OVERVIEW: What every practitioner needs to know

Is modification of antiretroviral therapy required?

Modification of antiretroviral therapy may be required for two main reasons:

  • failure of the regimen to satisfactorily control human immunodeficiency virus (HIV) replication

  • patient intolerance to or toxicity of the therapeutic regimen

    Continue Reading

In what circumstances is a switch required?

Virological failure

Virological failure may occur for a number of reasons, although it most commonly occurs in the presence of suboptimal medication adherence. Other variables associated with virological failure include previous virological failure, a higher baseline viral load, a lower baseline (pretreatment) or nadir cluster of differentiation (CD)4+ count, a prior diagnosis of acquired immunodeficiency syndrome (AIDS), the presence of comorbidities, such as mental illness or active substance abuse, and the presence of drug resistant virus (transmitted or acquired).

Patient intolerance to or toxicity from the regimen

Intolerance to or toxicity from combination antiretroviral therapy is an unpredictable phenomenon. It is believed that the field of pharmacogenomics will eventually allow prescribers to tailor therapies to specific individuals based on a genomic understanding of agents that might be best suited to that individual. In HIV medicine, there is one specific example that demonstrates this principle, the use of abacavir. However, despite this example, the great majority of intolerabilities/toxicities are not so readily predicted and the clinician must manage these as they arise in an expectant and empirical fashion.

Which individuals are at greater risk of a need to switch antiretroviral therapy?

Virological failure

Patients may fail regimens because of aspects of the drug regimen, such as adverse events associated with the medication (making it difficult for the patient to adequately adhere to treatment). Failure may also be associated with adverse drug-drug interactions between the antiretrovirals and other medications (prescribed, over the counter, or alternative) that patients may be receiving. Failure may be associated with inadequate dosing of medications in the antiretroviral regimen, leading to insufficient exposure and selection of drug resistance, or over-exposure, leading to toxicity and intolerability with resultant poor adherence. An assessment should be made as to whether the patient has an intercurrent illness that may be associated with a brief bout of detectable viremia despite good adherence to antiretroviral treatment (ART).

It is vital that the patient feel that honest and candid answers can be given to these questions without penalty or embarrassment. The patient must not be made to feel as if the failure is theirs.

Patient intolerance to or toxicity from the regimen

There is little understanding about how to predict which patients will develop intolerance or intolerability to any specific component of an antiretroviral regimen other than in one specific instance, the use of abacavir. Table I summarizes the association between various antiretroviral drugs and common adverse reactions.

Table I.
Adverse effects N(t)RTIs NNRTIs PIs Entry inhibitors Integrase inhibitors
Bone marrow suppression Zidovudine (anemia, neutropenia)
Cardiovascular disease All PIs (myocardial infarction and CVA, greatest risk in those with conventional risk factors)Boosted-saquinavir, boosted-atazanavir, boosted-lopinavir (PR interval prolongation)bSQV (QT prolongation; healthy volunteer study)
Central nervous system effects Stavudine—Guillain-Barré like syndrome (rare) Efavirenz—abnormal dreams, sleepiness, poor concentration, depersonalization, depression, psychosis
Diabetes mellitus Stavudine, zidovudine, didanosine, indinavir
Dyslipidemia Stavudine, zidovudine, abacavir Efavirenz All PIs (degree of dyslipidemia generally proportional to dose of ritonavir booster)
Gastrointestinal Zidovudine, didanosine All PIs (adverse effect(s) generally proportional to dose of ritonavir booster)
Hepatic Stavudine, zidovudine, didanosine (steatosis)Didanosine—prolonged exposure associated with noncirrhotic portal hypertension Nevirapine (hepatitis, may be associated with systemic hypersensitivity reaction) All PIs (hepatitis and hepatic decompensation)
Hypersensitivity reaction (not simple rash) Abacavir Nevirapine
Lactic acidosis Stavudine, zidovudine, didanosine
Lipoatrophy Thymidine-analogue NRTIs (stavudine>zidovudine)
Myopathy, elevated CK Zidovudine Raltegravir
Nephrotoxicity Tenofovir (acute and chronic renal impairment, Fanconi syndrome [rare]) Indinavir, atazanavir (crystal and stone formation and renal colic)Indinavir (chronic renal impairment)
Osteopenia/osteoporosis Tenofovir>zidovudine, stavudine, abacavir
Rash All NNRTIs Atazanavir, fosamprenavir, darunavir maraviroc

bSQV, boosted saquinavir’; CVA, cerebrovascular accident; NNRTI, non-nucleoside reverse transcriptase inhibitors; N(t)RTI, nucleoside and nucleotide reverse transcriptase inhibitor; PI, protease inhibitor.

