A Primer on HIV Testing and Initial Treatment

Introduction

Since the introduction of potent combination antiretroviral therapy (ART) in the mid-1990s, HIV disease has been transformed from a progressive and usually fatal condition to a chronic, treatable disease that can be managed for decades.[1] As a result, the care of patients with HIV will increasingly be shared among HIV specialty providers, primary care clinicians, and subspecialists, much as individuals with other chronic diseases receive care from a broad range of clinicians. In addition, practitioners in primary care, emergency medicine, and obstetrics and gynecology will be the key drivers of the move to reduce the number of patients with HIV who are unaware of their infection.

The goal of this review is to summarize the current standard of care for HIV diagnosis, initial assessment, and treatment. The intended audience is primary care clinicians and other healthcare providers who may not have a large HIV patient population in their practice. Each section will begin with a brief clinical vignette to help frame the particular topic being discussed.
Pre-Assessment: Measuring Educational Impact

HIV Testing: Practical Issues and the Problem of Late Diagnosis

A 48-year-old man visits his primary care physician complaining of fatigue. Physical examination, a comprehensive metabolic panel, and a complete blood count are all normal. Although the patient is reassured, over the next 2 years his symptoms persist and he loses approximately 10 pounds. After developing fever and headache, he goes to a local emergency room and is diagnosed with cryptococcal meningitis. An HIV antibody test is sent and returns positive. The baseline CD4 count is 10 cells/µL.

The above clinical anecdote, drawn from my clinical practice, is all too familiar to experienced HIV providers. Despite the extraordinary advances in ART since combination treatment became the standard of care in the mid-1990s, a surprisingly high proportion (estimated to be 40%) of patients with HIV infection present for care with advanced immunosuppression and a diagnosis of AIDS.[2] Although data clearly support a survival benefit to starting ART prior to the onset of AIDS or the CD4 count falling below 200 cells/µL, the average CD4 count of patients in the United States who are beginning ART is between 150 and 200 cells/µL,[3] and the CD4 count at which patients present for care has actually gone down since 1990 (Figure 1).[4]

Click to zoom
(Enlarge Image)
Figure 1.

Median CD4 count of ART-naive patients by year of presentation for care at the Johns Hopkins HIV Clinic in Baltimore, Maryland.
From: Keruly JC, Moore RD.[4] Clin Infect Dis. 2007;45:1369-1374.

Approximately 1 out of 5 of the 1.1 million individuals living with HIV infection in the United States are unaware that they have the virus.[5] Not only would diagnosing HIV infection in these individuals benefit them personally, but epidemiologic studies also suggest that they account for a significant proportion of new HIV infections[6] and may explain why HIV incidence has not declined in this country since the early 1990s (Figure 2).[7] Encouraging behavioral studies suggest that individuals who are diagnosed with HIV infection significantly reduce their high-risk sexual behavior.[8]

Click to zoom
(Enlarge Image)
Figure 2.

Estimated new HIV infections in the United States.
From: Hall HI, et al.[7] JAMA. 2008;300:520-529.
[ CLOSE WINDOW ]

Figure 2.
Estimated new HIV infections in the United States.
From: Hall HI, et al.[7] JAMA. 2008;300:520-529.

On the basis of these data, it was clear to HIV providers and public health officials that the standard approach to HIV testing in clinical practice -- which was based largely on provider risk assessment -- was not working. In 2006, the CDC issued a major revision to their guidelines for HIV testing.[9] The key components of these recommendations are as follows:

* HIV screening is recommended for patients aged 13-64 years in all healthcare settings unless the prevalence of undiagnosed HIV infection in patients has been documented to be < 0.1%;
* Patients should be notified that testing will be performed unless the patient declines ("opt-out screening");
* Persons at high risk (in the United States: men who have sex with men, injection-drug users, and sexual partners of HIV-infected or high-risk individuals) should have repeat testing annually;
* Separate written consent for HIV testing should not be required; and
* Counseling is still recommended, but pretest counseling should not be required with HIV diagnostic testing or as part of HIV screening programs in healthcare settings.

These recommendations have been bolstered further by largely similar content from the American College of Physicians.[10] Although these guidelines are not binding, they have had an effect on state and local statutes. A recent review found that 34 states have regulations that are largely in concordance with these revised guidelines[11] In addition, HIV testing frequency in San Francisco significantly increased after removal of the requirement for written informed consent (Figure 3). Even more important is that this increase in the number of tests led to a greater number of individuals finding out that they had HIV infection and could now take advantage of potentially life-saving treatment.[12] Currently, 9 states continue to require written informed consent for HIV testing, although in several of these states, providers, community representatives, and healthcare officials are engaged in efforts to remove these barriers.

Click to zoom
(Enlarge Image)
Figure 3.

Association between rates of HIV testing and elimination of written consent in San Francisco.
From: Zetola NM, et al.[12] JAMA. 2007;297:1061-1062.
[ CLOSE WINDOW ]

Figure 3.
Association between rates of HIV testing and elimination of written consent in San Francisco.
From: Zetola NM, et al.[12] JAMA. 2007;297:1061-1062.
Which of the following are significant barriers to screening for HIV in your practice? (Please select all that apply.)
Insufficient time
Unwillingness to become involved in HIV care and follow-up
Burdensome consent process
Burdensome counseling process
Insurance, reimbursement, and other cost issues
Patients unwilling to be tested for HIV
Not knowledgeable about HIV testing
Not knowledgeable about HIV treatment
Uncomfortable discussing HIV
My patients are not at risk
Which of the following is the most significant barrier to screening for HIV in your practice?
Insufficient time
Unwillingness to become involved in HIV care and follow-up
Burdensome consent process
Burdensome counseling process
Insurance, reimbursement, and other cost issues
Patients unwilling to be tested for HIV
Not knowledgeable about HIV testing
Not knowledgeable about HIV treatment
Uncomfortable discussing HIV
My patients are not at risk

One potential concern of expanding HIV testing to lower-risk individuals is that it will increase the likelihood of a false-positive result. However, standard HIV testing -- which consists of a screening enzyme-linked immunosorbent assay (ELISA) test performed first, followed by a confirmatory Western blot -- is extraordinarily accurate. In studies performed even in very low-prevalence settings, the rate of false positivity was .0004% to .0007%, or approximately 1 false-positive test result for every 200,000 tests.[13-15] As a result, it is highly unlikely that a primary care provider in a typical practice will ever encounter a truly false-positive result. "Incorrect" HIV diagnoses are most often the result of laboratory or clerical errors (such as mislabeled specimens or incorrect data entry) and, while potentially emotionally upsetting to a patient, can be clarified easily and corrected with a repeat test.

