Evaluation of the treatment-experienced patient failing HIV therapy

INTRODUCTION — The goal of antiretroviral therapy is to suppress the plasma HIV RNA below the limits of assay detection (eg, <20 copies/mL). Suppressing the viral load prevents the emergence of drug-resistant virus and disease progression [1,2]. However, certain patients with HIV will experience virologic failure on their antiretroviral regimen. Such individuals require a detailed evaluation to determine the reason for failure and the best approach to achieve viral suppression.

The evaluation of the patient failing their antiretroviral regimen will be reviewed here. Selecting a new regimen for patients with virologic failure is discussed elsewhere. (See "Selecting an antiretroviral regimen for treatment-experienced patients with HIV who are failing therapy".)

VIROLOGIC FAILURE — For individuals starting a new regimen, virologic failure is defined as the inability to achieve a viral load <200 copies/mL within 24 weeks of initiating antiretroviral therapy. For individuals who were initially able to suppress their viral load, virologic failure is defined as a recurrence of viremia to >200 copies/mL on two consecutive measurements taken approximately one month apart [3,4].

In contrast, intermittent "blips" of low level viremia (above the limits of assay detection to 200 copies/mL) are generally not associated with the emergence of drug-resistant virus; thus, intermittent viral blips should not prompt a change in therapy [5].

The clinical significance and management of persistent viremia detected at above the limits of assay detection to 200 copies/mL are less clear. We generally continue the current regimen with counseling to enhance adherence and to avoid drug-drug and drug-food interactions. However, in some cases (eg, patients who have continued difficulty adhering to a daily regimen), we change to a regimen with a high barrier to resistance, if possible, to minimize the risk of emerging resistance. Such regimens include a pharmacologically boosted protease inhibitor or a dolutegravir- or bictegravir-containing regimen (if the patient is not already on this class of drugs). (See "Selecting an antiretroviral regimen for treatment-experienced patients with HIV who are failing therapy".)

REASONS FOR VIROLOGIC FAILURE — Virologic failure typically results from poor adherence to an antiretroviral therapy (ART) regimen, or less commonly, from drug-drug or drug-food interactions that impair absorption. Patients with virologic failure may or may not have drug-resistant virus. The risk of developing drug resistance depends, in part, upon the type of regimen (eg, one that contains an agent with a high versus low barrier to resistance) as well as the level of adherence. (See 'Nonadherence' below and 'Poor absorption or altered metabolism' below and 'Resistance' below.)

With the introduction of more potent antiretroviral agents, the risk of developing virologic failure to first- or second-line regimens has decreased. As an example, in a retrospective study of 13,165 patients who were initiated on a new regimen after an initial treatment failure, the crude incidence of a subsequent treatment failure significantly declined from 56 events per 100 person-years in 1996 to 16 events per 100 person-years in 2005 [6].

More detailed discussions of specific antiretroviral agents are found elsewhere. (See "Selecting antiretroviral regimens for treatment-naïve persons with HIV-1: General approach" and "Selecting an antiretroviral regimen for treatment-experienced patients with HIV who are failing therapy" and "Overview of antiretroviral agents used to treat HIV".)

Nonadherence — Patients with poor adherence to their antiretroviral regimen can develop virologic failure. Those with detectable viremia and very low levels of adherence (eg, those not taking any medications) typically have a lower risk of developing resistance compared with patients who have intermediate levels of adherence. On occasion, a patient can develop drug-resistant virus despite excellent adherence; such individuals may have had drug-resistant virus prior to treatment initiation (ie, transmitted drug resistance) that was not detected by standard assays. (See "Selecting antiretroviral regimens for treatment-naïve persons with HIV-1: General approach", section on 'Considerations prior to initiating treatment'.)

If the patient admits to difficulties with adherence, potential barriers include the number and timing of doses, the size of the pills, and the presence of treatment-limiting side effects. Other barriers include depression, substance abuse, and housing instability [7]. (See 'Assess medication adherence' below.)

Poor absorption or altered metabolism — Pharmacokinetic causes of virologic failure should be explored, particularly in patients with low-level virologic failure (viral load <1000 to 2000 copies/mL). Poor absorption can result from drug-drug interactions (eg, atazanavir or rilpivirine administered with a proton pump inhibitor) or drug-food interactions (eg, rilpivirine administered without a meal). It can also result from underlying medical conditions that result in malabsorption. (See 'Assess for poor absorption or altered metabolism' below.)