Beware: there are other diseases that can mimic this disease:

Virological failure

Virological failure must be clearly delineated from viral “blips.” “Blips” are defined as transient elevations in the plasma viral load in patients with previously “undetectable” viremia that return to “undetectable” in the absence of a change in regimen. Such “blips” may reflect recent suboptimal adherence, a burst of viremia during an intercurrent infection or simply a result reflecting technical variance in the measurement.

What is the optimal threshold for the definition of "virological failure"?

There is no current agreement regarding the degree of virological suppression required to provide durable antiretroviral efficacy in the absence of viral evolution and selection of drug resistance mutations. Although the gold standard of recent years has been full suppression of HIV replication in plasma to less than 50 copies/mL, the most recent set of USA Department of Health and Human Services (DHHS) adult and adolescent HIV treatment guidelines has amended this to a goal of durable plasma HIV replication suppression to less than 200 copies/mL. This change has been driven by at least two factors.

First is the observation that, since the introduction of the new generation real-time polymerase chain reaction (PCR) technology for measuring HIV viremia (e.g., Abbott M2000 and Roche Taqman assays), there has tended to be a wider degree of test variability at the lower range of the assay limit, such that many patients who on prior monitoring platforms had repeat HIV ribonucleic acid (RNA) less than 50 copies/mL may, after switching to the latest technology return repeat viral loads in the range of 50 to 200 copies/mL. Although this has caused anxiety among patients and providers, there is no clear evidence to date that these patients are at an increased risk of progression to virological failure.

Second, and perhaps more importantly, the Aids Clinical Trials Group (ACTG) has performed an analysis of two recent ACTG studies to determine whether a viral load threshold of 200 copies/mL might be an alternative to the use of 50 copies/mL. The two studies (A5095 and A5142) allowed patients to continue randomized ART after reaching criteria for virological failure, making it possible to perform the analysis. A false-positive virological failure was defined as a confirmed viral load of less than 50 copies/mL after a prior test reaching criteria for virological failure. Across four independent measures of time to virological failure, the threshold of less than 50 copies/mL gave higher false-positive results for virological failure (25%) compared with an alternative threshold of 200 copies/mL. This suggests suggest that a virological threshold of 50 copies/mL falsely declares virological failure for an unacceptably high number of patients who ultimately re-suppress less than 50 copies/mL without change in combination ART (cART), and that a threshold of 200 copies/mL may be preferable.

The US DHHS Guidelines have subsequently adopted 200 copies/mL as the preferred cut-off for definition of virological failure in clinical practice.

What laboratory studies should you order and what should you expect to find?

Results consistent with the diagnosis

Virological failure

The only essential test to repeat is the viral load. This should be done promptly, as waiting for the next routine appointment in 3 to 4 months risks prolonged failure with accumulation of resistance mutations. Conversely, one needs to allow enough time for viral re-suppression to occur if the patient has been transiently unwell or has had a recent episode of poor compliance for reasons other than drug intolerability and toxicity. A repeat test done sometime between 2 and 8 weeks after the original test is reasonable.

In addition to a viral load, most clinicians order an HIV resistance test. These come in three forms: genotypic tests, phenotypic tests, and virtual phenotypic tests. The virtual phenotypic test bears special mention: it is an algorithm for interpreting genotypic tests, by pairing genotype with a paired phenotype in the Virdo correlative database. All three are returned to the physician with a report that guides optimal drug selection by the physician within each ART class. Because the wild type (original infecting viral strain) virus can “overgrow” the mutated virus once the selective pressure of the combination ART regimen is removed, drug resistance testing should be performed while the patient is continuing to receive the failing regimen or within a month of stopping the failing regimen. Tests performed outside these conditions that fail to demonstrate mutations despite virological failure (and an assessment that the patient has been adequately adherent) should be treated with suspicion.

