September 22, 2008
Listen to Audio (57 min.)
Please note: These files can be quite large. Allow some time for them to download.
Table of Contents
Note: This CME/CE activity expired on Sept. 22, 2009. For a list of currently available activities, click here.
The agenda for today's program includes the following:
U.S. DHHS recommendations for constructing an initial antiretroviral regimen. Click here to download the full guidelines (PDF).
With respect to the timing of the initiation of antiretroviral therapy, the new DHHS guidelines recommend initiating therapy in patients who have a history of AIDS-defining illnesses and those with CD4+ cell counts of less than 350 cells/mm3.1 Treatment should also be initiated under specific circumstances, such as in all pregnant women, people with HIV-associated nephropathy and patients with hepatitis B coinfection who require treatment for their hepatitis B.
In patients who have CD4+ cell counts greater than 350 cells/mm3, the benefits and risks of therapy -- including comorbidities, patient readiness, adherence issues, drug-drug interactions and other factors -- should be discussed with them before making the decision regarding whether to treat or delay the initiation of therapy.
The DHHS guidelines committee has developed recommendations for preferred initial antiretroviral combinations. The guidelines are used by selecting one of two different agents within column A and then a fixed nucleoside reverse transcriptase inhibitor [NRTI] combination in column B.
Among the preferred non-nucleoside reverse transcriptase inhibitor [NNRTI] agents in column A, efavirenz [EFV, Sustiva, Stocrin] is included. Remember that efavirenz should be used cautiously in a woman with childbearing potential who cannot reliably use birth control and that it should not be used during the first trimester in pregnant women.
Nevirapine [NVP, Viramune] is an alternative agent in the NNRTI category. It should be used cautiously depending upon a patient's CD4+ cell count because of an increased risk of the development of a hypersensitivity reaction to that agent [i.e., rash or hepatoxicity].
Among the protease inhibitors [PIs], the guidelines committee recommends boosted protease inhibitors, including:
with the two latter agents recommended as twice-a-day therapy rather than as once-a-day therapy.
In column B, initial therapy is recommended with fixed-dose abacavir/lamivudine [ABC/3TC, Epzicom, Kivexa] for patients who are HLA-B*5701 negative. Patients who are HLA-B*5701 positive are at risk for the hypersensitivity reaction associated with abacavir [ABC, Ziagen]. The second recommended dual-NRTI is fixed-dose tenofovir/emtricitabine [TDF/FTC, Truvada].
I want to discuss two recent studies that have been important in influencing antiretroviral treatment guidelines. The first is ACTG [AIDS Clinical Trials Group] 5142.2 This is a strategic trial in which patients who are antiretroviral-naive to therapy, but have baseline CD4+ cell counts of any value and viral loads of greater than 2,000 copies/mL, are randomized to one of three treatment arms:
Randomization was stratified by a baseline viral load of less than or greater than 100,000 copies/mL, the presence of hepatitis B or hepatitis C coinfection, and the nucleosides that are selected in the regimen.
Patients' nucleosides were selected by physicians and investigators. All patients in the nucleoside-receiving treatment arms of the study received lamivudine [3TC, Epivir] either administered twice a day or once a day as one of their nucleosides. The other nucleoside was either zidovudine [AZT, Retrovir]; stavudine-XR [d4T-XR, Zerit-XR], which is a long-acting form of stavudine that is not available in the United States; or tenofovir [TDF, Viread]. Patients continued on their original nucleoside assignment throughout the entire course of the study.
The slide above shows the proportion of patients with viral loads of less than 50 copies/mL after 96 weeks on therapy. You can see that the greatest response rate is in the two nukes [nucleosides] plus efavirenz group. Patients on that arm achieved viral loads of less than 50 copies/mL sooner than those in the two nukes plus lopinavir/ritonavir arm, or the nuke-sparing arm.
At the time this study was initiated, the feeling by many providers was that a boosted protease inhibitor was the most potent form of therapy and that it should be the basis for therapy in patients who start with low CD4+ cell counts or high viral loads. This study demonstrated that, in fact, two nukes plus efavirenz is potentially more potent than two nukes plus lopinavir/ritonavir.
ACTG 5202 is the other trial that has influenced antiretroviral treatment guidelines.3 This is a trial that compares different initial antiretroviral combinations in naive patients; compares two different nucleoside combinations, abacavir/3TC and tenofovir/FTC; and also compares an NNRTI-based strategy with a boosted protease inhibitor-based strategy, so that patients are also assigned either to efavirenz or boosted atazanavir.
