurea: 1 (4%)

D4T (30-40 mg bid, adjusted for weight) and NFV (750 mg tid)were started on day 1, with no washout period. On day 7, PK samples were collected over an 8 hour period. NVP was added on day 8 at 200 mg once a day, increasing to 200 mg bid after 14 days (dose escalation). On day 36, PK samples were collected for all 3 drugs over 12 hours.

The standard dose escalation for initiating NVP therapy was used as described in previous paragraph. It is used to lessen the chance of developing rash associated with NVP therapy. The investigators characterized the differences in AUC and Cmax for NVP between day 7 and day 36 as being unchanged.

At nine weeks there were 19 evaluable study participants. 84% (16/19) were <400 copies/ml (undetectable by Roche Amplicor PCR Test). The median viral load reduction was about 1.6 log. The mean CD4 increase from baseline was about 94.

Investigators reported details of the 3 permanent discontinuations, and 4 persons who had drug interruptions:

Rash
One person developed rash on day 7. Person removed from study, NVP not initiated. Rash resolved by day 16 and d4TNFV was successfully reinstated with ddI without reoccurrence of rash.

Elevated lipase
2nd person developed elevated lipase and GGTP on day 7; d4T was changed to AZT, NFV was continued and NVP started on day 8. Elevated lipase resolved by day 16, but elevations recurred at day 36. No symptoms were associated with lab abnormalities. Viral load was suppressed from 118,608 to below detection at day 36. Subsequently treated with AZT/3TC?NVP and viral load remains below detection.

Hepatitis and Rash
Person developed rash, fever and elevated SGOT/SGPT (LFTs) and bilirubin on day 34. LST elevations resolved within 4 weeks and restarted on d4T/3TC/NFV without reoccurrence of symptoms or lab abnormalities.

Rash
Person developed rash on day 8 within hours of adding NVP to d4T/NFV. Rash resolved within 12 days and nevirapine restarted along with Benadryl, and there was no recurrence of the rash. Person continues in study.

Hepatitis
Person developed hepatitis, nausea, vomiting, diarrhea, and dehydration at week 21, following ingestion of large amounts of alcohol. All 3 drugs were discontinued. LFTs returned towards normal and plan is to restart d4T?NVP?NFV when LFTs are within normal range.

Rash
Person developed circumferential rash around waist/groin and a low grade fever within 12 hours of starting NFV. NFV was withheld and person treated with Benadryl. Rash resolved within 7 days and NFV was restarted, along with Benadryl. There was no recurrence of rash.

Hepatitis
Hepatitis developed at week 13 associated with hep B conversion and alcohol consumption. All study meds stopped, LFTs have declined. Plan is to start d4T?NVP/NFV when LFTs are normal.

The INCAS study was started prior to the establishment that triple drug therapy was the most effective way to suppress viral load. Adherence was not an established principle and ddI was not well tolerated. As result Boerhinger Ingleheim says that the results of this study were effected by these obstacles.

INCAS was a double blind, randomized placebo controlled study. 34 patients who continued NVP based triple therapy after the completion of INCAS were encouraged to participate in this follow-up. Patients were allowed to change their NRTI therapy while continuing NVP. This analysis focuses on the 34 patients in the triple therapy arm who participated in the follow-up. There are some obvious limitations to this analysis: adherence to taking ddI was difficult, this follow-up does not track the original patients in the study but only a portion of them, many treatment failures and obviously dropouts in original INCAS study were not followed, and patients were permitted to change their NRTI therapy.

Median baseline HIV RNA for the 34 patients in the follow-up group was 4.22 log (about 17,000 copies/ml) when they first enrolled in INCAS. Their median baseline CD4 was 369. 44% (15/34) added 3TC during the follow-up period.

Investigators reported that of the 19 patients who were below detection (20 copies/ml)at 1 year only 2 increased above detection .

The authors said the goal of therapy is to maximally suppress viral replication as low as possible for as long as possible. 57 treatment naive and experienced individuals received IDV+NVP with either AZT/3TC or d4T/3TC.

In this study IDV was administered 1000 mg tid, NVP was taken 200 mg bid, AZT 300 mg bid, d4T 40 mg bid, and 3TC 150 mg bid. Ethnic demographics at baseline in naive group were: 6 white, 9 latino, 2 black. In the experienced group they were: 17 white, 18 latino, 4 black.

In the treatment naive group (n=18), the mean baseline CD4 was 342 , and mean viral load 82,140 copies/ml. In the treatment experienced group (n=39), all 39 had previous NRTI experience, 13/39 had previous saquinavir (Invirase) experience of <1 year (they may not have failed saquinavir), mean baseline CD4 was 184, and mean viral load was 169,403 copies/ml after a 13 day washout. The Roche Amplicor test was used to measure viral load and the Ultra Direct was used when individuals were <400 copies/ml.

One treatment naive and three treatment experienced (2 for adverse events) individuals left the study prior to 24 weeks. The author stated that treatment experienced group had used all NRTIs.

There were two serious adverse events requiring changing therapy. One person in the naive and one in the experienced groups experienced nephrolithiasis (kidney stones) which is a side effect from indinavir.

• 2 NVP rashes (grade 2)*
• 2 adverse gastrointestinal effects (bloating, GERD)

• 1 nephrolithiasis (resolved with oral fluid and pain control)
• 4 NVP rashes (grade 2)*

*Rashes managed with, at most, antihistamines, corticosteroids, and a "new" ramping period.

• naive group: no new opportunistic infections, malignancies, or AIDS defining conditions
• experienced group:
    -1 case of reactivation of peripheral CMV (4 weeks after initiation of therapy)
    - no new cases of CMV, MAC, PCP, KS
    -significant improvement in: molluscum contagiosum, KS, 2 cases of lymphoma, 50% decrease in OI prophylaxus

Previously, Dr Julio Montaner reported week 20 data for this study. M Harris reported the update in Chicago. 22 heavily pre-treated individuals with low median baseline CD4 of 30 and high median baseline viral load of 150,000 copies/ml received NVP 200 mg daily for 14 days then 200 mg twice daily (dose escalation), IDV 800 mg every 8 hours, and 3TC 150 mg twice daily. It’s been reported that when IDV and NVP are coadministered, IDV blood levels are reduced-- AUC -28%, Cmax -11%, Cmin 38%. Despite that, the normal IDV dose was used in this study.

19 participants had previously used 3TC, and 15 had a prior AIDS diagnosis. One person had previously used ritonavir, and two persons had previously used loviride (NNRTI).

After 20 weeks, Montaner reported a CD4 increase of 100, a 3 log reduction in viral load from baseline, 55% <500 copies/ml, 20% < 20 copies/ml. The Ultra-sensitive test was used to measure the viral load reduction.

• 4 persons withdrew due to adverse events before 8 weeks- nausea & vomiting (2), rash (1), urinary frequency/nocturia (1)
• 2 for personal reasons unrelated to toxicity or treatment failure
• 5 withdrew due to treatment failure (viral load >5,000 copies/ml)

After 1 year 11 patients remained on NVP/IDV/3TC

• 7 were <20 copies/ml (7/22, 32%)
• 3 were between 20-400 copies/ml
• 1 person had 1635 copies/ml

1 person had grade 3 hyperbilirubinemia which resolved. Another person had neutropenia which resolved (d4T substituted for 3TC)

In January ‘98 NATAP Reports, we reviewed 32 week data for DLV+AZT+3TC. The newsletter is available on the web site. A more extensive detail of the data is available in the New Article section.