Intolerance or toxicity

In most cases the specific adverse event or toxicity and the temporal relationship with commencing therapy is highly suggestive of a causal association. However, in some cases, an investigation of plausible alternative explanations is reasonable (e.g., check viral hepatitis serology and recent alcohol intake in cases of suspected ART related hepatitis; check stools for pathogens in the case of suspected antiretroviral-induced loose diarrhea).

What laboratory tests should be ordered to help manage virological failure and how should the results be interpreted?

There are three types of resistance tests that can be performed. Each has advantages and disadvantages.

Recently, a group has developed a new format for interpretation of resistance and recommendations for future options based not on individual drugs but on regimens. This group, the Resistance Database Initiative (RDI), has developed a mathematical modeling approach, using random forest modeling, for the prediction of treatment response following antiretroviral failure (defined as the probability of the plasma viral load reaching <50 copies/mL on a particular regimen) to a number of combination antiretroviral regimens. The system makes its predictions of response to a new antiretroviral treatment based on the individual patient’s HIV genotype, treatment history, viral load, CD4 count, and time to follow-up.

A version has been developed that performs the same function in the absence of HIV genotype resistance data. The models were developed using thousands of “treatment change episodes” drawn from various patient databases. These were then validated and refined against independent training sets. Although the early experience suggests these models perform better than standard genotypic sensitivity scores, more experience is required before this technique could be widely applied. (HIV Resistance Response Database Initiative. HIV Treatment Response Prediction System. Available at: http://www.hivrdi.org/treps. Accessed 19 July 2012.)

Results that confirm the diagnosis

Virological failure

If the repeat test continues to show a detectable viral load despite good reported adherence to medication, it is time to consider a switch to a new regimen.


In some cases (not those potentially life threatening) in which a causal link is unclear, re-challenge with the drug or continuation with minor intervention (e.g., addition of an anti-gut-motility agent in the case of suspected antiretroviral-associated diarrhea) is reasonable.

Which initial antiretroviral therapy regimens are associated with less likelihood of selection of resistance?

The evidence suggests that the degree of resistance at virological failure is determined by the components of the antiretroviral regimen. Although most studies have consistently suggested that cART composed of an non-nucleoside reverse transcriptase inhibitor (NNRTI) plus two nucleoside and nucleotide reverse transcriptase inhibitors (2N(t)RTIs) confer the greatest likelihood of virological success when compared to a starting regimen consisting of a boosted-protease inhibitor (PI) plus 2N(t)RTIs, patients failing a NNRTI+2N(t)RTIs regimen are at greater risk of failing with resistance to both the N(t)RTI and NNRTI class; patients failing a first-line boosted-PIs plus 2N(t)RTIs generally fail with evidence of resistance to the N(t)RTI class alone. There is little accumulated experience with first-line regimens consisting of an InSTI plus 2N(t)RTIs. In the STARTMRK study, after 156 weeks of treatment, of the 19 patients with virological failure at a level greater than 400 copies/mL, four had demonstrable evidence of resistance to raltegravir.

Although there is some controversy about the extent to which low level virological failure (e.g., HIV RNA >50 but <1,000 copies/mL) risks evolution of HIV with accumulation of resistance, recent evidence suggest this is a clearly definable risk, with greater risk in those with higher viral loads during episodes of low level viremia (i.e., greater risk for those with a viral load of 500–1,000 copies/mL than for those with 50–500 copies/mL).

After first line virological failure or intolerance, Is there any indication for the use of boosted-protease inhibitor monotherapy?

A number of studies over the past few years have examined the use of boosted-PI monotherapy. Most have examined the use of boosted-lopinavir; two studies have examined the use of boosted-darunavir monotherapy. Although over a 48 to 96 week period boosted-PI monotherapy performs remarkably well, no study has been able to demonstrate that the strategy is noninferior to the use of a conventional cART regimen consisting of a boosted PI plus 2N(t)RTIs. Therefore, no guidelines currently recommend the use of a boosted PI monotherapy as a preferred strategy.