One strategy for HIV testing that is prone to false-positive results is the rapid HIV test, especially the type that uses saliva. In this technique -- which can be done at the point of care and does not require a central laboratory -- a buccal swab is used to obtain the specimen. Results are generally available within 20 minutes and hence can be given to a patient at that time without the requirement for a return visit. A negative test generally rules out HIV infection with the same degree of sensitivity as standard antibody testing, but a reactive rapid test must be viewed as preliminary or inconclusive. Several studies have demonstrated a high rate of false-positive rapid results, especially when the test is done in low-prevalence areas.[16,17] As a result, all reactive rapid tests must be confirmed with standard ELISA/Western blot testing.

An important exception to the high sensitivity of HIV antibody testing is the patient who presents with recently acquired HIV, as HIV antibody is not detectable for 3-6 weeks after HIV acquisition.[18,19] During this "window" period, HIV can only be diagnosed using direct viral detection methods. The recommended tests for these purposes are any of the quantitative HIV RNA or viral load tests or the HIV-1 RNA qualitative assay. The former are widely available for patient monitoring and are therefore the tests used most often.

After HIV acquisition, an estimated 40% to 90% of patients will experience a self-limited febrile syndrome that may mimic many commonly encountered illnesses.[20,21] Typical symptoms include fever, pharyngitis, lymphadenopathy, and rash, indistinguishable from mononucleosis or other viral syndromes. More severe cases may include weight loss, oral and anogenital ulcers, and a wide range of neurologic syndromes, including meningoencephalitis, radiculitis, and mononeuritis multiplex. When acute HIV is diagnosed, the HIV RNA is invariably extremely high, generally measuring at > 100,000 copies/mL and sometimes > 1 million copies/mL.

It is not surprising that in light of the nonspecific nature of the presentation, acute HIV infection is often not diagnosed in clinical practice. However, because individuals who are worried about recent HIV exposure may seek HIV testing, it is important for the clinician to use both HIV antibody and HIV RNA testing in this particular situation, especially if the patient has symptoms consistent with acute HIV. The time course for various diagnostic tests to reveal a positive result relative to symptom onset is shown in Figure 4.

Click to zoom
(Enlarge Image)
Figure 4.

Dynamics of HIV viremia and antibody detection relative to symptom onset.
From: Branson BM.[18] Clin Infect Dis. 2007; 45:S221-S225. Based on data from: Fiebig EW, et al.[19] AIDS. 2003;17:1871-1879.

Baseline Evaluation

A 36-year-old woman applies for life insurance. She receives a letter stating that her HIV test is positive and that she should seek medical attention from her primary care provider. She is from sub-Saharan Africa and was tested once before in the late 1990s when she immigrated to the United States. She currently has one sexual partner whose HIV status is unknown.

Patients may be diagnosed with HIV infection in a variety of different settings, including routine screening by their primary care providers, during pregnancy, when donating blood, or, as occurred here, while applying for various forms of insurance.

Just as with any other significant disease, a new diagnosis of HIV requires a thorough history and physical examination. Certain aspects of the history are more relevant for this population, however, and deserve special emphasis. It is helpful to establish an approximate date of HIV acquisition; although this is not possible in all patients, many will be able to do so on the basis of symptoms compatible with acute HIV infection, prior negative HIV antibody tests, or the timing of high-risk exposures. A complete medication list -- one that also includes illicit and alternative therapies -- will help avoid potentially problematic drug-drug interactions with antiretroviral agents, particularly the nonnucleoside reverse transcriptase inhibitors (NNRTIs) and the protease inhibitors (PIs). For example, many of the commonly used HMG-CoA reductase inhibitors, or "statins," are contraindicated with PI-based ART, and the same is true with several of the antiseizure medications.

Particular focus should be placed on trying to assess the degree of immune competence of the patient. Worrisome symptoms or signs include weight loss, diarrhea, oral candidiasis, impaired memory, and recurrent bacterial infections, even though none of these conditions alone would typically lead to a diagnosis of AIDS. Because HIV is transmitted via the sexual and parenteral routes, queries about other sexually transmitted infections or those communicable through needle sharing (especially viral hepatitis) are indicated. Patients with HIV are at increased risk for reactivation of latent infections, so it is helpful to establish a thorough exposure history for such infections as tuberculosis and those caused by endemic fungi (histoplasmosis, coccidioidomycosis, and blastomycosis). Similarly, clinicians should document prior tuberculin skin testing results as well as whether preventive therapy was received.

Depression is more common among those with HIV infection than age-matched HIV-negative controls, and hence the review of systems should query particularly about symptoms possibly related to mood disorders. In addition, one of the most commonly used antiretroviral agents, efavirenz (EFV), may exacerbate depressive symptoms; therefore, an alternative agent should be chosen if possible.

The baseline laboratory assessment recommended by the DHHS guidelines for the use of antiretroviral agents in HIV-1-infected adults and adolescents[22] is extensive and is summarized in Table 1. Of the specific HIV-related laboratory tests, the most important are the CD4 count and the HIV RNA level (viral load). a CD4 count provides the best measure of current immune function and prognosis; it is specifically used to assess the risk for opportunistic infections and hence the indications for prophylaxis, and it also is used to assess whether asymptomatic patients should receive ART. The HIV RNA level provides information about the likely pace of CD4 decline in untreated individuals. Although some guidelines recommend performing these tests twice to provide a more accurate baseline, if the CD4 count is found to be significantly depleted ( < 200 cells/µL), it is unlikely that any subsequent measurement would change the fact that ART is indicated.

