Antiretroviral Update: Applying Emerging Trial Data to Clinical Practice in the Management of HIV

At a symposium held in conjunction with the American Academy of Physician Assistants’ 32nd Annual Physician Assistant Conference, two leaders in the use of antiretroviral medications presented new information describing emerging data concerning antiretroviral combinations that may result in suboptimal virologic outcomes, standard nucleoside reverse transcriptase inhibitor (NRTI), non-nucleoside reverse transcriptase inhibitor (NNRTI), and protease inhibitor (PI) regimens in treatment-naive HIV-infected patients and reporting recommendations concerning the timing of initial therapy
and the preferred/alternative regimens for the treatment of HIV.

Speakers
Daniel Bowers, MD
Partner
Pacific Oaks Medical Group
Beverly Hills, California

Susan LeLacheur, MPH, PA-C
Physician Assistant
George Washington University
Washington, DC

This program was supported by an unrestricted educational grant from Bristol-Myers Squibb Company.

Emerging Clinical Trial Data: Evidence of Early Virologic Failure with Triple-Nucleoside Containing

There are four primary goals of therapy in patients with Human Immunodeficiency Virus (HIV). They include maximal and long-lasting suppression of viral replication, maintenance or restoration of immunological function, improved quality of life, and reduction of disease-related mortality.

“We start with a drug regimen that is going to influence all of these criteria,” said Daniel Bowers, MD, Partner, Pacific Oaks Medical Group, Beverly Hills, CA. “Important considerations include potency, toxicity, tolerability, and convenience of therapeutic choices. The most important variables may be the characteristics of the particular patient.” (Table 1)

Patient factors include the ability to adhere to the regimen and whether the patient has a drug-resistant strain of the virus.
If all does not go well, either there will never be viral suppression or there is viral rebound following initial suppression. Failure is often caused by suboptimal adherence to treatment, emergence of resistant mutations, or inadequate medication potency because of factors such as poor absorption or drug-to-drug interactions. Even with the best of regimens and adherence, no medication routine will be perfect.

Inadequate potency is not always related to the drugs themselves. Some medications for other problems, such as proton pump inhibitors, may interfere with the effectiveness of antiretroviral medications. In other cases, the body may actively pump drugs from the intracellular space.

The first three-drug regimens were protease inhibitor (PI) based. However, the use of these medications also brought problems. There was a high pill burden and various food/water requirements making adherence difficult. There were tolerability issues, especially with liquid ritonavir. Lipodystrophy and other similar concerns became prominent as more people were living, and taking the medications, longer.

One of the first studies to look at a NNRTI-based regimen was the DMP 006 by Tashima, et al. This was an open-label, multicenter trial. It had three arms comparing efavirenz plus lamivudine and zidovudine, one that looked at efavirenz and indinavir, and indinavir with the lamivudine/zidovudine combination (Staszewski S, et al. NEJM.1999;341: 1865).

When they reviewed the viral load (VL) data at 48 weeks, the percentage of patients with HIV RNA levels <400 copies/mL, according to as-treated analysis (ATA), showed that the efavirenz/atazanavir/lamivudine arm was more efficacious than the regimen of indinavir/atazanavir/lamivudine at all time points and achieved statistically significant differences at weeks 2, 4, 8,16, and 48. The regimen of indinavir/atazanavir/ lamivudine performed significantly better than the efavirenz/indinavir arm at week 36. A similar trend was seen for those with HIV RNA <50 copies/mL in the as-treated analysis.

In the more rigorous intent-to-treat (ITT) analysis, the efavirenz/atazanavir/ lamivudine arm performed significantly better than the indinavir/atazanavir/ lamivudine arm at all time points with response rates of 70% and 48% at week 48 for the efavirenz and indinavir arms, respectively. Similarly, in those with HVI RNA levels <50 copies/mL, the ITT analysis indicated that the efavirenz/ atazanavir/lamivudine arm was significantly more effective at all time points.

When response rates (defined as HIV RNA < 50 copies) were analyzed based on HIV RNA strata at baseline, results showed that efavirenz/atazanavir/ lamivudine arm was significantly more effective than the indinavir arm for subjects with HIV RNA levels above 100,000 copies/mL at baseline. Signif-icant increases from baseline in CD4 cell counts were found in all three treatment arms.

Triple nucleoside regimens were tried next. In a trial comparing abacavir versus indinavir when used in combination with the zidovudine/lamivudine backbone, 51% of both groups reached the main endpoint of a VL < 400 copies. When they looked at patients with HIV RNA levels < 50 copies at 48 weeks, there was no statistical significance between the arms either as-treated or ITT. (Staszweski S, et al. JAMA. 2001; 285:1155).

