Chapter 5
Antiretroviral Therapy

John G. Bartlett, MD

The Principles of Antiretroviral Therapy (ART)   TOP

What are the goals of therapy?

The overall objective of treatment for HIV disease, as with treatment for many other infectious diseases, is to control the putative agent with antimicrobial agents while providing other appropriate therapies for HIV-related complications. The single most important goal of HIV antiretroviral therapy is to reduce the HIV viral load to as low as possible for as long as possible. The 5 other goals of antiretroviral therapy (ART) are to

• Prevent HIV-associated complications
• Avoid the long-term and short-term adverse drug reactions associated with antiretroviral agents
• Prevent HIV transmission
• Avoid HIV resistance
• Preserve HIV treatment options

Does antiretroviral therapy work?

Treatment recommendations have evolved greatly since zidovudine (AZT), a nucleoside, was first tested in 1986. The strategy was revolutionized in 1996-97 with the introduction of protease inhibitors (PIs) and then subsequently non-nucleoside reverse transcriptase inhibitors (NNRTIs). Combined with AZT and other nucleosides (see Table 5-1), the result was potent combination ART (also referred to as highly active antiretroviral therapy, or HAART) that had an immediate and dramatic impact on the prognosis for HIV infection, in fact one of the most impressive changes in any disease since the introduction of penicillin in the 1940s. Within 2 years there was a 60%-80% decrease in mortality, AIDS rates, and hospitalizations for HIV-associated complications. Nevertheless, this treatment was also associated with some disappointments: there is still no cure for HIV infection, many of the patients given ART develop serious side effects, and drug resistance causes many patients to eventually have virologic failure so that long-term benefit may be difficult to sustain. Also, adherence has been shown to be critical to treatment success, and the level of adherence required is among the stiffest for the treatment of any disease in medicine.

When to Start Therapy   TOP

What are the criteria for starting therapy?

The decision to start therapy, like most medical decisions, depends on the risk-to-benefit ratio of treatment. ART is given to control HIV replication and the consequent immune dysfunction, but therapy is also associated with the development of some substantial side effects and the risk of developing resistance that would limit future options. The reduction in CD4 cells is the pivotal event of HIV disease that renders the patient susceptible to the unique opportunistic infections (OIs) and tumors that have come to be known as AIDS-defining diagnoses (see Table 5-2). The patient becomes vulnerable to these diseases when the CD4 cell count decreases from normal levels (500-1500 cells/mm³) to <200 cells/mm³; the association with complications is the rationale for this CD4 cell threshold to be used by CDC to define AIDS.

Virtually all guidelines in the world have reached a consensus that a CD4 cell count of <200/mm³ is an indication for ART. The only question is the CD4 level above which treatment should be started. According to the DHHS Guidelines for the Use of Antiretroviral Agents in HIV-Infected Adults and Adolescents (available at http://www.aidsinfo.nih.gov) the CD4 cell count range of 200-350/mm³ is also an indication for ART, but the decision is tempered by the viral load, the rapidity of the CD4 cell count decline (CD4 cell slope), and patient readiness. The role of viral load in this decision is based on 2 observations: 1) the viral load is an independent predictor of progression and 2) the CD4 decline slope is directly correlated with viral load. The implication is that a high viral load imparts a risk for more rapid progression to AIDS and a more rapid decline in the CD4 cell count. The threshold that appears important in the CD4 cell strata of 200-350 is =20,000 copies of virus/mL (c/mL). Some authorities also treat patients with a viral load of 50,000-100,000 c/mL even when the CD4 cell count is >350/mm³, but there is no consensus here. It should be emphasized that multiple cohort studies involving thousands of patients clearly indicate that therapy is necessary when the CD4 cell count is <200/mm³; some of these studies also show a benefit when the CD4 cell count is 200-350/mm³, but these findings are somewhat inconsistent from study to study.

What patient factors influence this decision?

Patient readiness may be the most important factor in the decision to treat most patients. ART is never an emergency, although the CD4 cell count clearly indicates the magnitude of risk from waiting. Most experienced HIV providers will never start ART on the first patient visit, which is designed for evaluation and the initiation of patient education. The patient must understand not only the risks and benefits of HIV treatment, but also the critical role of adherence to what may be one of the most complicated medical treatment regimens for any disease (see Chapter 7: Adherence to HIV Therapies).