Table 1. Tests Recommended for HIV-Infected Patients Entering Care
Standard and HIV-Specific Laboratory Tests Screen for Other Medical Conditions

* CD4 count
* HIV RNA
* Resistance genotype
* Complete blood count
* Blood chemistry profile
* Liver enzymes
* Bilirubin
* Blood urea nitrogen
* Creatinine
* Urinalysis
* Fasting blood glucose
* Fasting serum lipids



* Syphilis
* Tuberculosis
* Toxoplasma gondii
* Hepatitis A, B, C
* Pap smear (women)
* Pregnancy (women)

Adapted from: US Department of Health and Human Services Guidelines for the Use of Antiretroviral Agents in
HIV-1-Infected Adults and Adolescents.[22]
When to Start Therapy?

The 36-year-old woman with a new HIV diagnosis is seen by her primary care provider. Additional history is obtained and is notable for the absence of weight loss or other symptoms suggestive of advanced immunodeficiency, but she reported a case of localized herpes zoster 2 years prior, which was uncomplicated. The physical examination is normal; liver and kidney tests are unremarkable. The CD4 count is 590 cells/µL and the HIV RNA is 12,500 copies/mL. Tests for viral hepatitis show immunity to hepatitis A and B, with a negative hepatitis C antibody. No HIV resistance is detected on genotype testing. She reports that her current sexual partner has been tested for HIV and that his test was negative.

Guidelines issued by the DHHS and the International AIDS Society (IAS)-USA provide evidence-based criteria for initiating ART.[22,23] All patients with HIV-related symptoms should be treated, as should those without symptoms who have a CD4 count < 350 cells/µL. Certain concomitant conditions should also prompt therapy, such as hepatitis B infection that requires treatment, HIV nephropathy, and pregnancy. When the CD4 count is > 350 cells/µL, the optimal time to initiate therapy is not well defined. The DHHS guidelines[22] recommend following CD4 counts every 3-4 months to determine when to start ART. The frequency of CD4 count monitoring may be extended to every 6 months if the patient has been on successful ART for more than 2-3 years. Clinicians might consider other factors, including age, patient readiness, potential impact on quality of life, likely patient adherence and, according to some experts, rapid decline in CD4 T cells (eg, a decrease of more than 120 cells/µL per year) or a high HIV RNA level (> 100,000 copies/mL).

Because there has never been a comparative clinical trial of immediate vs deferred therapy in patients with a CD4 count > 350 cells/µL, the bulk of evidence for answering this question arises from cohort studies and an important retrospective analysis from the SMART study. The SMART study compared a strategy of continuous vs intermittent therapy, the latter driven by the CD4 count. Treatment was stopped when the CD4 count increased to > 350 cells/µL, and then resumed when it fell to < 250 cells/µL.[24] The overall study was stopped early due to an increased risk for HIV-related opportunistic diseases or death in the intermittent-therapy arm. Important to the issue of when to start therapy in asymptomatic patients is that this clinical benefit of continuous therapy was seen even among those who entered the trial with a CD4 count > 350 cells/µL.[25]

One large cohort study suggests that treatment should be initiated at CD4 counts even higher than 350 cells/µL. In the NA-ACCORD study, all-cause mortality was evaluated among more than 9000 treatment-naive patients receiving care in North America from 1996 to 2006.[26,27] A group who started therapy when the CD4 count was > 500 cells/µL was compared with those who waited until it was between 351 and 500 cells/µL, with the endpoint of interest being all-cause mortality. Compared with the early-therapy group, individuals who waited until the CD4 count was < 500 cells/µL had a 94% increased risk for death.

A second recent cohort study also evaluated the question of when to start therapy. It included more than 20,000 patients who started therapy with a CD4 count < 550 cells/µL.[28] Rates of AIDS and death were compared in adjacent 100 cells/µL starting-threshold strata. In this study, a delay of ART to < 350 cells/µL (but not < 375 cells/µL) was associated with an increased risk for AIDS or death.

There are important limitations to all cohort analyses, the most important being that these studies cannot fully account for unmeasured differences between the 2 groups which might have influenced the outcome. For example, if the group who started treatment early had more "health-seeking" behavior, this might have improved survival due to factors unrelated to ART. In addition, the NA-ACCORD study to date has not provided the specific causes of death, which are unlikely to be solely HIV related given the high CD4 counts of the study groups. Nonetheless, the large size of the NA-ACCORD study and its sophisticated statistical methods provide the best data thus far that starting treatment with a CD4 count > 500 cells/µL has clinical benefit.

An additional benefit to starting therapy is that doing so reduces the risk for HIV transmission. It has long been known that HIV RNA is the strongest predictor of the risk for viral transmission.[29] Furthermore, a retrospective observational study of HIV-discordant heterosexual couples found that being on ART significantly reduced the prevalence of HIV infection in the uninfected partner.[30]

On the basis of these data and the observation that the decline in HIV RNA in genital secretions generally parallels that in plasma, the Swiss Commission Fédérale pour les Problèmes Liés au Sida issued a statement in January 2008 that those on suppressive ART could not transmit HIV to their seronegative partner provided that there was no concomitant sexually transmitted infection.[31] Although other health authorities cautioned that the risk should not be considered "zero" -- further evidenced by a case report of likely transmission in such a setting[32] -- the principle that ART reduces (but does not eliminate) infectivity remains scientifically valid. Indeed, a prospective study of nearly 3000 discordant couples in Africa found that when the infected partner was receiving ART, there was a 3- to 5-fold decreased likelihood of viral transmission (Figure 5).[33]

Click to zoom
(Enlarge Image)
Figure 5.