There may, however, be a place for the use of PI monotherapy in specific populations in which adherence has been or is anticipated to be poor and intermittent. In this population, an argument can be mounted that boosted PI monotherapy confers useful antiretroviral activity while the patient adheres and is not associated with penalty when the patient stops, as studies to date suggest that intermittent dosing with a boosted-PI is infrequently associated with the selection of PI resistance mutations.

Switch for virological failure

There is little evidence to guide the choice of an optimal second-line regimen following the virological failure of first-line cART. The components of a modified regimen to salvage virological failure depend on the components of the first. For instance:

  • A first-line cART regimen consisting of an NNRTI + 2N(t)RTI

    As a general rule, and in the context of standard 3 to 4 monthly monitoring of viral load, virological failure is associated with the selection of a single mutation conferring high level drug resistance to the NNRTIs (e.g., K103N for efavirenz and Y181C for nevirapine) with or without the signature mutation conferring high level drug resistance to lamivudine or emtricitabine (i.e., M184V). In this context the use of an N(t)RTI agent such as tenofovir in a new regimen is likely to retain adequate activity in a second-line regimen if combined with a robust agent selected from a new ART class to which the patient has not been previously exposed.

    The most experience in this regard is with the use of a ritonavir-boosted PI. This class is the most robust of all of the ART classes (meaning it confers durable efficacy with the least amount of antiretroviral drug support), a quality generally attributed to a high barrier to drug resistance, a high inhibitory quotient (the ratio of plasma drug exposure to viral susceptibility), and a high degree of specificity of these agents for the protease binding site.

    The specific choice of boosted PI depends on a number of factors. The most evidence exists for the use of either atazanavir, darunavir, or lopinavir as the preferred agent. Their main characteristics are summarized in Table II. The use of less robust agents in other classes with a lower genetic barrier (e.g., an InSTI or a second-generation NNRTI) added to 2N(t)RTIs should be avoided, as there is a high risk of subsequent failure and further selection of resistance.

  • Boosted PI + 2N(t)RTIs

    These first-line regimens generally fail with evidence of resistance to one or two (less commonly) of the N(t)RTI components alone. For reasons only partially understood, it is rare to fail first-line boosted PI containing regimens with evidence of resistance to the PI. In this context careful questioning regarding adherence to therapy must be undertaken.

    There is no good evidence to guide how best to manage this situation. If questioning reveals a specific issue or set of issues regarding timing of doses, intolerance, or toxicity, the regimen may be altered in such a way to continue with a bPI + 2N(t)RTIs (particularly if failure has occurred in the absence of or minimal N(t)RTI resistance) or with the introduction of new agents from new classes that may remove or diminish the identified problem. If there is no evidence of resistance and no apparent problem, the patient must be assessed for their motivation to take an active role in their disease management including adherence to ART.

  • InSTI + 2N(t)RTIs

    As a result of the relative novelty of this combination as a first-line selection, there is as little evidence to guide how best to modify ART on virological failure. The evidence to date suggests that these regimens fail with selection of resistance to the InSTI, (although there is some early suggestion that another InSTI currently in phase 3 development (dolutegravir) at 50 mg bid may retain at least a degree of activity against HIV variants carrying the typical RAL resistance mutations.

    As in the case with first line NNRTI+2N(t)RTI regimens, there may be selection of resistance to 3TC/FTC. Extrapolating from this, and applying the guideline that a modified regimen should contain at least two, if not three, fully active agents, a salvage regimen consisting of a bPI + 1-2N(t)RTIs +/- a second agent from an independent class would likely regain full virological control. Clearer guidelines await future studies.