What to Start   TOP

How do you suppress HIV replication?

The goal of treatment is to suppress HIV as much as possible for as long as possible. This requires 3-4 drugs in combinations that have been proven to have merit in clinical trials (see Table 5-1).

Do you ever start with 1 or 2 drugs?

No, with a rare exception; some providers treat pregnant women with a CD4 cell count of >350/mm³ and a viral load of <1,000 c/mL with zidovudine (AZT) alone. With this exception, all patients should be treated with 3-4 drugs combined in a fashion that has established merit.

How do you decide on a specific regimen?

The selection of a regimen is based on the experience with completed trials and idiosyncrasies of the specific patients. No single regimen is appropriate for all patients, but virtually all experienced AIDS treaters have a few favorites. Most start with a 3- or 4-drug regimen that includes 1) an NNRTI combined with 2 nucleosides, 2) a PI either alone or boosted with ritonavir (RTV) plus 2 nucleosides, or 3) a 3-nucleoside regimen that includes abacavir (ABC), lamivudine (3TC), and zidovudine (combined as Trizivir). This last regimen has been popular for patients who struggle with complicated medical regimens because it is only 1 pill twice a day; however, the regimen is considered suboptimal compared with the PI- or NNRTI-based regimens.

Is once-a-day therapy possible?

This is now feasible and clearly preferred by some patients. The following drugs may be given once a day: tenofovir (TDF), lamivudine (3TC), emtricitabine (FTC), didanosine (ddI), stavudine (d4T), efavirenz (EFV), atazanavir (ATV), and amprenavir plus ritonavir (APV/r).

When to Change Therapy   TOP

When do you change therapy because of virologic failure?

The goal of therapy is to reduce the viral load to undetectable levels. The expectation is a reduction by 1 log10 c/mL within 1-4 weeks, a viral load of <500 c/mL at 16-24 weeks, and a viral load of <50 c/mL after 24 weeks. Viral load measurements are subsequently made at 3-month intervals and are expected to remain at <50 c/mL. Temporary increases above this level are sometimes called "blips." Virologic failure is defined as 2 consecutive viral load counts of >500 c/mL (see Table 5-3).

What causes virologic failure?

There are 2 causes: 1) resistance, meaning that the strain of HIV is resistant to one or more of the drugs in the regimen, and 2) failure of the drug to reach the virus, which may be due to failure of adherence, drug interactions, or drug malabsorption. The most common cause is lack of adequate adherence.

What do you do when the patient has side effects or intolerance to the regimen?

Side effects of antiretroviral agents are common and highly variable in severity and implications. Nausea is particularly common and may become a huge impediment to adherence. Other side effects may be more serious in terms of specific medical consequences, including several that are potentially lethal: pancreatitis due to didanosine (ddI), hypersensitivity due to abacavir (ABC), Stevens-Johnson syndrome due to an NNRTI, nevirapine-associated hepatic necrosis, and lactic acidosis due to nucleosides, especially stavudine (d4T). If the side effect precludes adherence or is regarded as a serious consequence for adherence, the regimen must be changed. It is important to warn patients about common and serious side effects before treatment begins.

Are there other forms of treatment failure?

Immunologic failure, as opposed to virologic failure, is characterized by the failure of the CD4 count to rebound, and clinical failure is characterized by AIDS-defining events (see Table 5-4). Success in treatment is based on viral load response, and this dictates therapy decisions. No one knows what to do with immunologic or clinical failure if there is viral suppression.

Table 5-4. Definitions of Therapeutic Failure

Type	Definition
Virologic failure	Initial therapy: failure to decrease viral load by 1 log10 c/mL by 1-4 weeks, to <500 c/mL by 8-16 weeks, or to <50 c/mL by 16-24 weeks. Chronic therapy: viral load >400 c/mL with 2 measurements after achieving viral suppression.
Immunologic failure	Failure to increase CD4 to >25-50/mm3/year (this is an arbitrary definition; the average increase with viral suppression by the above definition is an increase of 100-150 c/mm3/year).
Clinical failure	HIV-related complication, usually an AIDS-defining condition after antiretroviral therapy for 3 months.