Reduction of HIV transmission risk and high-risk sex while prescribed ART: Results from discordant couples in Rwanda and Zambia.
From: Sullivan P, et al.[33] Program and abstracts of the 16th Conference on Retroviruses and Opportunistic Infections; February 8-11, 2009; Montreal, Canada. Abstract 52bLB.
[ CLOSE WINDOW ]

Figure 5.
Reduction of HIV transmission risk and high-risk sex while prescribed ART: Results from discordant couples in Rwanda and Zambia.
From: Sullivan P, et al.[33] Program and abstracts of the 16th Conference on Retroviruses and Opportunistic Infections; February 8-11, 2009; Montreal, Canada. Abstract 52bLB.

A summary of the potential risks and benefits of early ART is shown in Table 2.

Table 2. Benefits and Risks of Initiating Antiretroviral Therapy in Asymptomatic Patients With CD4 Count > 350 cells/µL
Potential Benefits of Early Therapy Include:
Maintenance of a higher CD4 count and prevention of potentially irreversible damage to the immune system
Decreased risk for HIV-associated complications that can sometimes occur at CD4 counts > 350 cells/µL, including tuberculosis, non-Hodgkin's lymphoma, Kaposi's sarcoma, peripheral neuropathy, HPV-associated malignancies, and HIV-associated cognitive impairment
Decreased risk for nonopportunistic conditions, including cardiovascular disease, renal disease, liver disease, and non-AIDS-associated malignancies and infections
Decreased risk for HIV transmission to others, which will have positive public health implications
Potential Risks of Early Therapy Include:
Development of treatment-related side effects and toxicities
Development of drug resistance because of incomplete viral suppression, resulting in loss of future treatment options
Less time for the patient to learn about HIV and its treatment, and less time to prepare for the need for adherence to therapy
Increased total time on medication, with greater chance of treatment fatigue
Premature use of therapy before the development of more effective, less toxic, and/or better studied combinations of antiretroviral drugs
Transmission of drug-resistant virus in patients who do not maintain full virologic suppression

Adapted from: US Department of Health and Human Services (DHHS) Guidelines for the Use of Antiretroviral Agents in HIV-1-Infected Adults and Adolescents.[22]

Which Therapy to Start?
Initial Treatment: Basic Principles

Selection of the optimal initial ART regimen generally falls into the province of a clinician with prior experience managing patients with HIV. However, despite the numerous antiretroviral agents available, certain basic principles apply which are well within the domain of the primary care provider. In addition, the increased simplicity of current treatments makes selection of the proper regimen easier than ever despite the fact that there are so many agents available.

Goals of initial therapy include reduction of HIV-related morbidity and mortality, improved quality of life, restoration and preservation of immune function, and control of HIV replication. This last goal is achieved when the HIV RNA is suppressed to below the limits of currently available assays. The assays currently widely available have a lower limit of detection of 40-75 copies/mL.

On the basis of current treatment guidelines, all recommended initial regimens consist of 2 nucleoside reverse transcriptase inhibitors (dual NRTIs) plus either a nonnucleoside reverse transcriptase inhibitor (NNRTI) or a ritonavir (RTV)-boosted protease inhibitor (PI/r). One of the 2 NRTIs should be lamivudine (3TC) or emtricitabine (FTC). The currently preferred regimens are shown in Table 3.[22] The subsequent sections will address the treatment decisions facing clinicians when selecting from among these choices.

Table 3. Antiretroviral Therapy for Treatment-Naive Patients*
NNRTI Recommendations Population in Which to Avoid or Use With Caution
Preferred Efavirenz Do not use during first trimester of pregnancy or in those with high pregnancy potential.
Use with caution in patients with unstable psychiatric disease.
Alternative Nevirapine Do not use in patients with moderate-to-severe hepatic impairment (Child-Pugh score B or C).
Do not use in women with pre-ART CD4 > 250 cells/µL or in men with pre-ART CD4 > 400 cells/µL.
Use with caution in patients on tenofovir/emtricitabine (or lamivudine); early virologic failure has been reported with this combination.
PI Recommendations
Preferred Atazanavir + ritonavir
(once daily) Do not use in patients who require high-dose (> 20 mg omeprazole equivalent/day) PPIs.
Use with caution in patients on PPIs (any dose), H2 blockers, or antacids.
Darunavir + ritonavir
(once daily)
Fosamprenavir + ritonavir
(twice daily)
Lopinavir/ritonavir
(once or twice daily) Do not use once-daily lopinavir/ritonavir in pregnant women.
Alternative Atazanavir, unboosted
(once daily) Do not use in combination with tenofovir or didanosine/lamivudine.
Fosamprenavir + ritonavir
(once daily)
or
Fosamprenavir, unboosted
(twice daily)
Saquinavir + ritonavir
(twice daily)
Dual-NRTI Recommendations
Preferred Tenofovir + emtricitabine Do not use in combination with unboosted atazanavir.
Use with caution:
• with nevirapine due to reports of early virologic failure
• in patients with underlying renal insufficiency
Alternatives Abacavir + lamivudine Do not use in patients who test positive for HLA-B*5701.
Use with caution in the presence of the following:
• HIV RNA > 100,000 copies/mL -- higher rate of virologic failure reported in ACTG 5202; or
• High risk for cardiovascular disease
Didanosine + lamivudine (or emtricitabine) Do not use in combination with unboosted atazanavir.
Do not use in patients with a history of pancreatitis or peripheral neuropathy.
Zidovudine + lamivudine Use with caution in the presence of pretreatment anemia and/or neutropenia (may improve or worsen with zidovudine).