Table II.
Boosted-atazanavir Boosted-darunavir Boosted -lopinavir
Potency (PI-naives) High High High
Pill burden 2 3 (qd), 6 (bid) 4
Common adverse effects Hyperbilirubinemia GI adverse events, rash GI adverse events
Dosing qd qd (800mg) bid (600mg) qd (800mg)bid(400mg)
Lipid profile* Minor change Minor change (qd)adverse (bid) Adverse
Resistance barrier High Very high High
Drug-drug interaction potential HighShould not be administered with proton pump inhibitor agents High High

Metabolic changes are in general proportional to the total daily dose of ritonavir required for boosting

Switch for antiretroviral intolerance/toxicity

There are a number of situations in which ART might be modified in the absence of virological failure. The main reasons are as follows:

  • Modification to eliminate or modify a specific adverse event (e.g., anemia, neuropathy, neuropsychological disturbance, lipodystrophy) or to confer an improved risk profile (e.g., lipids) See Table III. Recommendations for reasonable switch options to eliminate or modify a specific adverse event.

  • Modification to alter as unacceptable risk for the development of a serious non-AIDS event (e.g., myocardial infarction). See Table IV.

    The SMART study demonstrated that those receiving continuous ART aimed at fully suppressing HIV replication were less likely to develop serious non-AIDS events (SNAEs), compared to those receiving intermittent CD4 guided treatment. There is, nevertheless, concern about associations between particular ARVs and the risk of SNAEs. Table III lists those agents that have been associated with SNAEs and suggests possible strategies. (Note – Although the associations have been described, there are few data on which to base recommendations to ameliorate any potential risk.)

  • Modification to confer a specific desirable effect (e.g., better penetration into the central nervous system). See Table V.

    This is a controversial issue. The pathogenesis of HIV-associated neurocognitive disease (HAND) is poorly understood. One prominent school of thought holds that, to treat (or avoid) neurocognitive defects associated with HIV disease, patients should be administered a selection of ARVs with a demonstrable capacity to cross the blood-brain barrier. The CHARTER group has developed a BBB penetration score as a guide for the selection of combination of ART that will maximize creating an antiretroviral regimen. Clinical experience using such regimens is limited and has given mixed results. There are no HIV treatment guidelines to date that specifically recommend the use of neuro-penetrative cART.

Table III.
Adverse event Associated ARV Switch candidate* Anticipated benefit
Anemia Zidovudine Stavudine (little used nowadays because of high risk of lipodystrophy) abacavir, tenofovir or other suitable agent Resolution
Neuropathy Stavudine, didanosine Abacavir, tenofovir, zidovudine or other suitable agent Reversal/reduction/stabilization
Lipoatrophy Thymidine analogue-NRTIs Abacavir, tenofovir other suitable agent Prevention of further lipoatrophy, mild reversal
Neuropsychological disturbance Efavirenz Alternative NNRTI, boosted-PI, CCR5 antagonist Resolution
Diarrhea Boosted-lopinavir, nelfinavir, boosted-fosamprenavir Other boosted-PI, InSTI, CCR5 antagonist Resolution/reduction
Dyslipidemia All boosted PIs (degree generally proportional to ritonavir booster dose)Abacavir, zidovudine, stavudine Switch to unboosted atazanavir or atazanavir given 200mg bid or switch to an alternative ART classTenofovir or other suitable agent Resolution/improvementImproved lipid profile
Injection site reactions Enfuvirtide Raltegravir or other suitable agent
Portal hypertension and nodular regenerative hyperplasia Didanosine Other suitable agent prevention of progression
Renal impairment Indinavir, boosted indinavir, tenofovir** Switch to an alternative NRTI, boosted-PI or other suitable agent Prevention of progression/resolution
Hepatitis Didanosine, stavudine, nevirapine, tipranivir, ritonavir (full antiretroviral [not booster]dose) Switch to a suitable NRTI, NNRTI, boosted-PI or other suitable agent Resolution
Cardiac failure Tenofovir Any other suitable agent Reduced risk of cardiac failure

Switch only in those situations in which the resultant regimen can be assured to provide continuing durable virological efficacy (i.e., in those patients in whom the full activity of the supporting antiretroviral agents is assured).

In patients experiencing renal impairment on tenofovir in combination with a boosted-PI, a switch from the boosted-PI to an agent from another class (without the need for ritonavir-boosting) may result in improvement alone.