There are four treatment arms to the trial with patients getting abacavir/3TC and either efavirenz or boosted atazanavir or tenofovir/FTC and either efavirenz and boosted atazanavir. The study is partially blinded so that subjects and providers are unaware of the nucleoside treatment assignment, but know whether patients get efavirenz or boosted atazanavir.
This study is ongoing but, following a DSMB [Data Safety Monitoring Board] review earlier this year, the patients whose baseline viral loads were greater than 100,000 copies/mL were unblinded because there was evidence that patients who were assigned to abacavir/3TC had a higher risk of virologic failure than patients who got tenofovir/FTC.4
At the recent International AIDS Conference in Mexico City, we saw the percentages of patients who experienced virologic failure.3 The virologic failure rate in the abacavir/3TC group was 14% compared to only 7% in the patients who got tenofovir/FTC.
There were actually two different definitions of virologic failure in the study. Virologic failure was defined as having a viral load of greater than 1,000 copies/mL between week 16 and week 24 or having a confirmed viral load of greater than 200 copies/mL after week 24.
There were approximately twice the number of virologic failures in the abacavir/3TC group compared to the tenofovir/FTC group with either definition of virologic failure. So there were more patients with viral loads of greater than 1,000 copies/mL after four months of treatment and there were a higher number of patients with viral loads of greater than 200 copies/mL after 24 weeks of therapy. That included a greater virologic failure rate in patients who had a viral load of less than 200 copies/mL and experienced a rebound in viral load of greater than 200 copies/mL as well as patients who did not suppress down below 200 copies/mL after 24 weeks of therapy.
There was no difference in virologic failure rates comparing the patients who got efavirenz and the patients who got boosted atazanavir, and there were no differences in virologic failure rates for patients whose baseline viral loads were less than 100,000 copies/mL. The study continues in that subset of the population.
We don't know, at this point, what proportion of patients came into the trial with baseline resistance and we don't know what the baseline viral loads were in the patients who experienced virologic failure. I think, at this moment, it's too soon to conclude that abacavir/3TC is less potent than tenofovir/FTC and, in fact, there were data from the abacavir clinical trials presented in Mexico City comparing virologic outcomes in patients with higher and lower viral loads.
In pharma-sponsored studies there does not seem to be a greater risk of virologic failure in patients who get abacavir/3TC with a baseline viral load of greater than 100,000 copies/mL versus those who start on therapy with a viral load of less than 100,000 copies/mL.5
I think that the 5202 story is still not entirely complete. We need a little bit more data before we can absolutely conclude that abacavir/3TC is less potent than tenofovir/FTC in patients whose baseline viral loads are high.
It's important to remember that the initial treatment selection is part of an overall antiretroviral sequencing strategy. We need to consider what therapeutic options we have if patients fail on an initial combination. So we need to think about how to use the nucleosides sequentially. We need to think about how we may switch from an NNRTI-based regimen to a boosted protease inhibitor-based regimen. If we initiate therapy on a boosted protease inhibitor regimen, we need to think how we may want to incorporate a non-nucleoside reverse transcriptase inhibitor into future antiretroviral combinations.
Now let's discuss the use of resistance testing and how we need to incorporate that into our initial treatment decision. All of the major guideline-issuing organizations -- the IAS-USA [International AIDS Society-USA] group,6 the U.S. DHHS guidelines committee1 and the European guidelines committee7 -- recommend baseline resistance testing prior to the initiation of antiretroviral therapy in patients with primary or acute infection.
For patients with chronic infection who are antiretroviral naive, there are different strengths of recommendations by each of the groups. But each of the groups suggests that initial baseline resistance testing should be considered. All of the groups agree that in regions where baseline drug resistance has a prevalence of 5% or greater, resistance testing should be done.
Unfortunately, in most urban centers in the United States and in many rural areas, the prevalence of baseline resistance is 5% or more.8
There are two important issues to remember when ordering a resistance test before initiating antiretroviral therapy. First of all, a genotypic test is sufficient in this clinical situation. If the pattern of resistance comes back with multiple protease inhibitor-associated mutations, then a phenotypic resistance test may also be desirable prior to the initiation of therapy. But in general, a genotypic resistance test will give you all of the information you need to make an initial decision.