“We were finding naive patients with viral loads of 100,000 copies and T-cells of 50 responding to highly active retroviral therapy,” said Dr. Bowers. “That was when the triple NRTI regimens began to move up on the horizon of treatment options.”

Further studies have now reassessed the long-term potency of triple NRTI regimens.

Gulick and others undertook a study to determine safety and tolerability of three regimens (abacavir/lamivudine/ zidovudine versus lamivudine/zidovudine/efavirenz versus abacavir/lamivudine/zidovudine/efavirenz) and investigate the non-inferiority of regimens based on virologic failure rates (2 consecutive HIV RNA tests, 200 or more copies per mL at week 16 or later). (Gulick RM, et al. 2nd IAS. Paris, France. July 2003. Oral presentation and Abstract 41).

An interim analysis at week 16 indicated that patients receiving only abacavir/lamivudine/zidovudine had significantly greater risk of virologic failure than those in either of the efavirenz arms. Of the 167 patients with virologic failure, almost twice as many occurred in the abacavir/lamivudine/zidovudine arms than in the efavirenz arms (p < 0.001).

“There was a high failure rate in the triple NRTI arm, even among those who at one time had viral loads suppressed to 200,” said Dr. Bowers. “No matter which way you cut the data, there was a statistically significant difference in the efavirenz arms versus the three NRTI arms.”

Nassar, et al. confirmed that efavirenz-based treatments seem better than three-NRTI regimens in a retrospective study using the Parkland HIV database. They found that a regimen of lamivudine/zidovudine + efavirenz was significantly better than abacavir/lamivudine/zidovudine combination at viral levels both under and above 100,000 copies/mL (Nassar N, et al. 2nd IAS, Paris France, 2003 Abstract 554). (Figure 1)

The next step in the evolution of antiretroviral therapy was to see if sequencing the medications impacted on outcomes. The Clinically Significant Long-term Antiretroviral Sequential Sequencing Study (CLASS) was conducted at 43 sites. They enrolled 291 HIV-positive patients with an initial VL >5,000 copies/mL and followed them for 48 weeks. There was an abacavir and lamivudine backbone with a third medication from one of three classes—a NNRTI (efavirenz), the NRTI stavudine, or a PI combination (amprenavir/ritonavir)—added.

At week 48, 76% of the NRTI arm achieved viral loads <50 copies compared to 59% for PI arm and 62% for the NRTI on an ITT analysis (p=0.47). The week 96 analysis is pending. (Bartlett JA, et al. XIV International AIDS Confer-ence. Barcelona, Spain. 2002. Abstract TuOrB1189).

Gallant and colleagues had planned to look at a lamivudine/abacavir backbone combined with either efavirenz or tenofovir. As an open-label study, the investigators began to notice poor efficacy in the tenofovir arm and completed an unplanned, early interim analysis. At eight weeks, almost 50% (50/102) of the patients in the tenofovir arm were non-responders as compared to 5% (5/92) in the efavirenz. The 12-week results were similar.

When they did genotyping on the 36 non-responders, 13 had M184V and 23 had M184V and K65R together. There are currently no explanations for this phenomenon (Gallant J, et al. 43rd Interscience Conference on Antimicrobial Agents and Chemotherapy. September 2003, Chicago, IL. Poster H-1722a).

“The triple NRTI combinations seems to be more convenient,” said Dr. Bowers. “But there is some evidence they may have a lower potency. This will mean balancing what you think will be best for your patient with what they are going to be able to take.”

Recommendations for the Initial Timing and Treatment of HIV-Infected Patients

“For those of you in primary care, clinical practice is easier, believe it or not,” said Susan LeLacheur, MPH, PA-C, Physician Assistant at George Washington University, Washington, DC. “Guidelines allow you to easily apply the data from the research to your practice.”

One of the first considerations is when to begin treatment. Aggressive treatment with large doses of medication lead to toxicities. Waiting until there is damage and then treating is not productive either.

Early treatment does have some advantages. It is easier to get viral counts down and protect the immune system. From a public health viewpoint, low viral load means less likelihood of transmission. The disadvantages include having to take medications every day, various toxicities and- since it is not often that the virus is completely suppressed- resistance (DHHS, Guidelines for the Use of Antiretroviral Agents in HIV-Infected Adults and Adolescents. http://aidsinfo. nih.gov). (Table 1)

The guidelines suggest that everyone with a CD4 count of 200 cells/mm3 and below or who are symptomatic should start treatment. Those with CD4 levels between 200 and 325 should be considered for treatment according to the guidelines, but this is controversial.