Table IV.
SNAE Drug Switch Candidate Expected Benefit
Acute myocardial infarction (AMI) Abacavir Other suitable agent except didanosine Reduced risk of AMI
Cerebro-vascular accident (CVA) Abacavir Other suitable agent Reduced risk of CVA
Acute or chronic kidney disease (AKD/CKD) Tenofovir indinavir Other suitable agent Reduced risk of AKD/CKD
End-stage liver disease (ESLD) Didanosine Other suitable agent Reduced risk of ESLD
Severe infections (pneumonia) Enfuvirtide Raltegravir or other suitable agent Reduced risk of pneumonia
Table V.
CPE Score 4 3 2 1
ART drug class
N(t)RTIs zidovudine abacaviremtricitabine didanosinelamivudine stavudine tenofovirzalcitabine
NNRTIs nevirapine efavirenzdelavirdine etravirine
PIs boosted-indinavir boosted-darunavirboosted-fosamprenavirindinavirboosted-lopinavir atazanavirboosted-atazanavirfosamprenavir nelfinavirritonavirsaquinavirboosted-saquinavirboosted-tipranavir
InSTIs raltegravir
Entry inhibitors maraviroc enfuvirtide

The other reservoir for which adequate antiretroviral penetration has been considered desirable is the genital tract and, in particular, the genital secretions. This consideration is based on the public health interest in the reduction of infectiousness of the infected individual.

Table IV. Serious non-AIDS events (SNAE) associated with ART and suggested management strategies

Management of virological failure of second-line cART and beyond

Over the past decade, a series of studies have been reported that examined the use of new antiretroviral agents (new more potent drugs in existing classes or new drugs in new classes) against placebo, both combined with an “optimized background antiretroviral regimen” in the management of HIV-infected patients with evidence of resistance to various agents and classes of ART. In these studies, the background regimen was optimized by the use of various forms (and sometimes combinations) of resistance tests. Although the studies are not homogenous in design, study population, or analysis, their results suggest that the optimal response to any ART combination will most likely to be achieved if at least two, and perhaps three, agents to which the virus remains fully susceptible are employed in the new regimen.

There are some additional considerations before constructing a new regimen in the context of confirmed virological failure:

  • The aim of a new regimen should be full and durable viral suppression.

  • Plan for success but try to maintain future options if you encounter failure.

  • Take care with the potential for negative drug-drug interactions, particularly when combining agents metabolized through the CYP450 metabolic enzyme system (i.e., PIs, NNRTIs and CCR5 antagonists).

  • Exclusive CCR5 tropism becomes less likely as immunodeficiency progresses (think about using maraviroc when the option exists).

  • Convenience is important but is short-sighted if the regimen is doomed to fail (explain to the patient that after optimizing the chances of success, even if this means twice daily dosing, consideration can be given to switching to a more convenient strategy once virological control has been achieved).

  • The recycling of N(t)RTIs is controversial and lacks data from RCTs specifically designed to answer the question. N(t)RTI-sparing regimens are potent, and data are emerging that suggest that recycling N(t)RTIs may be unnecessary and may add to pill burden and toxicity.

  • There are few data from studies to make recommendations about the safety and efficacy of specific combinations of therapies in late salvage. However, the data to date (generated in ART-naïve patients) does lend general support to the use of the following combinations:

    a boosted PI in combination with an InSTI

    a boosted PI +/- a CCR5 antagonist

    a boosted PI plus an NNRTI

Of all of these combinations, the N(t)RTI-sparing option of a bPI plus an NNRTI has the most evidence. Although it confers comparable efficacy to a combination of an NNRTI plus bPI in at least two large RCTs, it is associated with an adverse metabolic profile compared to conventional regimens using N(t)RTI backbones. None of these combinations is specifically recommended to date in guidelines.

Salvaging virological failure of an InSTI + 2N(t)RTIs

Because of the relative novelty of this combination as a first-line selection, there is as little evidence to guide how best to modify ART on virological failure. The evidence to date suggests that these regimens fail with selection of resistance to the InSTI (although there is some early suggestion that dolutegravir at 50 mg bid (double the standard starting dose for ART naives) may retain at least a degree of activity against HIV variants carrying the typical RAL resistance mutations). As in the case with first line NNRTI+2N(t)RTI regimens, there may be selection of resistance to 3TC/FTC. Extrapolating from this and applying the guideline that a modified regimen should contain at least two, if not three, fully active agents, a salvage regimen consisting of a bPI + 1-2N(t)RTIs +/- a second agent from an independent class would likely regain full virological control. Clearer guidelines await future studies.