The second point is that the resistance tests should be part of the initial evaluation of a patient. Don't wait to order the resistance test until after a patient's CD4+ cell count has fallen and you're ready to initiate therapy. The reason for that is that the virus may revert back to wild type and you won't be able to identify the presence of any resistance-associated mutations that were prevalent when the patient initially came into care.9
Let's review several studies that report on the prevalence of baseline resistance in antiretroviral-naive patients. The first study includes two different data sets that have been developed by the U.S. Centers for Disease Control and Prevention [CDC].8,10 It looks at the prevalence of resistance to the non-nucleosides, nucleosides and protease inhibitors, as well as multiple drug resistance, from 1998 through 2006. You can see in the slide above that over this period of time, the prevalence of resistance has increased such that people who came into care more recently have about a 10% risk of having baseline resistance.
Resistance to the non-nucleoside reverse transcriptase inhibitors has increased over time and now surpasses the rate of resistance to the nucleoside reverse transcriptase inhibitors.
In contrast, resistance to the protease inhibitors, as well as multiple drug resistance, is relatively uncommon.
The above slide from Lisa Ross et al also looks at the prevalence of resistance over a period of time from 2000 to 2005 and includes data from an antiretroviral therapy development program of a pharmaceutical manufacturer and includes data from patients participating in antiretroviral-naive clinical trials.11 You can see in the slide that in 2005, the prevalence of resistance to any drug was as high as 18%. Again, non-nucleoside reverse transcriptase inhibitor resistance is most common. Prevalence of resistance to protease inhibitors was about 2% to 4% over this five-year time interval.
Some epidemiological studies focusing on specific geographic areas occasionally report a higher prevalence of resistance. Anita Shet et al in New York City evaluated the pattern of resistance in patients with acute or recent HIV infection.12 Her study included 112 individuals. You can see from the slide below that in 2003 to 2004 -- the most recent data -- approximately 25% of patients had some baseline resistance. Resistance to the nucleoside and the non-nucleoside reverse transcriptase inhibitors were both common, occurring in about 15% of patients and, alarmingly, the presence of multiple drug resistance was approximately 10% in this cohort.
There are some studies, however, that suggest that the transmission of drug resistance, or the prevalence of drug resistance, in antiretroviral-naive patients may be declining.
The slide below contains data from the U.K. Collaborative Group on HIV Drug Resistance et al. It looks at the prevalence of resistance in antiretroviral-naive patients, both those acutely infected and those with chronic disease, from 1997 through 2005.13 You can see that the presence of resistance peaked in the year 2002, and has since been declining.
But again, resistance to the non-nucleoside reverse transcriptase inhibitors, looking at the above panel on the right in this same slide, has been increasing over time and has plateaued. In contrast, resistance to the nucleoside reverse transcriptase inhibitors and protease inhibitors has declined.
It's important to remember that patients with established disease may have detectable resistant virus even though they're not on therapy. One of the rules of ordering resistance tests for patients who are on therapy is to try to obtain the test while patients are still taking their failing regimen. But when patients are infected with resistant virus, the presence of that resistant virus may persist for years.
In the slide below you can view data from several different studies that examined patients with acute HIV infection who had baseline resistance to the non-nucleoside reverse transcriptase inhibitors, the nucleosides or the protease inhibitors; in a handful of cases, patients had the presence of multiple drug resistant virus.14-17 Patients were followed prospectively to see how long it took before the virus reverted back to wild type.
In only one of these cases -- which include patients who have been followed for over five years -- did the resistant virus revert to wild type. Even though patients may have never been on therapy and were infected a number of years ago, they still may have resistant virus that's detectable.
Baseline resistance is really important in predicting who's going to fail on therapy. The slide below contains data from one particular study known as the FTC-301 trial, which evaluated emtricitabine [FTC, Emtriva] in antiretroviral-naive patients.18
Patients in this prospective, international trial were randomized to receive:
As you can see from the above slide, the proportion of patients with virologic failure who come into the study with wild-type virus in the didanosine plus emtricitabine group was 4%. In the didanosine plus stavudine group it was 11%.
In contrast, patients who were enrolled in the study who had a K103N mutation had a virologic failure rate that varied between 43% and 71% depending upon the original treatment assignment. Baseline resistance testing was not performed in this trial, so these were patients who were entering the study with efavirenz resistance.
Patients who come into the study with nucleoside resistance, in general, also have a higher virologic failure rate than do patients who come into the study with wild-type virus.
These data are among the most convincing to demonstrate that patients who are most likely to experience virologic failure are those for whom the selection of an initial combination includes a drug to which the patient's virus is resistant.