Resistance — Patients who develop virologic failure may develop drug resistance mutations to agents with a low genetic barrier to resistance (eg, lamivudine, emtricitabine, raltegravir, elvitegravir, non-nucleoside reverse transcriptase inhibitors [NNRTIs]). The spectrum of drug resistance in an individual patient can range from narrow resistance that affects the activity of one or two drugs (eg, the M184V mutation associated with resistance to lamivudine or emtricitabine) to multidrug resistance that includes resistance to several drug classes (eg, nucleoside reverse transcriptase inhibitors [NRTIs], NNRTIs, and/or integrase strand transfer inhibitors [INSTIs]). (See "Interpretation of HIV drug resistance testing".)

However, the risk of developing multidrug-resistant virus is much lower than in the past due to simpler regimens that are well tolerated and are less likely to induce drug resistance mutations. (See "Selecting antiretroviral regimens for treatment-naïve persons with HIV-1: General approach".)

INITIAL EVALUATION — Treatment-experienced patients are defined as those who have received at least one antiretroviral regimen in the past. The common names and abbreviations of currently available antiretroviral agents are listed by class in the table (table 1). Detailed discussions of the different types of antiretroviral therapy (ART) regimens are found elsewhere. (See "Selecting antiretroviral regimens for treatment-naïve persons with HIV-1: General approach" and "Selecting an antiretroviral regimen for treatment-experienced patients with HIV who are failing therapy" and "Overview of antiretroviral agents used to treat HIV".)

After virologic failure is confirmed (see 'Virologic failure' above), the initial assessment focuses on why the patient is failing their regimen and if drug-resistant virus is present (algorithm 1). This evaluation is best performed when a patient is on treatment, since this allows the most accurate assessment of drug resistance mutations. However, certain individuals will present off treatment, and the initial assessment includes a careful evaluation of the patient's prior treatment history and other information from past records. (See 'Patients who present off treatment' below.)

Patients who present on treatment — In patients who present with virologic failure on treatment, we try to assess if this is due to poor adherence, pharmacokinetic factors, or the development of drug-resistant virus. If poor adherence or pharmacokinetic factors are suspected, we will often try and address these first before obtaining resistance testing.

Assess medication adherence — If failure is confirmed, we first evaluate whether the patient is taking their medication correctly. If poor adherence is suspected, we assess the barriers to adherence and see if they can be modified:

●For some patients, adherence can be improved with the use of interventions such as pill packs, alarms, and/or enhanced family or caregiver support. For such patients, we try to implement these measures and repeat the viral load in one month's time. If the viral load is >200 copies/mL at that time, we perform resistance testing. (See 'Assess for drug resistance' below.)

●Some patients have behavioral barriers to adherence (eg, depression, substance use) that may require longer, more intensive interventions. In this case, we perform resistance testing and, if possible, modify their antiretroviral regimen to one that includes an agent with a high barrier to resistance (eg, a pharmacologically boosted protease inhibitor [PI] or an integrase strand transfer inhibitor [INSTI], such as dolutegravir or bictegravir). (See "Selecting an antiretroviral regimen for treatment-experienced patients with HIV who are failing therapy", section on 'Patients without drug-resistant virus'.)

Assess for poor absorption or altered metabolism — We assess all patients with virologic failure for the presence of pharmacokinetic factors that could impair drug absorption or alter metabolism. These include:

●The presence of drug-drug interactions, either within the antiretroviral regimen or with other medications the patient is currently taking (eg, proton pump inhibitors)

●If the patient is taking antiretroviral medications at the appropriate time intervals

●If dietary considerations are being followed (eg, rilpivirine with food)

●The presence of a concurrent medical condition that could impact absorption (eg diarrhea)

●If the current antiretroviral regimen is associated with inadequate trough drug levels

If identified, we modify these factors and repeat a viral load in one month's time; resistance testing should be performed if the repeat viral load is >200 copies/mL. (See 'Assess for drug resistance' below.)