WHAT'S THE EVIDENCE for specific management and treatment recommendations?

Carr, A, Workman, C, Smith, DE. “Abacavir substitution for nucleoside analogs in patients with HIV lipoatrophy: a randomized trial”. JAMA. vol. 288. 2002. pp. 207-15. (In this sample of lipoatrophic HIV-infected adults, switching from stavudine or zidovudine to abacavir for 24 weeks led to significant, albeit modest, objectively measured increases in limb fat.)

Choi, AI, Vittinghoff, E, Deeks, SG, Weekley, CC, Li, Y, Shlipak, MG.. “Cardiovascular risks associated with abacavir and tenofovir exposure in HIV-infected persons”. AIDS. vol. 25. 2011. pp. 1289-98. (This study of a large observational database found that recent abacavir exposure was independently associated with an increased risk of cardiovascular events and recent tenofovir exposure with heart failure.)

Cozzi-Lepri, A, Phillips, AN, Ruiz, L. “Evolution of drug resistance in HIV-infected patients remaining on a virologically failing combination antiretroviral therapy regimen”. AIDS. vol. 21. 2007. pp. 721-32. (In patients kept on the same virologically failing cART regimen for a median of 6 months, there was considerable accumulation of drug resistance mutations.)

Cruciani, M, Zanichelli, V, Serpelloni, G. “Abacavir use and cardiovascular disease events: a meta-analysis of published and unpublished data”. AIDS. vol. 25. 2011. pp. 1993-2004. (This meta-analysis of RCTs does not support the hypothesis that ABC-containing combination ART regimens carry a greater risk of MI or major cardiovascular events relative to comparator regimens.)

“Use of nucleoside reverse transcriptase inhibitors and risk of myocardial infarction in HIV-infected patients enrolled in the D:A:D study: a multi-cohort collaboration”. Lancet. vol. 371. 2008. pp. 1417-26. (This cohort study found an increased risk of myocardial infarction in patients exposed to abacavir and didanosine within the preceding 6 months. The excess risk could not be explained by underlying established cardiovascular risk factors and was not present beyond 6 months after cessation of drug.)

Else, LJ, Taylor, S, Back, DJ, Khoo, SH.. “Pharmacokinetics of antiretroviral drugs in anatomical sanctuary sites: the male and female genital tract”. Antivir Ther. vol. 16. 2011. pp. 1149-67. (A review article that summarizes the state of knowledge regarding the degree of penetration of current antiretroviral agents into the genital tract.)

Elzi, L, Marzolini, C, Furrer, H. “Treatment modification in human immunodeficiency virus-infected individuals starting combination antiretroviral therapy between 2005 and 2008”. Arch Intern Med. vol. 170. 2010. pp. 57-65. (Drug toxicity remains a frequent reason for treatment modification; however, it does not affect treatment success.)

Eron, JJ, Young, B, Cooper, DA. “Switch to a raltegravir-based regimen versus continuation of a lopinavir-ritonavir-based regimen in stable HIV-infected patients with suppressed viraemia (SWITCHMRK 1 and 2): two multicentre, double-blind, randomized controlled trials”. Lancet. vol. 375. 2010. pp. 396-407. (Although switching to raltegravir was associated with greater reductions in serum lipid concentrations than continuation of ritonavir-boosted lopinavir, efficacy results did not establish non-inferiority of raltegravir to ritonavir-boosted lopinavir.)

“Clinical management and treatment of HIV-infected patients in Europe”. (This HIV treatment guideline is extensively referenced and updated at regular intervals.)

Katlama, C, Haubrich, R, Lalezari, J. “Efficacy and safety of etravirine in treatment-experienced, HIV-1 patients: pooled 48 week analysis of two randomized, controlled trials”. AIDS. vol. 23. 2009. pp. 2289-300. (Treatment-experienced patients receiving etravirine plus an optimized background regimen had superior antiviral responses with comparable tolerability compared with those receiving placebo plus background regimen.)