Another prospective study, this one from Mark Oette et al, looked at outcomes in relation to the presence of a pattern of baseline resistance and a particular selection of the initial combination.19 In this case, all treatment assignments were made on the basis of the knowledge of the baseline resistance test.
In contrast to the FTC-301 trial, these are the results of a study done at multiple centers, in which they evaluated the treatment response in patients with baseline resistance or wild-type virus and in which the initial combination was made on the basis of the initial resistance test. Approximately 11% of the patients came into the trial with baseline resistance, mostly to the nucleoside reverse transcriptase inhibitors.
In this study, providers selected agents to which the virus was fully susceptible and there was no difference in treatment outcome at week 48 in an intent-to-treat analysis or in an on-treatment analysis comparing patients who came into the study with baseline resistance or no resistance at all. All patients received multiple drugs to which their virus was susceptible.
Some people wonder whether baseline resistance testing is cost effective. Studies have been conducted looking at just that. Paul Sax et al examined the cost-benefit of baseline resistance testing in a simulation model.20 There are a couple of assumptions that are built into the model. In this case it was assumed that baseline resistance would be present in approximately 8.3% of patients. If resistance is higher, then the baseline cost-effectiveness of initial resistance testing becomes less expensive.
Sax found that the cost per each quality-of-life year preserved is approximately $20,000 for baseline resistance testing following antiretroviral failure. It's approximately the same, $23,900, for baseline resistance testing prior to the initiation of antiretroviral therapy. That's quite consistent with the cost per quality-of-life year saved by antiretroviral therapy alone.
There's another form of baseline resistance that people have been discussing, and that is minority species. The resistance tests that we have available to us clinically are only able to detect the presence of mutations if those mutations are present in 20% or more of the population.21 There are now more sensitive resistance assays that are available in the context of research protocols.
Data from Roger Paredes et al from the ACTG 5095 study evaluates the impact of minority variance on virologic response rates.22 ACTG 5095 was a phase 3, randomized, double-blind, placebo-controlled comparison of three protease-inhibitor-sparing regimens in treatment-naive HIV-infected patients with HIV RNA of more than 400 copies/mL. The three regimens were:
This slide shows the patients who received the efavirenz-inclusive arms; specifically, we can see the prevalence of virologic failure in the population according to the presence of baseline resistance.
In the most left-hand column of the slide, you can see that by bulk sequencing -- which is the kind of resistance test available to us in clinical practice -- about 5% of patients came into the study with baseline resistance.
However, by more sensitive sequencing methodologies, amongst patients who came into the trial with wild-type virus -- and that's wild-type virus by bulk sequencing -- 4.4% of these patients with wild-type virus were found to actually have low levels of a K103N mutation.
Of these patients, 29.5% had baseline resistance with a Y181C mutation and 6% of these patients had baseline resistance with a K103N mutation and a Y181C mutation. All of these different patterns of resistance are associated with resistance to efavirenz.
Patients who came into the study with a Y181C mutation detectable only by these more sensitive sequencing methods were three times more likely to have virologic failure than patients who came into the study with efavirenz resistance that was detectable by bulk sequencing.
We've just been discussing the prevalence of transmitted resistance. We've also talked about some of the newer concepts concerning rare resistance-associated mutations that may be transmitted, but at a level below which can be detected by the commercially available assays.
These low-frequency resistance mutations may have important implications for virologic response, but unfortunately the methods we need to detect them are only available in the context of studies.
Now what I want to do is switch gears and discuss specific mutations that you may encounter when you get genotypic results back and how to use that information to select an initial combination for a treatment-naive patient.
We're going to start by looking at resistance to the reverse transcriptase inhibitors and, specifically, the nucleoside reverse transcriptase inhibitors.23
The International AIDS Society-USA puts out a blueprint that lists specific mutations associated with resistance to agents. This is a nice reference to review when trying to figure out which mutations may inhibit the activity of certain agents.
The most common mutation we'll see transmitted is the M184V mutation. The M184V mutation is associated with resistance to lamivudine and emtricitabine, and in the presence of other mutations, can confer some resistance to abacavir.
We also may see resistance to abacavir and didanosine conferred by the L74V mutation. In rare cases, we'll see a K65R mutation that confers resistance to a number of agents, including abacavir, didanosine, lamivudine, emtricitabine and tenofovir.
The thymidine analog mutations include six specific mutations at codons 41, 67, 70, 210, 215 and 219. These mutations confer resistance to stavudine and zidovudine when present alone. When multiple mutations are present, they confer broader cross-resistance, and I'll specifically mention which mutations confer resistance to which agents in a second.