*Patients naive to antiretroviral therapy should be started on a combination regimen that consists of either: 1 NNRTI + 2 NRTIs or 1 PI (preferably boosted with ritonavir) + 2 NRTIs.
ART = antiretroviral therapy; NNRTI = nonnucleoside reverse transcriptase inhibitor; NRTI = nucleoside reverse transcriptase inhibitor; PI = protease inhibitor; PPIs = proton-pump inhibitors
Adapted from: US Department of Health and Human Services (DHHS) Guidelines for the Use of Antiretroviral Agents in HIV-1-Infected Adults and Adolescents.[22]

Selection of NNRTI- vs PI-Based Treatment

All newly diagnosed patients with HIV infection should undergo genotypic resistance testing. Approximately 5% to 10% of patients are found to have resistance on baseline testing, indicating that they have acquired a resistant strain. Among the patterns of resistance observed, the most important clinically is NNRTI resistance; several studies have shown a significantly higher rate of virologic failure for NNRTI-based regimens when such resistance is identified at baseline.[22]

In the absence of baseline resistance, other factors come into play in the decision to choose NNRTI- or PI-based treatment. If we focus exclusively on EFV-based treatments as the only preferred NNRTI-containing strategy, then these regimens are -- compared with PI-based therapies -- overall simpler, associated with a lower rate of metabolic and gastrointestinal side effects, and generally have demonstrated greater antiviral potency. EFV is available in a once-daily formulation or in a coformulated pill as EFV/tenofovir (TDF)/emtricitabine (FTC), which is taken as 1 pill daily -- the simplest treatment regimen currently available.

The most important comparative clinical trial of EFV vs a boosted PI is ACTG 5142.[34] ACTG 5142 was an open-label clinical trial of 750 treatment-naive patients with a median CD4 count of 191 cells/µL; patients were randomized to receive EFV + 2 NRTIs, lopinavir/r (LPV/r) + 2 NRTIs, or an NRTI-sparing strategy of LPV/r + EFV. After a median of 112 weeks of follow-up, virologic failure was significantly less common with the EFV-based regimens compared with the LPV/r-based regimen, and the EFV + NRTI regimen was associated with a lower rate of metabolic side effects, particularly triglyceride elevation.[34] Of note, the LPV/r-containing arm had a greater CD4 response compared with EFV; in addition, there was a lower rate of protocol-specified lipoatrophy, though the rate was overall quite low in all patients who received TDF-containing regimens. A second study comparing LPV/r with EFV, both with zidovudine (ZDV)/lamivudine (3TC) and limited to patients with CD4 counts < 200 cells/µL, again demonstrated the virologic superiority of the EFV-containing treatment.[35] The difference was particularly striking in the patients with high baseline HIV RNA, reinforcing a finding in other studies demonstrating the potent antiviral activity of EFV in treatment-naive patients.

Despite the advantages of EFV-based therapy over LPV/r in these studies, EFV is not an appropriate treatment for all patients. Baseline NNRTI resistance, as described above, would mandate an alternative choice. In addition, treatment with EFV is associated with nearly universal central nervous system (CNS) side effects and therefore must be taken before bedtime so that daytime activities are not affected. Early in therapy, the CNS side effects are manifested as dizziness and grogginess; these early symptoms subside after 1-2 weeks in most patients, but many continue to experience them to some degree, most commonly alterations in sleep (insomnia, vivid dreams, and nightmares). A small subset of patients on EFV will have mood disorders, predominantly depression. Given these side effects, clinicians should use EFV with caution in patients with preexisting psychiatric disease.

EFV has been associated with fetal CNS abnormalities in animal reproductive studies involving nonhuman primates. Furthermore, there are case reports of human infant neural tube defects when used in the first trimester of pregnancy.[36] It is therefore listed as a "category D" drug for pregnancy and should be avoided in women of childbearing potential who are not using reliable contraception.

A final issue to consider with EFV is its relatively low barrier to resistance. In ACTG 5142, although EFV treatment carried a lower risk for virologic failure compared with LPV/r, when failure did occur, detectable resistance mutations were more commonly observed in the EFV arm than in the LPV/r arm. This lower risk for resistance in the LPV/r arm has been observed in all prospective studies of boosted PIs in treatment-naive patients; as such, a boosted PI might be a less risky choice for individuals who have trouble taking their ART regimen consistently.
Selection of an NNRTI for Initial Therapy

Among the 4 available NNRTIs, only EFV and nevirapine (NVP) are options for initial therapy. Etravirine is specifically indicated for treatment-experienced patients, with no clinical studies done in those who are treatment naive. Due to a higher pill burden and less favorable clinical trial results compared with EFV and NVP, delavirdine is rarely used and hence will not be considered further.

Between EFV and NVP, EFV is generally preferred due to the abundant clinical trial data supporting its potent antiviral activity, the simplicity of dosing, and its overall favorable safety and tolerability profile. Possible circumstances in which EFV might not be the optimal choice are described above. One prospective clinical trial has compared EFV and NVP: the 2NN study.[37] Treatment-naive patients were randomized to receive EFV, NVP twice daily, NVP once daily, or the combination of NVP + EFV; all study subjects also received stavudine (d4T) and lamivudine (3TC). In this study, EFV treatment led to a lower rate of virologic failure compared with the 2 NVP arms; the NVP + EFV arm was associated with increased toxicity without antiviral benefit. Furthermore, there were 2 drug-related deaths in the NVP-treated subjects, one from Stevens-Johnson syndrome and the other from NVP hepatotoxicity.

Although infrequent, these treatment-related deaths in the 2NN study highlight the potentially life-threatening hypersensitivity reaction seen rarely with NVP therapy. In a prospective clinical trial, 17% of NVP recipients experienced hepatotoxicity, with 2 deaths.[38] In a comprehensive review conducted by the makers of NVP, a 12-fold higher incidence of symptomatic hepatic events was seen in women (including pregnant women) with CD4 counts > 250 cells/µL at the time of NVP initiation when compared with women with CD4 counts < 250 cells/µL (11.0% vs 0.9%). An increased risk was also seen in men with pretreatment CD4 counts > 400 cells/µL when compared with men with pretreatment CD4 counts < 400 cells/µL (6.3% vs 1.2%).[39] NVP should therefore be avoided in these patient populations. In order to further reduce the incidence of hypersensitivity, NVP requires dose escalation, with a starting dose of 200 mg daily for 14 days, then increasing to a maintenance dose of 200 mg twice daily; liver function tests should be monitored closely, and the drug should be stopped for transaminase elevation or the occurrence of fever and rash.