Letendre, SL, Ellis, RJ, Ances, BM, McCutchan, JA.. “Neurologic complications of HIV disease and their treatment”. Top HIV Medicine. vol. 18. 2010. pp. 45-55. (This review summarizes the CHARTER groups antiretroviral CNS penetration effectiveness score (CPE) score and the supporting evidence.)

Madruga, JV, Berger, D, McMurchie, M. “Efficacy and safety of darunavir-ritonavir compared with that of lopinavir-ritonavir at 48 weeks in treatment-experienced, HIV-infected patients in TITAN: a randomized controlled phase III trial”. Lancet. vol. 370. 2007. pp. 49-58. (In lopinavir-naive, treatment-experienced patients, darunavir-ritonavir provided non-inferior antiviral efficacy compared to boosted-lopinavir.)

Martínez, E, Arnaiz, JA, Podzamczer, D. “Substitution of nevirapine, efavirenz, or abacavir for protease inhibitors in patients with human immunodeficiency virus infection”. N Engl J Med. vol. 349. 2003. pp. 1036-46. (When therapy was switched from a protease inhibitor to nevirapine, efavirenz, or abacavir in patients with virological suppression, there was a trend toward a higher rate of virological failure among those given abacavir.)

Molina, JM, Andrade-Villanueva, J, Echevarria, J. “Once-daily atazanavir/ritonavir versus twice-daily lopinavir/ritonavir, each in combination with tenofovir and emtricitabine, for management of antiretroviral-naive HIV-1-infected patients: 48 week efficacy and safety results of the CASTLE study”. Lancet. vol. 372. 2008. pp. 646-55. (In treatment-naive patients, boosted-atazanavir qd demonstrated non-inferior antiviral efficacy compared to lopinavir/ritonavir bid with less gastrointestinal toxicity but more hyperbilirubinemia.)

Ortiz, R, Dejesus, E, Khanlou, H. “Efficacy and safety of once-daily darunavir/ritonavir versus lopinavir/ritonavir in treatment-naive HIV-1-infected patients at week 48”. AIDS. vol. 22. 2008. pp. 1389-97. (In treatment naïve patients, boosted-darunavir qd provided to be non-inferior to boosted-lopinavir bid at 48 weeks, with a more favorable safety profile.)

Guidelines for the use of antiretroviral agents in HIV-1-infected adults and adolescents. January 10, 2011. pp. 1-166. (This HIV treatment guideline is extensively referenced and updated at regular intervals.)

Revell, AD, Wang, D, Boyd, MA. “The development of an expert system to predict virological response to HIV therapy as part of an online treatment support tool”. AIDS. vol. 25. 2011. pp. 1855-63. (The models achieved a consistent, high level of accuracy in predicting treatment responses, which was superior to that of genotypic sensitivity scores.)

Riddler, SA, Haubrich, R, DiRienzo, AG. “Class-sparing regimens for initial treatment of HIV-1 infection”. N Engl J Med. vol. 358. 2008. pp. 2095-106. (Virologic failure was less likely in the efavirenz group than in the lopinavir-ritonavir group. The virologic efficacy of the NRTI-sparing regimen was similar to that of the efavirenz regimen but was more likely to be associated with drug resistance.)

Günthard, HF, Aberg, JA, Eron, JJ, Hoy, JF. “Antiretroviral treatment of adult HIV infection: 2014 recommendations of the International Antiviral Society-USA Panel”. JAMA. vol. 312. 2014 Jul 23-30. pp. 410-25. (This HIV treatment guideline is extensively referenced and updated at regular intervals.)

Tashima, KT, Smeaton, LM, Fichtenbaum, CJ, Andrade, A. “HIV Salvage Therapy Does Not Require Nucleoside Reverse Transcriptase Inhibitors: A Randomized, Controlled Trial”. Ann Intern Med.. vol. 163. 2015 Dec 15. pp. 908-17. (A study to settle a long-lasting controversy regarding the role of NRTIs in salvage therapy.)