It's important to consider which mutations preserve certain nucleoside options. HIV with the M184V mutation is still susceptible to abacavir if that M184V mutation is present alone; the virus is also susceptible to didanosine, stavudine, tenofovir and zidovudine in the presence of the M184V mutation.
The L74V mutation, like I said, confers resistance to abacavir and didanosine, but lamivudine, emtricitabine, stavudine, tenofovir and zidovudine are all still active.
The only nucleoside options available, when a K65R mutation is present, are stavudine and zidovudine. Because these are both thymidine analogs, you cannot use these two drugs together. So when the K65R mutation is present, the only active nuke that can be used is one of these two thymidine analogs. To find other agents that are active in the presence of the K65R mutation, you need to go outside the nucleoside class.
When a single thymidine analog mutation -- and that's one of those six mutations I mentioned a moment ago at codons 41, 67, 70, 210, 215 or 219 -- is present, zidovudine and stavudine are less active, but all of the other nucleosides are active and are options.
When two specific thymidine analog mutations -- 41 and 210 -- are present, abacavir cannot be used. Tenofovir is not effective if the 215 mutation is present. It may be possible that none of the other nucleosides will be effective. Thus, in the presence of three thymidine analog mutations that include 41 or 210, you will probably want to go outside the nucleoside class. However, if the thymidine analogs include 67 and 70 and don't include 41 and 210, then didanosine, lamivudine, emtricitabine and tenofovir will still be active.
So how do you construct a potent combination in the presence of nucleoside reverse transcriptase inhibitor resistance? We can turn to some antiretroviral therapy clinical trial data to give us some clues about what combinations are going to be effective and how they may be active in comparison to other combinations you may use in a treatment-naive patient with resistance to nucleosides.
I'm going to return to the ACTG 5142 study to discuss how to construct an initial treatment regimen for patients with nucleoside resistance.2 Remember that the most potent arm in this study was the two nukes plus efavirenz arm. In ACTG 5142, there was also a nucleoside-sparing arm consisting of efavirenz plus lopinavir/ritonavir. That arm performed almost as well as the two nukes plus efavirenz arm. So if a patient who comes into care already has broad nucleoside resistance, one option is a nucleoside-sparing combination of efavirenz plus lopinavir/ritonavir.
Patients who receive two nukes plus lopinavir/ritonavir did well also. If you can select two active nucleosides and pair them with a boosted protease inhibitor, you can also construct a potent combination.
Let's consider transmitted non-nucleoside reverse transcriptase inhibitor resistance. The most common mutation we see transmitted is the K103N mutation. That mutation confers resistance to the first-generation non-nucleoside reverse transcriptase inhibitors.
Etravirine [TMC125, Intelence], a second-generation non-nucleoside reverse transcriptase inhibitor, maintains full activity in the presence of the K103N mutation. With the first-generation of NNRTIs, we typically think that one mutation confers resistance to the entire class, but etravirine is more like a nucleoside reverse transcriptase inhibitor or a protease inhibitor in that each mutation confers some step-wise resistance. So etravirine is still an option in patients who have just a K103N or Y181C mutation, and few additional other mutations associated with resistance to etravirine.
Again, when thinking about how to put together an effective combination in patients with NNRTI resistance at baseline, we need to think about the use of a boosted protease inhibitor to form the backbone of this initial combination.
The MONARK study looked at the response of patients who got lopinavir/ritonavir alone compared to patients who received lopinavir/ritonavir, zidovudine and lamivudine.24
Among patients who received just the boosted protease inhibitor in an on-treatment analysis, 84% achieved viral loads that were less than 50 copies/mL at one year of therapy.
Among patients in that study who received lopinavir/ritonavir together with the nucleoside combination of zidovudine and lamivudine, 98% of those patients achieved a viral load of less than 50 copies/mL at week 48.
People will do better if they get a boosted protease inhibitor plus some nucleosides that are active, but most of the activity is being conferred by the boosted protease inhibitor. Thus, a patient who comes into study with resistance to just the non-nucleoside reverse transcriptase inhibitors can get a boosted protease inhibitor plus nucleosides and do very well.
One question becomes: Is there a difference among boosted protease inhibitors when the virus is fully susceptible to a boosted protease inhibitor? The answer is probably not.