Therapeutic drug monitoring is rarely performed in patients with virologic failure, since the reproducibility of these methods is highly variable [8]. If such testing is obtained (eg, in the setting of complicated drug-drug interactions, pregnancy), the results should be interpreted with expert guidance.

Assess for drug resistance — We perform resistance testing in patients with virologic failure if no adherence or pharmacokinetic barriers are identified or if interventions to improve adherence and/or potential drug interactions were not effective (ie, the viral load is still >200 copies/mL) (algorithm 1). Although genotype testing below 200 copies/mL may produce clinically meaningful data, the chance of obtaining an interpretable genotype is low [9].

Testing to assess for drug resistance is best done while the patient is on their failing regimen or within the first four weeks of discontinuing treatment. Resistance testing is less accurate if performed while off therapy, since low levels of resistant virus cannot be excluded in the absence of drug pressure. (See "Overview of HIV drug resistance testing assays".)

Genotypic testing is the preferred drug resistance assay for patients with virologic failure while taking their first or second regimen. These assays can detect mutations in the coding regions for HIV reverse transcriptase, integrase, and protease that have been associated with drug resistance. Genotypic testing is readily available and relatively inexpensive compared with phenotypic assays. However, standard genotype tests do not always evaluate for INSTI resistance, so, if appropriate, this must be specifically requested. (See "Interpretation of HIV drug resistance testing" and "Overview of HIV drug resistance testing assays".)

Additional resistance tests may be needed to assess those with multiple mutations, since interpretation becomes more complex with an increasing number of mutations, especially to PIs. (See 'Additional information for patients with highly resistant virus' below.)

Patients who present off treatment — Not infrequently, a patient will present off treatment for many months with a reported history of treatment failure and/or drug-resistant virus. In this setting, the initial approach typically involves restarting a regimen and then assessing the patient's response to therapy.

Choosing a new regimen — When selecting a regimen, we assess for prior resistance to the best of our ability and initiate a regimen with a high barrier to resistance.

●Factors informing regimen selection – We obtain a careful history to review the patient's prior regimens and do everything possible to obtain previous medical records, including prior drug resistance testing. Some experts also routinely obtain resistance testing as part of the initial evaluation.

•Antiretroviral history – We review the patient's past antiretroviral treatment history and prior resistance results.

When reviewing the patient's treatment history, we determine if the patient was just failing their prior regimen or if they had failed multiple regimens in the past. This is important to assess which drugs will likely be active. (See "Selecting an antiretroviral regimen for treatment-experienced patients with HIV who are failing therapy", section on 'Patients failing their initial regimen' and "Selecting an antiretroviral regimen for treatment-experienced patients with HIV who are failing therapy", section on 'Patients who have failed multiple regimens'.)

In addition, for certain drugs, treatment history alone can often predict HIV susceptibility. As an example, it should be assumed that documented virologic failure while adherent to medications with a low barrier to resistance (eg, emtricitabine, lamivudine, non-nucleoside reverse transcriptase inhibitors [NNRTIs], elvitegravir, raltegravir, enfuvirtide) has selected for resistance to these agents.

When reviewing prior genotypes, it is important to look at all that are available, not just the most recent, since resistance mutations that are present at one point in time may not be seen on subsequent tests. These "archived" mutations can still limit the activity of future regimens. As an example, a patient who had a change in regimen because of a resistant virus five years ago may now present with a different mutation pattern given the selective pressures of their current regimen. In this case, the clinician must take into consideration the earlier mutations, which are likely to still be present as a minority viral population, even though they may not be detected with the most recent resistance testing. (See "Interpretation of HIV drug resistance testing".)

•Patient related factors – When selecting a regimen, the presence of certain comorbid conditions impacts which agents a specific patient should receive. These include factors such as renal insufficiency, active hepatitis B viral infection, heart disease, and, if considering abacavir, the presence of the HLA-B*5701 allele. These considerations are discussed in detail in separate topic reviews. (See "Overview of antiretroviral agents used to treat HIV" and "Selecting an antiretroviral regimen for treatment-experienced patients with HIV who are failing therapy", section on 'Overview of specific antiretroviral agents' and "Selecting antiretroviral regimens for treatment-naïve persons with HIV-1: General approach", section on 'Additional considerations'.)