“The bottom line is that treatment depends on the patient,” she said. “Medications should be started before the CD4 count drops below 200 because the treatments do not work as well below that threshold and the danger of illness dramatically increases.” One approach would be to use two NRTIs such as zidovudine or didanosine with a non-nucleoside. Efavirenz is generally preferred when using a NNRTI. The NRTIs might be zidovudine, didanosine or stavudine with either lamivudine or emtricitabine used along with efavirenz.

“The decision on which medications to use first focuses on the NRTI,” said Ms. LeLacheur.

Zidovudine’s adverse impact on anemia means that it is probably not a first-line medication in patients who are anemic. Stavudine may cause neuropathies and is strongly implicated in lipodystrophy and other metabolic complications. Emtricitabine is similar to lamivudine and they should not be used together. However, emtricitabine may be a good alternative choice because of its longer half-life.

In choosing an NNRTI, efavirenz should never be used in women who are pregnant or might become pregnant. Nevirapine has a good safety profile in pregnancy and it may help prevent mother-to-child viral transmission, but it does potentially cause hepatic toxicity in women with higher CD4 counts.

Staszweksi, et al. looked at 600 naive patients who were placed on a efavirenz/ lam-ivudine regimen and randomized to either tenofovir+ a stavudine placebo or stavudine+ a tenofovir placebo and followed for 96 weeks. In this cohort, the tenofovir regimen worked equally well with the stavudine. Thus, tenofovir’s inclusion as a preferred option (Staszweski S, et al. 10th CROI, 2003;Boston, MA. Abstract 564b).

The COMBINE study used a zidovudine/lamivudine backbone to compare nevirapine with nelfinavir. The results indicated that the two medications were roughly equivalent. The results were, not however, statistically significant due largely to a small sample size (Podzamczer D, et al. IAS Conference on HIV Pathology and Treatment. 2001. Abstract 7).

The side effect profile may be the deciding consideration when beginning treatment. Both efavirenz and nevirapine can cause rashes, tend to select for class resistance (K103N) with failure, and both can increase PI catabolism complicating therapy. Efavirenz can have central nervous system (CNS) effects, especially vivid dreaming, that are short-lived. Nevirapine can have liver toxicities.

“When you can’t completely get rid of a bug, you’re going to end up with resistance and we cannot completely get rid of this bug,” said Ms. LeLacheur. “Once you lose the ability for either of these drugs to work, you have lost the whole class.”

VanLeth and others looked at two dosages of nevirapine (once-a-day 400 mg or twice-a-day 200 mg dose) versus efavirenz alone and nevirapine+efavirenz. They found that using the two drugs together resulted in higher toxicities. The once-daily 400 mg dose increased hepatic toxicity. This study was underpowered to address efavirenz vs. nevirapine comparisons (van Leth F, et al. Lancet. 2004;363:1253).

Clifford and others studied the CNS concerns of efavirenz. Initial problems with vivid dreams, sleep problems, and the digit-symbol test for eye-hand coordination had all resolved by week 24 (Clifford D, et al. 2nd IAS. 2003;Paris, France Abstract 54).

“There are some caveats to remember when prescribing nevirapine,” noted Ms. LeLacheur. “Women with CD4 counts above 250 cells/mm3 have a greater risk of hepatotoxicity. The greatest risk of potentially fatal events comes in the first six-weeks of treatment. Patients should be monitored closely for 18 weeks as the injury may continue after the drug is stopped.”

The other option for treatment-naive patients is a PI-containing regimen. One of the more widely studied combinations is lopinavir with ritonavir, also known as “boosted lopinavir”.

“Although there are others ways to boost, it is suggested that primary care practitioners stick with ritonavir,” said Ms. LeLacheur. “Boosted lopinavir is recommended for use with lamivudine plus either zidovudine or stavudine.”

There are some new PIs and most of the research compares them to nelfinavir. Fosamprenavir is comparable to nelfinavir overall. However, when you look at viral load, patients with higher loads at the start did better on the newer medication (Rodriguez-French A, et al. JAIDS. 2004;35:22).

Boosted lopinavir also has durability advantages when compared with nelfinavir. The proportion of those taking lopinavir/ritonavir and a nelfinavir placebo with a HIV RNA < 50 c/mL was higher at 96 weeks (p=.001) than those taking nelfinavir with a lopinavir/ritonavir-placebo (Walmsley S, et al. NEJM. 2002;346:2039).