There are a number of studies now that compared different boosted protease inhibitors in antiretroviral-naive patients together with nucleosides. The CASTLE study compared boosted atazanavir and lopinavir/ritonavir.25 The KLEAN study compared boosted fosamprenavir and lopinavir/ritonavir.26 The GEMINI study compared lopinavir/ritonavir and boosted saquinavir [SQV, Invirase].27 The ARTEMIS study compared lopinavir/ritonavir administered as either a twice-a-day therapy or once-a-day therapy versus boosted darunavir [TMC114, Prezista].28
In all of these studies, there was little difference when comparing outcomes -- the proportion of patients with viral loads that were less than 50 copies/mL after a year on therapy -- between these different boosted-protease-inhibitor regimens. The only exception was that patients who got lopinavir/ritonavir 800 mg/200 mg once a day did not do as well as those who received boosted darunavir.
In the ARTEMIS study, I should mention that the boosted darunavir dose is 800 mg of darunavir plus 100 mg of ritonavir [RTV, Norvir], a different dose than is used in treatment-experienced patients.28
There's not that much of a difference between different boosted-protease-inhibitor options in patients whose virus is fully susceptible. However, there may be a concern about using boosted protease inhibitors in patients whose viral loads are high.
Again, looking at different studies, there is little difference in outcome in patients who get one boosted protease inhibitor versus another if their baseline viral loads are less than 100,000 copies/mL compared to those patients whose viral loads are greater than 100,000 copies/mL.
Again, the only caution here is that patients with high baseline viral loads should not be given lopinavir/ritonavir once a day because there are a number of studies now that suggest that patients with high baseline viral loads will not do as well if they get lopinavir/ritonavir administered once a day compared to lopinavir/ritonavir twice a day.
I want to turn to etravirine again, because there may be a desire to use a second-generation NNRTI in patients who have some baseline nucleoside resistance. It's important to become familiar with the mutations associated with resistance to etravirine because they're not the typical NNRTI-associated mutations that we're used to.
There are 13 mutations that have been found to confer resistance to etravirine that have an impact on virologic response in clinical trials. The clinical trials that have been used to form this data are the DUET studies, which have now been published in Lancet.29,30
The 13 mutations involve different amino acid substitutions at eight different codons. Those codons are 90, 98, 100, 101, 106, 179, 181 and 190. When there are three or more of these mutations present, patients are not as likely to respond to etravirine, so that information is important when trying to determine whether etravirine is an option in patients who may have transmitted nucleoside resistance.
There's also important data from a phase 2 study of etravirine done by Brian Woodfall in treatment-experienced patients.31 The study compared patients with varying nucleoside and documented non-nucleoside reverse transcriptase inhibitor resistance.
Patients were randomized to either etravirine or a protease inhibitor and nucleosides that were chosen by the investigators.
In this study, patients did better if they were on the boosted protease inhibitor compared to patients who were on etravirine. This response occurred not because etravirine was not active, but because response was influenced by the number of nucleoside mutations that were present. When patients had no nucleoside mutations, and they just had NNRTI resistance, those patients had an almost 2-log reduction in viral load. But when one or more nucleoside mutations were present, patients did not respond at all, or as well.
If there is NNRTI resistance, and just the K103N mutation is present, then you can use etravirine together with some active nucleosides, but I would not use etravirine together with nucleosides if there is any nucleoside resistance, just if there's non-nucleoside resistance. In the presence of nucleoside and non-nucleoside reverse transcriptase inhibitor resistance, I would base my combination on a boosted protease inhibitor.
Resistance to the protease inhibitors is the most complicated pattern of resistance. There are some protease inhibitors that will be resistant in the presence of a single mutation. For example, atazanavir will not be effective in the presence of an I50L mutation. There are other protease inhibitors, such as lopinavir/ritonavir or darunavir that will preserve activity in the presence of a single mutation.
Frequently, in the setting of transmitted protease inhibitor resistance, there may be one option in the class that maintains full phenotypic activity. Fortunately, we don't see multiple protease inhibitor mutations transmitted very frequently. We'll often see only a single major mutation, and some secondary mutations. It's rare to see transmitted resistance to multiple protease inhibitors.
In the presence of a few PI mutations, you may want to get a phenotypic resistance test prior to the initiation of therapy if you're going to use a protease inhibitor, in order to determine what the best options in this class are.
To consider which options may be the best in this category, if there is a single mutation or a few mutations that are transmitted, one study to examine is the TITAN trial.32 The TITAN trial compared lopinavir/ritonavir versus boosted darunavir in treatment-experienced patients.
To enroll in this study, patients had to be failing on a stable antiretroviral combination that they had been on for at least 12 weeks or had interrupted therapy for a minimum of four weeks. There were 595 patients enrolled who were randomized to receive lopinavir/ritonavir or boosted darunavir.