What to Change to   TOP

How do you address intolerance or serious side effects?

Assuming virologic control has been achieved, the goal is to find a substitute for the offending agent. In most cases, this is simply a single drug substitution using an agent in the same class with comparable potency. For example, for zidovudine (AZT) intolerance, stavudine (d4T) is commonly substituted. For indinavir-associated nephrolithiasis, the usual tactic is to reduce the dose, increase fluid intake, or substitute another PI for indinavir (IDV).

What is intensification?

Intensification is the addition of a drug, usually a single drug, to enhance antiretroviral activity in a patient who has a good but suboptimal response to therapy. This needs to be done at a time when the viral load has not increased to too high a level; for example, most authorities consider a viral load of >5,000 c/mL to be an indication for a completely new regimen. Common ploys for intensification are the addition of tenofovir (TDF) or abacavir (ABC), or a boosting of a PI with ritonavir (RTV) if that was not done initially.

How do you deal with virologic failure?

The usual method is to measure HIV resistance. If the strain is sensitive, it is important to closely examine the adherence issue or look for drug interactions or some other reason that the antiretroviral agent does not reach the virus. If the virus is resistant, this information is used to select the next regimen as described below.

Resistance Testing   TOP

What are the tests?

There are 2 types of resistance tests: genotypic and phenotypic. Genotypic resistance measures mutations to the genes that are the target of antiretroviral drugs. These mutational changes predict resistance. Phenotypic resistance measures the activity of various drugs against the patient's virus compared with "wild type" virus (untreated HIV strains). These tests are quite different in method of performance, but there is no consensus regarding their relative merit. Many people use genotypic resistance testing most frequently because it is faster and cheaper. Under any circumstances, interpreting both tests requires substantial expertise. It should be emphasized that the testing is valid only for the drugs that are being given at the time the test is done; drugs discontinued >2-6 weeks earlier may not be reliably tested because the resistant strains often become "minority species" when drug pressure is removed. Thus, a history of prolonged ART and virologic failure in the past often indicates resistance even if this is not revealed with the current test.

What are the indications and requirements for resistance testing?

The major indication for resistance testing is virologic failure, but testing is also sometimes done prior to initiation of therapy (see Table 5-5). The requirement for testing after virologic failure is an adequate viral load (usually >1,000 c/mL) to do the test. For patients who have failed therapy it is standard practice to sample while the drugs of interest are being given. Resistance to drugs discontinued >2-6 weeks before testing may not be expressed in the resistance testing, although this probably differs by individual patient and specific agent and has not been extensively studied.

How should you treat someone who has run out of options?

Extensive use of ART over several years combined with virologic failure and resistance leads to limited options for many patients. However, it appears that antiretroviral drugs are still beneficial in these patients because of reduced "fitness" of the virus. The implication is that the antiretroviral drugs force development and persistence of resistant mutations, and these mutated strains have reduced replicative capacity in vitro and presumably in vivo as well. Several studies show that discontinuation of ART is often followed after 4-8 weeks by a significant increase in viral load and a very rapid decline in the CD4 cell count. Thus, the goal of therapy for heavily treated patients who have few therapeutic options is often to continue these drugs despite virologic failure with the hope of maintaining the CD4 cell count and preventing HIV-related complications. Virologic control would be nice, but may be impossible or unrealistic.

What is therapeutic drug monitoring (TDM)?

TDM is the measurement of serum concentrations of antiretroviral drugs to determine if there are adequate levels to assure antiviral activity or to account for toxicity. Relatively few laboratories offer the test, and interpretation is confounded by limited experience and great variability among laboratories testing the same samples. Nucleosides are not tested because this would require measuring intracellular concentrations. It is usually PIs, which may have marginal levels, and to a lesser extent NNRTIs that are tested. Although TDM is not common in clinical practice at present, many experts feel it may become a standard component of treatment in the future with better standardization and more information about how to use the tests. The anticipated use is in situations in which levels are difficult to predict, as in renal failure, hepatic failure, pregnancy, concurrent use of drugs with possible drug interactions, use of once-a-day PIs with concerns about trough levels, and the monitoring of adherence.