Despite these concerns, NVP has extensive use worldwide as an initial antiviral option (often coformulated with d4T/3TC) and as part of a strategy to reduce perinatal transmission of HIV. Small clinical trials of NVP used with TDF and either FTC or 3TC have raised concerns about possible early virologic failures[40,41]; fully powered studies of NVP with TDF/FTC are ongoing.
Selection of a Boosted PI for Initial Therapy

The strategy of using low-dose RTV -- 100 mg or 200 mg daily -- in combination with another PI increases plasma levels of that PI and may allow for less frequent dosing. In addition, RTV boosting typically leads to a higher trough concentration, which may explain the favorable activity of some RTV-boosted PIs against some PI-resistant viruses.

Several boosted PIs are listed as either preferred or alternative initial therapies for patients with HIV: atazanavir (ATV), darunavir (DRV), fosamprenavir (FPV), lopinavir (LPV), and saquinavir (SQV). Only LPV is available coformulated with RTV (LPV/r) with each tablet consisting of 200 mg LPV and 50 mg RTV. This coformulation is highly convenient, does not require cold storage, and may reduce pharmacy copays. After LPV/r demonstrated superior antiviral activity compared with nelfinavir (NFV),[42] it became the gold standard for initial therapy among PI options.

Several recent clinical trials, however, have shown that other boosted PIs have comparable -- or, in some populations, superior -- antiviral efficacy to LPV/r. Furthermore, those boosted-PI strategies that use a lower daily dose of RTV (100 mg a day vs the 200 mg a day in LPV/r) have lower rates of gastrointestinal side effects and hypertriglyceridemia. As such, many clinicians have moved away from LPV/r and instead are using either once-daily ATV/r or DRV/r.

The CASTLE study was an open-label, randomized clinical trial comparing once-daily ATV/r (300 mg/100 mg, respectively) vs twice-daily LPV/r (400 mg/100 mg); study subjects also received coformulated TDF/FTC.[43] At 48 weeks, virologic suppression rates between the 2 study arms were similar (78% for ATV/r, 76% for LPV/r), meeting the protocol's prespecified criteria for noninferiority (Figure 6).

Click to zoom
(Enlarge Image)
Figure 6.

48-week efficacy and safety results of the CASTLE study: initial therapy with TDF/FTC + ATV/r vs TDF/FTC + ATV/r.
From: Molina JM, et al.[43] CASTLE Study Team. Lancet. 2008;372:646-655.
[ CLOSE WINDOW ]

Figure 6.
48-week efficacy and safety results of the CASTLE study: initial therapy with TDF/FTC + ATV/r vs TDF/FTC + ATV/r.
From: Molina JM, et al.[43] CASTLE Study Team. Lancet. 2008;372:646-655.

Treatment responses to ATV/r were similar regardless of baseline CD4 count; however, in the LPV/r group, there was a trend toward lower virologic suppression rate in those with progressively lower CD4 counts. Treatment with ATV/r in this study was associated with a lower incidence of both diarrhea and nausea vs treatment with LPV/r.

Additional favorable features of ATV/r are that it has the lowest daily pill count of any boosted-PI option (1 pill each of ATV and RTV) and has less of an effect on plasma lipids than LPV/r. The most common adverse effect is indirect hyperbilirubinemia; in the CASTLE study,[43] symptomatic jaundice was experienced by 4% of ATV/r patients vs 0% for LPV/r. In clinical practice, the jaundice may (rarely) be severe enough to cause cosmetic concerns for the patient, and an alternative regimen must be chosen. In addition, ATV therapy has been associated with nephrolithiasis; sometimes stone analysis will demonstrate ATV as a component of the stone.

Coadministration of ATV with tenofovir significantly reduces ATV concentrations. As a result, RTV boosting of ATV is required whenever the 2 drugs are given together. In addition, stomach acidity is required for optimal ATV absorption. Although H2-blockers and proton-pump inhibitors may be used in certain situations with ATV, clinicians should consult the ATV package insert carefully to review acceptable strategies. In general, our practice is to advise patients to avoid use of acid-reducing therapies when receiving ATV-based therapy; if such treatment is required (such as for therapy of reflux disease), an alternative boosted PI should be strongly considered.

Due to its in-vitro activity against many viruses that have resistance to other PIs, DRV was initially tested in clinical trials of treatment-experienced patients. The dose selected for this population was 600 mg of DRV with 100 mg of RTV, both given twice daily. However, pharmacokinetic studies suggested that DRV given in a once-daily 800 mg dose with 100 mg of RTV would yield trough concentrations greatly in excess of the minimum inhibitory concentration for wild-type virus. As a result, this dose was tested in treatment-naive patients: The ARTEMIS study compared DRV/r once daily with LPV/r given either once- or twice daily in 689 treatment-naive patients; again, the NRTIs in this study were coformulated TDF/FTC.[44] At 48 weeks, 84% of the DRV/r vs 78% of the LPV/r group had an HIV RNA < 50 copies/mL; virologic responses in those patients who began the study with an HIV RNA > 100,000 copies/mL were significantly better with DRV/r than with LPV/r (Figure 7), and a similar trend favoring DRV/r was seen in patients with lower CD4 counts (Figure 8).

Click to zoom
(Enlarge Image)
Figure 7.

Once-daily DRV/r + TDF/FTC vs LPV/r + TDF/FTC in treatment-naive HIV-1-infected patients at week 48 by baseline viral load.
From: Ortiz R, et al.[44] AIDS. 2008;22:1389-1397.
[ CLOSE WINDOW ]

Figure 7.
Once-daily DRV/r + TDF/FTC vs LPV/r + TDF/FTC in treatment-naive HIV-1-infected patients at week 48 by baseline viral load.
From: Ortiz R, et al.[44] AIDS. 2008;22:1389-1397.

Click to zoom
(Enlarge Image)
Figure 8.