It's important to remember that these patients were lopinavir/ritonavir naive, so we're not comparing patients with previous lopinavir/ritonavir experience to patients on a new boosted protease inhibitor that they had never experienced.
About 32% of these patients were PI naive. They had been treated with a different antiretroviral combination, but had never been on a protease inhibitor previously.
About 35% to 40% of patients, depending upon the arm, had received one previous protease inhibitor-inclusive combination.
If you look at treatment outcomes -- the proportion of patients with viral loads of less than 50 copies/mL after one year on therapy -- patients who came into the study with no PI mutations did just as well on lopinavir/ritonavir as did patients on boosted darunavir.
However, if patients had only a single IAS-USA protease inhibitor mutation -- and you can refer to the IAS-USA list of primary PI mutations23 -- they did better if they got boosted darunavir compared to those who got lopinavir/ritonavir. That's true even when the phenotypic fold change to lopinavir/ritonavir was less than 10.
If you look at the phenotype of virus that had only a single mutation to lopinavir [LPV] in the study, there is very little impact on the fold change in the susceptibility of that virus. Single mutations may confer very small incremental fold changes that have important effects on virologic response, so it's important when using boosted protease inhibitors to use the agent that is the most active where there's the lowest fold change in virus susceptibility. That is true for patients who are getting a boosted protease inhibitor with a single mutation or multiple mutations, whether they're PI naive or have extensive PI experience.
It may be that some patients are going to be infected with virus that's resistant to multiple agents in multiple classes. We're going to need to think about using agents in different classes in order to have enough active agents available to construct a potent combination.
Let's think about some of the agents in the new classes. Raltegravir [MK-0518, Isentress] is the first FDA [U.S. Food and Drug Administration]-approved integrase inhibitor. Although we may be most familiar with the BENCHMRK studies looking at raltegravir use in treatment-experienced patients,33,34 there is another study focusing on an antiretroviral-naive population that compared different doses of raltegravir versus efavirenz in combination with two nucleosides.35
The 48-week data of this treatment-naive trial shows no difference in virologic response rates in patients who get two nukes plus raltegravir versus the patients who get two nukes plus efavirenz. In fact, patients achieved a viral load of less than 50 copies/mL at earlier time points in the raltegravir arms at any raltegravir dose compared to the efavirenz arm.
Basically, raltegravir is a potent agent that can be used in combination with other drugs if needed due to the presence of multiple-class resistance.
Another new agent is the first FDA-approved CCR5 inhibitor: maraviroc [MVC, Selzentry, Celsentri]. In a study by Michael Saag et al known as the MERIT study, fixed-dose zidovudine/lamivudine plus maraviroc was compared to fixed-dose zidovudine/lamivudine plus efavirenz in antiretroviral-naive patients who had pure R5 virus (virus that gains entry into cells using the CCR5 coreceptor).36
Patients in this study who received the nucleosides plus efavirenz experienced less virologic failure than did the patients who received two nukes plus maraviroc. Thus, we don't know for sure that maraviroc is as potent as efavirenz together with two nucleosides in an antiretroviral-naive patient.
It's important to note that some patients came into that study with CCR5-using virus at the time of screening, but had X4 [CXCR4] or dual-tropic virus at the time they started on the drug. If we exclude those patients from the analysis, then patients did just as well on two nukes plus maraviroc as they did on two nukes plus efavirenz.
If you're going to consider using maraviroc in a patient, you need to first give him or her a Trofile assay. It is the most sensitive commercially available phenotyping assay we have at our disposal. The Trofile assay distinguishes virus that uses CCR5 and virus that uses CXCR4 for entry. The CCR5 inhibitors will not have any activity against virus that uses CXCR4 or against dual-tropic virus, so it's important to define the tropism of the virus before using a CCR5 inhibitor.
This assay has sensitivity for X4 or dual-tropic virus of 1% or less.37 This is a more sensitive assay than has previously been made available. It is about 10 times more sensitive than the previous assay and it needs to be ordered if you're going to consider using a CCR5 inhibitor. You should only use a CCR5 inhibitor if the virus is pure R5.
Let's try to put some of the data and some of the clinical trial information together to think about how to construct an active combination in the presence of certain patterns of resistance.