•Resistance testing – The use of resistance testing as part of the initial evaluation in patients who present off treatment is unclear. Testing to assess for drug resistance is best done while the patient is on their failing regimen or within the first four weeks of discontinuing treatment, since low levels of resistant virus cannot be excluded in the absence of drug pressure. Thus, some providers feel the results of resistance testing in patients who present off therapy can be misleading and difficult to interpret, whereas others find that testing can still be helpful because the presence of any resistance mutations may provide useful information about the type of resistant virus that exists.

If resistance testing is performed, we obtain a genotype. In patients who had previously taken an INSTI, testing for INSTI resistance should be performed as well. There has been increasing interest in the use of a proviral DNA assay that detects the presence of archived resistance mutations. However, this test is generally used in patients who are virologically suppressed on therapy, and even in that population, the utility of the test has not been definitively established. A more detailed discussion of resistance testing is presented elsewhere. (See "Overview of HIV drug resistance testing assays".)

●Selecting a regimen – The approach to choosing a regimen depends primarily upon the availability of prior treatment records and the presence of comorbid conditions, as described above. Resistance mutations identified on a genotype that is performed off therapy can also be used to inform regimen selection, but the absence of resistance may not be representative of the patient's full resistance profile. In these cases, review of the prior treatment history is critical.

Our general approach is as follows (algorithm 1):

•If detailed records are available, we choose a regimen based upon the prior treatment history and the results of resistance testing.

•If records are not available, we try to elicit information from the patient regarding past use of an unboosted PI (eg, nelfinavir) or an INSTI. We then initiate a regimen likely to include at least two fully active agents, at least one of which should have a high barrier to resistance (eg, boosted PI, dolutegravir, bictegravir).

•If there is insufficient information to even make this assessment, we usually initiate a regimen that includes a pharmacologically boosted PI, since this type of regimen is unlikely to select for resistance mutations beyond those that had been previously selected for. (See "Selecting an antiretroviral regimen for treatment-experienced patients with HIV who are failing therapy", section on 'General principles'.)

-We typically administer the boosted PI with a nucleoside reverse transcriptase inhibitor (NRTI) combination (eg, pharmacologically boosted darunavir with tenofovir alafenamide-emtricitabine). In patients at high risk for NRTI resistance (eg, prior treatment failure on a regimen that included a dual-NRTI combination plus an NNRTI), this combination can be administered with an integrase inhibitor that has a high barrier to resistance (eg, dolutegravir or bictegravir). Given the multiple coformulated tablets that are available, these regimens can be administered as one or two pills once daily (table 1).

-For those who are unable to take a first-line NRTI combination (eg, secondary to comorbid condition and/or reported history of toxicity), we administer a regimen that includes a pharmacologically boosted PI and dolutegravir. This regimen is a less desirable option for patients with a history of prior INSTI treatment, for which the extent of resistance is not known; however, if it is used, dolutegravir should be dosed twice daily.

-This strategy requires frequent viral load assessments (eg, monthly with low threshold to perform drug resistance testing if there is an inadequate virologic response), as outlined below. (See 'Monitoring response to treatment' below.)

Monitoring response to treatment — After initiating an ART regimen, we monitor the virologic response to therapy.

●For patients who were failing an initial regimen, we typically monitor the viral load response at four weeks and then at four- to eight-week intervals until the viral load falls below the assay's limit of detection.

●For patients who have failed multiple regimens, particularly if the past regimens are unknown, we perform viral load testing every four weeks.

●We also monitor the viral load every four weeks in patients who are started on an NRTI-sparing regimen if they have taken a PI or INSTI in the past.

If a patient experiences a poor virologic response (<1 log decrease in viral load in the first month or viral load >200 copies/mL by 12 to 24 weeks), resistance testing should then be performed while the patient is on therapy. (See "Overview of HIV drug resistance testing assays".)

If drug-resistant virus is detected, the regimen will need to be modified. A more detailed discussion of how to select an antiretroviral regimen for patients with drug-resistant virus is found elsewhere. (See "Selecting an antiretroviral regimen for treatment-experienced patients with HIV who are failing therapy".)

Evaluating barriers to adherence — When patients present off treatment, it is important to evaluate and address all barriers to adherence. This is discussed in detail above. (See 'Nonadherence' above and 'Assess medication adherence' above.)