When is structured treatment interruption (STI) indicated?

Although there are 4 reasons to suspend treatment temporarily, pulse therapy appears to be the most promising.

• Immunization. One rationale for STI, when used with patients who have prolonged virologic control, is to stop therapy to let the virus come back and "immunize" the patient in a fashion analogous to a vaccine. The theory seemed good, but has not proven successful, and most have abandoned this tactic except with the rare patient who was treated very early in the course of the disease.
• Treatment failure. A rationale for discontinuing treatment when it has failed due to drug resistance is to allow growth of "wild type virus" that is sensitive to antiretroviral agents. The theory, that STI would permit a new round of therapy against sensitive virus, has not proven successful for possibly predictable reasons. In many or most patients, the resistant strains are minority species that quickly become the dominant strains under renewed antiviral pressure.
• Intermittent therapy. The plan with intermittent therapy is to periodically discontinue treatment on a prearranged schedule, such as 1 week on and 1 week off or 5 days of treatment followed by a weekend off. The theory is that when therapy is discontinued there is usually sustained viral suppression for 10-14 days and treatment is restarted while the virus is still suppressed. This strategy could potentially reduce treatment-associated side effects and cost. Although the initial experience has been limited but promising, the strategy cannot be recommended until more substantial experience is gained.
• Pulse therapy. With this strategy, therapy is discontinued when the CD4 cell count increases to a level that makes the patient and physician comfortable doing so and is restarted when the CD4 count declines to a worrisome level. Initial experience has generally been that the viral load returns within 2 weeks and the CD4 cell count declines rapidly and then plateaus. Most experts restart therapy when the CD4 cell count reaches 350/mm³, but this strategy has not been systematically studied. Nevertheless, it does appear that there is no penalty in terms of resistance and the period off therapy is often 1-2 years, depending to a large extent on the CD4 cell count at the time treatment is discontinued and on the pre-therapy CD4 nadir. The patient must be warned that viral loads will return to pre-therapy levels, with important implications for risk of HIV transmission to others.
  

• The major indicator for the speed of progression in early stage disease is the viral load, which dictates the speed of CD4 cell decline. The major indicator for the risk of HIV-associated complications is the CD4 cell count.
• ART is directed toward preventing HIV-associated complications and hospitalizations and improving quality of life and survival.
• ART does not cure HIV infection, is expensive, is associated with substantial risks of short-term and long-term toxicity, and requires a level of adherence that is unmatched with any other antimicrobial therapy.
• ART is recommended when there are HIV-associated symptoms or when the CD4 count is <200-350/mm³.
• The standard of practice for ART is 3 or 4 drugs representing 2 classes.
• The major objective of ART is to inhibit HIV as reflected by viral load. Successful antiviral therapy is defined as a viral load of <50 c/mL by 16-24 weeks, usually accompanied by an increase in the CD4 cell count of 100-150/mm³/year.
• Failed therapy is defined as virologic (viral load >400 c/mL twice after 24 weeks), immunologic (CD4 count fails to increase by >25-50/mm³ at 6-12 months), or clinical (new AIDS-defining OI after >3 months of therapy).
• Treatment is changed for drug intolerance (substitution) or virologic failure (selection by resistance tests).
• The 3 factors that consistently predict outcome are CD4 cell count, patient adherence, and provider experience.

Suggested Resources   TOP

Deeks SG, Hirschel B. "Supervised interruptions of antiretroviral therapy." AIDS. 2002;16:S157-169.

Hirsch MS, Brun-Vezinet F, Clotet B, et al. "Antiretroviral drug resistance testing in adults with Human Immunodeficiency Virus type 1: 2003 recommendations of an International AIDS Society - USA panel." Clin Infect Dis. 2003;37:113-128.

Lawrence J, Mayers DL, Hullsiek KH, et al. "Structured treatment interruption in patients with multidrug-resistant human immunodeficiency virus." New Engl J Med. 2003;349:837-846.

Mellors JW, Munoz A, Giorgi JV, et al: "Plasma viral load and CD4+ lymphocytes as prognostic markers of HIV-1 infection." Ann Intern Med. 1997;126:946-954.