Choosing a Regimen in the Presence of NRTI Resistance Alone
In the presence of reverse transcriptase inhibitor resistance, if nucleoside resistance is present alone, then one option is an NNRTI plus a boosted protease inhibitor. A boosted protease inhibitor with active nucleosides is also a reasonable combination. Be cautious about using an NNRTI plus partially active nucleosides. There probably isn't a high enough genetic barrier of resistance to prevent the rapid selection of resistant virus in that case.
Choosing a Regimen in the Presence of NNRTI Resistance Alone
If non-nucleoside reverse transcriptase inhibitor resistance is present alone, then the best option for initial therapy is a boosted protease inhibitor plus two active nucleosides.
Choosing a Regimen in the Presence of PI Resistance Alone
For patients who just have protease inhibitor resistance, then a good initial combination is two nucleosides plus one non-nucleoside reverse transcriptase inhibitor.
As I said before, it's unusual to see primary protease inhibitor resistance transmitted with multiple mutations. If the patient has PI resistance and only a single mutation is present, then you probably have active agents within the class and you will want to choose your most active agent to use as part of that initial combination.
Choosing a Regimen in the Presence of Multiple Class Resistance
If the patient has multiple drug resistance -- multiple class mutations -- and if the virus is fully susceptible to protease inhibitors (that is, if the patient has nuke mutations plus NNRTI mutations, but no PI mutations), then you can use a boosted protease inhibitor plus one or two new agents. The options include: etravirine, which is an option even in the presence of a K103N mutation; maraviroc, if the virus is pure R5; and raltegravir. If the virus is resistant to the protease inhibitors, then you still may have an option in that class; use the boosted protease inhibitor that is phenotypically most active and combine that with two fully active drugs. You're probably going to need to go to new agents and new classes or, in extreme situations, consider the use of enfuvirtide [T-20, Fuzeon].
There may be situations when you're going to need to start a patient on an initial combination, but you're not going to have a resistance test available. One of the situations where this might arise is starting a patient on therapy who presents with an opportunistic infection.
Andy Zolopa, at the most recent retrovirus conference, presented the results of ACTG 5164, which randomized patients who presented with an opportunistic infection, or an AIDS-associated bacterial infection, to either immediate therapy or delayed therapy.38 Immediate therapy was considered treatment within 48 hours after a patient presented with an opportunistic infection and their HIV diagnosis, or was newly diagnosed with HIV, otherwise treatment was delayed for approximately six weeks.
Patients who came into the study had a median CD4+ cell count of 29 cells/mm3 and 70% had a CD4+ cell count of less than 50 cells/mm3. Ninety percent of these patients were antiretroviral naive.
Most of the patients received lopinavir/ritonavir and all received either lamivudine or emtricitabine. This is an important study because it demonstrated that patients who received immediate therapy were less likely to die or have new AIDS-defining complications compared to patients whose treatment was deferred. There were a similar number of patients with immune reconstitution in the study (10 in the immediate treatment group and 13 in the deferred treatment group), although most of the patients in this study had pneumocystis, and many of them received steroids. The real incidence of immune reconstitution may have been somewhat diminished because of the use of steroids in this population.
But this study suggests that we may need to start treatment earlier than we had initially been accustomed to. For that reason, we may not have a resistance test available when initiating some patients on therapy. What recourse do we have in that situation?
What you can do is order a resistance test in the hospital prior to initiating therapy. If you're going to start therapy in the hospital while waiting for the results of the resistance test, start the patient on a boosted protease inhibitor combination, because PI resistance is not as common as nuke or NNRTI resistance. Again, you may want to choose the agent that has the most activity in the presence of some PI resistance, because PI resistance will occur in a handful of cases. When the resistance test comes back, you can then simplify therapy, choosing among the agents that the virus is susceptible to. This is a way you can make an early treatment decision in the absence of a resistance test.
In conclusion, a genotypic resistance test should be obtained before prescribing an antiretroviral combination in every treatment-naive patient. In the presence of mutations, the combination that's prescribed should have at least two active drugs if you are using a boosted protease inhibitor -- that's a boosted protease inhibitor plus one or two active drugs, or preferably three fully active drugs. It has to be three fully active drugs if a non-nucleoside reverse transcriptase inhibitor is part of your initial combination.
Finally, if you're going to start a patient on antiretroviral therapy, in the absence of a resistance test, the safest approach is to use a boosted protease inhibitor, and then you can modify therapy when the results of your resistance test return.
Thanks for participating in this activity. I hope the information that was discussed helps you more effectively manage your patients.
This transcript has been lightly edited for clarity.
The content on this page is free of advertiser influence and was produced by our editorial team. See our content and advertising policies.