The patient is the best guide when deciding between using a PI and a NNRTI. PIs have more GI effects and more lipid problems. Patients with diabetes or a family history of heart disease may be better suited to efavirenz or nevirapine. For those with a psychiatric history, the concerns about vivid dreams should probably be considered prior to prescribing efavirenz. Those who are taking many medications for other reasons may not be good candidates for PI treatment because of high levels of drug interactions.

“It does seem from the research that it is better to change from a non-nucleoside to a PI-containing regimen than the other way around,” said Ms. LeLaucheur.

Another option for consideration when PIs or NNRTIs are not appropriate is triple nucleoside regimens. The only combinations considered to be viable are abacavir+ lamivudine + zidovudine or abacavir+ lamivudine + stavudine “Unless using zidovudine in pregnancy to prevent infection of the fetus and not to treat the mother, never let your patients out the door with only one medication,” stressed Ms. LeLaucheur. “There are a large number of other drugs and combinations that the guidelines say should not be considered. Familiarize yourself with them before prescribing.”

Case Studies
John came to Dr. Bowers’ practice in 1996 with a VL of 20,000 and a CD4 count of 362 cells/mm3. They initially decided not to treat. He was stable until September of 2003 when his viral load spiked to 322,000, although his CD4 count was stable at 387.

When seen a month later, his CD4 count had dropped to 89 and VL to 123,170. By November he had developed Kaposi’s sarcoma (KS), had rheumatoid arthritis (RA), and was taking hydroxychloroquine and celecoxib. He had expressed some depressive symptoms, but was not on medication.

“I can’t explain what happened because there were no other changes in his health and he was in a monogamous relationship with no history to suggest exposure to a new strain,” said Dr. Bowers. “So we had to stop following and begin to think about which regimen to start him on. I like to use two NRTIs as a backbone and add either a PI or NNRTI.”

Dr. Bowers was concerned about a NNRTI-containing regimen initially given his high viral load. The sudden explosion of KS increased the desire for greatest potency possible. He settled on atazanavir boosted by adding ritonavir with NRTIs tenofovir and atazanavir.

“You get four HIV physicians on a stage you will get six opinions on starter medications,” said Dr. Bowers. “I chose atazanavir/ritonavir because of once-a-day dosing, which is very important to our patients.”

A month later, the VL was down below 1,000 and T-cell percentage had increased from 9% to 20%. By 12 weeks, the viral load had dropped to 70.

This improvement was short-lived and by March his T-cells had drifted back down to 13% and the VL was 189. At this time, the doctor ordered phenotyping, which came back as wild-type.

The next month, his CD-4 count was 179 and VL 208. His KS was disappearing, the RA has resolved and he felt great. However, the gold standard is that VL should be <50 at six months.

“One option is to intensify the regimen,” said Dr. Bowers. “I am going to add didanosine because of its once-a-day dosing and use an interaction with tenofovir to see if that will help.”

JA is a patient of Ms. LeLacheur. He is newly diagnosed with HIV, has an undocumented history of thrush, a current CD4 count of 275, and VL of 125,000. He has a family history of cardiovascular problems and diabetes and is a smoker.

He is in the range where the decision to treat or not is unsettled. He is concerned about his test results and wants to start treatment immediately.

“He has a strong history of CVD and diabetes in his family and a highly variable schedule,” said Ms. LeLacheur. “So I gave him a combination of zidovudine and lamivudine plus efavirenz. This allowed us to avoid metabolic and lipid problems and maintain twice-a-day dosing.”

He had an urgent care visit with another practitioner a couple of weeks later, was found to have anemia and was sent home with iron and a vitamin. When he next saw Ms. LeLacheur, he was taken off zidovudine for possible medication-induced anemia. In its place, he was given stavudine with lamivudine and efavirenz.

She saw him again shortly thereafter and labs were back to normal. This time, he mentioned during a review of systems that he had some tingling in his toes. A new vibratory sensation test was compared with one from the initial assessment. There was a demonstrable change in sensation triggering a change in medications.

“We put him on once-daily tenofovir, switched him to a once-a-day dosing of lamivudine and continued the efavirenz,” said Ms. LeLacheur. “He got very nervous about switching medicine so frequently having heard that your first regimen is usually the best. I explained that this was still his first regimen since the drugs were working and toxicities were the reason for the changes.”