Palella FJ, Delaney KM, Moorman AC, et al: "Declining morbidity and mortality among patients with advanced human immunodeficiency virus infection." N Engl J Med. 1998;338:353-360.

Panel on Clinical Practices for Treatment of HIV Infection. Guidelines for the Use of Antiretroviral Agents in HIV-Infected Adults and Adolescents. Washington DC: U.S. Department of Health and Human Services. Nov 10, 2003.

Yeni PG, Hammer SM, Hirsch MS et al. "Treatment for Adult HIV Infection: 2004 recommendations of the International AIDS Society-USA Panel." JAMA. 2004; 292:251-265.

Youle M. "Strategies of HIV Management - When to start." AIDS. 2002;16:S145-149.

WEBSITES

AIDSInfo: Accessed 11/03 (Several treatment guidelines are continuously updated at this U.S. Department of Health and Human Services website).

Cases   TOP

1. A 36-year-old woman is asymptomatic and has a CD4 cell count of 210/mm³ and a baseline viral load of >750,000 copies/mL. Other laboratory studies are unremarkable. She seeks your advice regarding management.

Question: What can you recommend?

a. zidovudine (AZT) plus stavudine (d4T) plus efavirenz (EFV)
b. lamivudine (3TC) plus zidovudine plus nelfinavir (NFV)
c. Trizivir (AZT + 3TC + ABC) plus lopinavir/ritonavir (LPV/r)
d. Trizivir
e. lamivudine plus stavudine plus indinavir (IDV)

Answer:
Virtually all of these are feasible except option a, which combines zidovudine and stavudine, a combination that shows pharmacologic antagonism. The worrisome part of her presentation is that the CD4 cell count is quite low and approaching the threshold of vulnerability to opportunistic infections, and the viral load is very high, which poses substantial challenge to virologic control. The best drugs for baseline high viral load according to currently available data are 2 nucleosides combined with efavirenz or lopinavir/ritonavir. Thus, option c would be the best choice.

2. A 42-year-old truck driver has HIV infection that was treated in 1999 with efavirenz (EFV), stavudine (d4T), and lamivudine (3TC). He subsequently did well and maintained a viral load of <50 c/mL, and the CD4 cell count increased from 250 to 450/mm³. He feels well, is fully adherent with his regimen by history, and has no complaints except that he is bothered by facial thinning (buccal cheek atrophy).

Question: Which of the following would you recommend?

a. Discontinue treatment
b. Change stavudine to abacavir (ABC)
c. Switch efavirenz to indinavir (IDV) plus ritonavir (RTV)
d. Switch stavudine to zidovudine
e. Give zidovudine, lamivudine, and abacavir

Answer:
Discontinuing therapy is an option, although guidelines on when and how to do this are sparse because it has not been extensively studied. Anecdotal information suggests that this patient will have a sustained period with a CD4 cell count above the threshold for initiating treatment by current guidelines of 200-350/mm³. The best predictor of a prolonged drug-free interval is the baseline CD4 cell count, which in his case is low, but not severely low. The main problem is that the CD4 cell count at 450/mm³ is high, but possibly not high enough. The most likely cause of his change in face appearance is stavudine, but methods to reverse the change are not necessarily clear. If we stop stavudine, face thinning will not progress and some studies suggest there may be reversal after a prolonged period. Since this is important to the patient and this kind of change is relatively easy to make, we should probably do it, so the best answers are to use a potent regimen without stavudine treatment (options a, b, and d).

3. A 36-year-old woman sees you for evaluation of HIV infection, which she has had since 1985. There have been multiple courses of treatment, including nucleosides in the period 1987 to 1996 and since then nucleosides combined with NNRTIs and PIs. The longest course of treatment was with zidovudine (AZT), lamivudine (3TC), and efavirenz (EFV), which produced a temporary period of virologic control but then failed. Currently she is receiving amprenavir (APV), lopinavir/ritonavir (LPV/r), and tenofovir (TDF). You perform a resistance test, which shows mutations on the reverse transcriptase gene at codons 41, 210, and 215 and mutations on the protease gene at 30 and 82. The conclusion is that HIV is resistant to most NRTIs and PIs. Her current numbers show a CD4 cell count of 87/mm³ and a viral load of 210,000 c/mL.