Once-daily DRV/r + TDF/FTC vs LPV/r + TDF/FTC in treatment-naive HIV-1-infected patients at week 48 by baseline CD4 count.
From: Ortiz R, et al.[44] AIDS. 2008;22:1389-1397.
[ CLOSE WINDOW ]

Figure 8.
Once-daily DRV/r + TDF/FTC vs LPV/r + TDF/FTC in treatment-naive HIV-1-infected patients at week 48 by baseline CD4 count.
From: Ortiz R, et al.[44] AIDS. 2008;22:1389-1397.

As with the CASTLE study, patients who received DRV/r had a lower incidence of gastrointestinal side effects than those who received LPV/r. Ninety-six week follow-up of these data from the ARTEMIS study showed that DRV/r overall had a significantly lower rate of virologic failures than LPV/r, and that the difference between the 2 strategies was largely driven by lower response rates in those with suboptimal adherence to LPV/r.[45]

The results of the CASTLE[43] and ARTEMIS[44,45] studies demonstrate advantages of ATV/r and DRV/r, respectively, over LPV/r. By contrast, comparative clinical trials of twice-daily FPV/r[46] and SQV/r[47] showed nearly identical virologic response rates compared with LPV/r as well as similar adverse-event profiles. When given twice daily with RTV, both FPV and SQV have a daily pill count of 6/day, relatively high for current treatments. Of note, other dosing strategies of FPV in treatment-naive patients have also been FDA approved, including a once-daily 1400-mg dose with 100 mg or 200 mg of RTV for treatment-naive patients, and a 1400-mg dose twice daily without RTV for use in certain cases. Overall, there are fewer contemporary comparative data for these doses of FPV than for the 700 mg/100 mg FPV/r twice-daily dose.[48] SQV may have particular use in pregnant women, as drug levels appear to be reduced less with SQV than with other PIs. In addition, SQV will likely become the first PI available generically in the United States, which may reduce its cost. Finally, tipranavir (TPV) is a boosted PI that only should be used in treatment-experienced patients with PI-resistant virus, as treatment is associated with a higher rate of transaminase elevation and a greater number of drug-drug interactions than the other PI options.
Selection of the NRTI Pair

Fixed-dose formulations of NRTI pairs are available for abacavir (ABC)/3TC TDF/FTC, and ZDV/3TC. These coformulations reduce the pill burden of an antiviral regimen and may furthermore be associated with lower pharmacy copay. Treatments containing d4T are no longer recommended given the relatively high mitochondrial toxicity of this agent, a property that led to several complications, including lipoatrophy, peripheral neuropathy, lactic acidosis, and pancreatitis.[49,50] Study GS934 compared TDF/FTC and ZDV/3TC, both with EFV, in treatment-naive patients.[51] The study results demonstrated superior virologic activity of TDF/FTC (Figure 9) and also better safety and tolerability; more ZDV/3TC-treated patients required drug discontinuation (generally for anemia) than those receiving TDF/FTC and furthermore had a greater loss of limb fat as measured by DEXA scanning.

Click to zoom
(Enlarge Image)
Figure 9.

EFV + TDF/FTC vs EFV + ZDV/3TC in treatment-naive HIV-1-infected patients at week 48.
From: Gallant JE, et al.[51] N Engl J Med. 2006;354:251-260. Copyright © 2006 Massachusetts Medical Society. All rights reserved.
[ CLOSE WINDOW ]

Figure 9.
EFV + TDF/FTC vs EFV + ZDV/3TC in treatment-naive HIV-1-infected patients at week 48.
From: Gallant JE, et al.[51] N Engl J Med. 2006;354:251-260. Copyright © 2006 Massachusetts Medical Society. All rights reserved.

Since the results of this study, TDF/FTC has become the NRTI pair used most commonly in prospective studies comparing initial treatment strategies,[43,44,47] and all of these have demonstrated excellent virologic suppression rates.

Like TDF, ABC has low in-vitro mitochondrial toxicity and therefore provides a good alternative to ZDV- or d4T-containing NRTI pairs. In a prospective randomized study, ABC/3TC was compared with ZDV/3TC, with all study subjects also receiving EFV. The results demonstrated similar rates of virologic suppression, with ABC/3TC-treated subjects having a greater CD4 count increase.[52] A randomized open-label trial of either ABC or d4T with 3TC and EFV found higher rates of virologic suppression and lower rates of lipoatrophy, dyslipidemia, and drug discontinuation in the ABC arm.[53]

There have been 2 prospective clinical trials comparing TDF/FTC with ABC/3TC. In ACTG 5202, 1800 study subjects were randomized to 1 of the 2 NRTI-strategies and also were randomized to receive either EFV or ATV/r.[54] During the course of the study, an independent data and safety monitoring board halted the comparison between TDF/FTC and ABC/3TC for study participants with screening HIV RNA > 100,000 copies/mL due to a significantly higher rate of virologic failure in the ABC/3TC arm. In addition, ABC/3TC treatment was associated with a greater incidence of adverse events. By contrast, the HEAT study also compared TDF/FTC and ABC/3TC in 688 treatment-naive patients; here the third agent used was once-daily LPV/r. At 48 and 96 weeks, there was no difference between the 2 treatment arms in terms of virologic suppression.[55] Notable differences between A5202 and HEAT include sample size, different third drug strategy, definition of study endpoints, and loss to follow-up rates. These differences notwithstanding, major HIV treatment guidelines cite the ACTG 5202 results as a reason for using ABC/3TC with caution in patients with high baseline HIV RNA.[22,23,56,57]

Nephrotoxicity is the most important adverse event associated with TDF treatment. In general, prospective clinical trials and cohort studies have shown a lower rate of clinically significant renal dysfunction in patients who begin TDF/FTC therapy with normal or near normal renal function. For example, in study GS934,[51] creatinine clearance (as estimated by the Cockroft-Gault equation) remained within normal range for patients in both the TDF/FTC and ZDV/3TC arms, and there were no discontinuations due to renal complications in either group. Although renal complications are rare, clinicians should monitor creatinine and creatinine clearance periodically for all patients taking TDF and should be especially vigilant in patients with preexisting renal disease or risk factors for this condition, such as hypertension, diabetes, or receiving concomitant nephrotoxic agents. In general, we avoid use of TDF in patients with significantly impaired baseline impaired renal function; if TDF must be used, dose reduction is recommended using the strategies outlined in the package insert.