Question: What would you recommend?

a. zidovudine plus lamivudine plus tenofovir plus efavirenz
b. lopinavir/ritonavir plus efavirenz plus tenofovir
c. AZT/3TC/ABC (Trizivir) plus lopinavir/ritonavir
d. AZT/3TC/ABC plus tenofovir plus indinavir (IDV) plus ritonavir (RTV)
e. enfurvirtide (T-20) plus atazanavir plus lamivudine plus tenofovir

Answer:
The tricky part of this question is the need to assume resistance to efavirenz and lamivudine despite the failure to demonstrate the associated mutations: 103 and 184 on the RT gene. This reflects the fact that these drugs were not being given at the time the test was done, but history suggests that resistance to these drugs occurred at the time of failure. The point is that interpretation of resistance tests must take into account both the current pattern and the history of drug exposure in terms of specific agent, duration, and virologic outcome. This patient is running low on options and low on CD4 cells. She does have some PI options, but the most predictable response would probably be a regimen with the fusion inhibitor enfurvirtide (option e).

4. A 50-year-old secretary has just learned that he has HIV infection with a CD4 cell count of 49/mm³ and viral load of 280,000 c/mm³. He is quite shaken by this information, claims that he has never been able to take pills for anything and wants treatment, but wants it to be as simple as possible.

Question: What would you recommend?

a. Delay therapy until the patient is ready
b. AZT/3TC/ABC (Trizivir) plus efavirenz (EFV)
c. tenofovir (TDF), lamivudine (3TC), plus efavirenz
d. zidovudine (AZT), lamivudine, amprenavir (APV), and ritonavir (RTV)
e. AZT/3TC/ABC (Trizivir)
f. zidovudine plus efavirenz plus indinavir (IDV)

Answer:
This patient needs to be treated rapidly because he is highly vulnerable to major opportunistic infections. We emphasize the need for patient readiness, but this patient does not have much time to get ready. Training will take substantial effort as described in Chapter 7: Adherence to HIV Therapies. We would like potency plus convenience to facilitate adherence. The combination of lamivudine plus tenofovir plus efavirenz (option c) means 4 pills once a day, which could be taken, for example, when he shaves.

5. A doctor telephones you about a patient who has read about structured treatment interruption for patients who have failed therapy. You are told this patient has extensive experience with antiretroviral treatment for over 10 years, a CD4 cell count of 80-100/mm³, and a viral load that is back to baseline at 420,000 c/mL.

Question: What would you recommend to the physician?

a. Test resistance while receiving antiretroviral drugs and select the drugs based on the ART history and resistance tests
b. Discontinue treatment, test the HIV strain at 12 weeks, and treat accordingly
c. Continue therapy with same drugs
d. Treat with lopinavir/ritonavir (LPV/r), Trizivir (AZT/3TC/ABC), tenofovir (TDF), and either efavirenz (EFV) or nevirapine (NVP)

Answer:
The experience with STI for virologic failure is that wild-type virus that is less resistant replaces the predominant strain after about 6-11 weeks, but the CD4 cell count drops rapidly during the period off treatment; after new treatment, the more completely resistant pre-STI strain returns. The largest experience indicates only about 10%-15% of these attempts have been judged successful. Therefore, STI for virologic failure has fallen into disfavor. Nevertheless, if you wanted to play long odds, the right time to do the testing is after there has been a sufficient interval for ingrowth of a new, more sensitive strain of HIV making option b the correct option. However, most would prefer option a with expert interpretation of
the test results.

6. Question: Which of the following patients is the best candidate for genotype resistance testing?

a. A previously untreated pregnant woman with HIV
b. A chronically infected, previously untreated man with a partner who is currently taking triple therapy with poor control
c. A woman who has failed her initial treatment with a viral load of 7,000 c/mL after 1 year on indinavir (IDV), ritonavir (RTV), lamivudine (3TC), and stavudine (d4T)
d. A 40-year-old man with good virologic control, but a need to change therapy because of AZT-induced anemia

Answer:
Recommendations for the use of resistance testing are somewhat arbitrary, but the highest priority according to virtually all guidelines is for patients who have failed treatment for the purpose of identifying the next regimen; therefore, option c is the best answer.