Treatment with ABC may lead to a potentially life-threatening hypersensitivity reaction (HSR). In prospective studies, ABC HSR occurred in 5% to 8% of study subjects, generally within the first 6 weeks of treatment; continued therapy in the face of HSR or rechallenge with ABC after cessation of the drug for HSR can be fatal. The risk for HSR is strongly linked to having the HLA-B*5701 allele, and clinical studies have shown that a negative test for this allele virtually eliminates the risk for severe HSR.[58] As a result, all patients should be screened for HLA-B*5701 prior to receiving ABC therapy.

An additional toxicity potentially linked to ABC therapy is an increased risk for cardiovascular disease. The D:A:D study is a large, multinational cohort study focusing on adverse events associated with ART. In one report involving over 33,000 participants with more than 150,000 person-years of follow-up, recent use of ABC was associated with a 1.9-fold increased risk for myocardial infarction.[59] Follow-up data from this cohort did not show a similar association with TDF.[60] Three additional studies have also found a link between ABC use and myocardial infarction risk,[61-63] but this has not been seen in all reviews.[64,65] Although the link between ABC use and myocardial infarction is not proven given the heterogeneity of the above study results, it is prudent to avoid ABC-containing regimens in patients at excess risk for cardiovascular events.[66]

In summary, the 2 NRTI pairs that should be considered for initial therapy are TDF/FTC and ABC/3TC. On the basis of the results from ACTG 5202 showing a lower risk for virologic failure with TDF/FTC than with ABC/3TC in patients with high HIV RNA, as well as a possible link between ABC therapy and myocardial infarction risk, the DHHS guidelines now list only TDF/FTC as a "preferred" initial NRTI option; the other major guidelines continue to list ABC/3TC as well but cite the above data as caveats regarding ABC/3TC use.
Use of Newer Agents in Initial Therapy

Approved for use in 2007 in treatment-experienced patients, the integrase inhibitor raltegravir (RAL) has also been studied in treatment-naive subjects. In a placebo-controlled study of more than 500 patients, study subjects received either once-daily EFV or twice-daily RAL along with matching placebos and TDF/FTC as the NRTIs.[67] At week 48, 86% of the RAL group and 82% of the EFV group had HIV RNA levels < 50 copies/mL, a nonsignificant difference. As had been noted in previous studies, the RAL treatment led to a faster HIV RNA decline than EFV, the clinical significance of which is uncertain. Raltegravir-treated patients experienced fewer drug-specific adverse events and also a greater CD4 count increase than those receiving EFV. These data strongly support the use of RAL in treatment-naive subjects; FDA approval for this indication is likely.

The CCR5 antagonist maraviroc (MVC) has also been tested in this patient population. In a 3-arm clinical trial, patients with CCR5-tropic virus were randomized to receive EFV or MVC once- or twice daily.[68] Due to lower virologic efficacy at an interim review, the once-daily MVC arm was discontinued prematurely. Ultimately, the rates of virologic suppression were similar for EFV and twice-daily MVC but did not meet the stringent (10%) criteria for noninferiority as defined at the study outset. Rates of virologic failure were higher in the MVC arm, particularly in those with HIV RNA > 100,000 copies/mL. In retrospect, some of the patients entering the study had low levels of CXCR4 using virus that was not detected at screening; this group did particularly poorly on MVC, as MVC is only active against CCR5-tropic viruses. A reanalysis of this study using an improved tropism assay that has greater sensitivity for CXCR4-using variants yielded nearly identical rates of virologic suppression between MVC and EFV.[69] Additional studies of MVC in treatment-naive patients are planned.
Selected ART Strategies Generally Not Recommended

Certain antiviral strategies are not recommended due to either lower virologic efficacy rates or higher toxicity. The choices cited here achieve virologic suppression in some patients but not at the rate seen with conventional treatments. They might be appropriate in rare patient scenarios but should not be considered the standard of care. These include the following:

* Triple-NRTI-based therapy: Only ABC/3TC/ZDV and ZDV/3TC + TDF should be considered in the rare situation of a patient who cannot tolerate standard regimens. Other NRTI-only strategies should be avoided.
* Boosted-PI monotherapy: Regimens containing only RTV-boosted lopinavir and atazanavir have been published or presented; overall, these strategies have generally showed a lower rate of virologic suppression and a higher incidence of treatment-emergent resistance compared with standard triple therapies.[70,71]
* Other NRTI-sparing regimens: In ACTG 5142, the NRTI-sparing regimen of EFV + LPV/r had a virologic efficacy comparable to the EFV + 2 NRTIs arm but had a higher incidence of drug-related side effects and emergence of resistance.[34] Other NRTI-sparing strategies, such as a boosted PI + RAL, are under study

Summary

Despite the abundance of options available for treatment of HIV, data from clinical trials support the use of relatively few initial therapies as generally preferred. A starting strategy containing TDF/FTC, with EFV, ATV/r, or DRV/r as the third agent, would be optimal for most patients; raltegravir may also be a good initial third drug choice if longer-term safety data are benign. An example of the convergence of these options is provided in a recent paper which documented a pronounced homogeneity of initial ART for treatment-naive patients in recent years, a marked contrast from an earlier treatment era. In 2000 and 2001, 7 unique regimens accounted for 95% of initial therapy; by 2007, only 2 regimens, TDF/FTC + EFV or ATV/r, were used in 95% of patients.[72] Remarkably, HIV treatment has never been so effective and well tolerated, and the bar has been set extremely high for new candidates to become an optimal initial choice.

Source : http://cme.medscape.com/viewarticle/701845