ADDITIONAL INFORMATION FOR PATIENTS WITH HIGHLY RESISTANT VIRUS — Most patients with drug-resistant virus will require a change in their regimen to achieve sustained viral suppression. These patients should be managed in consultation with a clinician experienced in managing patients with HIV. (See "Selecting an antiretroviral regimen for treatment-experienced patients with HIV who are failing therapy".)

To determine which agents are most likely to result in virologic suppression, patients who have failed multiple regimens may need additional testing (eg, drug resistance phenotype and tropism testing), especially if the patient's virus shows resistance to multiple agents on genotype testing. (See 'Assess for drug resistance' above.)

These tests include:

●Specific genotype tests – Standard genotype tests do not always assess for resistance to integrase strand transfer inhibitors (INSTIs) and fusion inhibitors (eg, enfuvirtide). Thus, clinicians may need to request specific resistance testing for patients who experience viral failure on these agents. Such testing is not needed for patients who have never been treated with an INSTI or fusion inhibitors, since transmitted resistance with these agents is unlikely.

●Phenotypic testing – Phenotypic testing detects the fold-change in drug susceptibility of the patient's virus relative to a wild-type control strain. The utility of these assays has improved as clinical cutoffs corresponding to fold-change susceptibility have been better defined for each available drug [10]. Phenotypic testing has a longer turnaround time and is more expensive than genotypic testing. However, phenotypic testing appears to give beneficial information in the treatment-experienced patient with complex resistance patterns, particularly in the protease gene [11]. Standard phenotypic assays do not always give drug resistance testing results for fusion inhibitors and INSTIs; a separate assay may be required for each of these, if indicated. (See "Overview of HIV drug resistance testing assays", section on 'Phenotypic resistance assays'.)

●Tropism assay – A tropism assay is required if a CCR5 antagonist (ie, maraviroc) is being considered as a possible agent, since maraviroc only inhibits CCR5-utilizing virus (R5 virus). CCR5 antagonists should not be used for patients who have detectable CXCR4-using virus (X4 virus; ie, those who are reported as X4 or dual/mixed tropic). A commonly used assay assesses phenotypic tropism, but there are increasing data demonstrating the utility of genotypic tropism testing, which has a faster turnaround time and is less expensive. (See "Overview of HIV drug resistance testing assays", section on 'Tropism assays'.)

More detailed discussions of resistance testing are presented elsewhere. (See "Overview of HIV drug resistance testing assays" and "Interpretation of HIV drug resistance testing".)

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: HIV treatment in nonpregnant adults and adolescents".)

SUMMARY AND RECOMMENDATIONS

●For individuals starting a new regimen, virologic failure is defined as the inability to achieve a viral load <200 copies/mL within 24 weeks of initiating an antiretroviral regimen. For individuals who were initially able to suppress their viral load, virologic failure is defined as a recurrence of viremia to >200 copies/mL on two consecutive measurements taken approximately one month apart. (See 'Virologic failure' above.)

●Virologic failure typically results from poor adherence to an antiretroviral regimen or from drug-drug or drug-food interactions that impair absorption. Patients with virologic failure may or may not have drug-resistant virus. (See 'Reasons for virologic failure' above.)

●After virologic failure is confirmed, the initial assessment focuses on why the patient is failing their regimen and whether drug-resistant virus is present. This evaluation is best performed when a patient is on treatment, since this allows the most accurate assessment of drug resistance mutations (algorithm 1). (See 'Initial evaluation' above and 'Patients who present on treatment' above.)

●However, not infrequently, a patient will present off treatment for many months with a reported history of treatment failure and/or drug-resistant virus.

•In this setting, we do everything possible to obtain medical records. When this is not possible, we initiate an empiric antiretroviral regimen and then monitor the response to therapy (algorithm 1). For such patients, we typically prefer a regimen that includes a pharmacologically boosted protease inhibitor. (See 'Choosing a new regimen' above.)

•After initiating an antiretroviral regimen, we monitor the virologic response to therapy. If a patient experiences a poor virologic response (<1 log decrease in viral load in the first month or viral load >200 copies/mL by 12 to 24 weeks), resistance testing can then be performed while the patient is on therapy in order to define the extent of resistance to these classes of drug. (See 'Monitoring response to treatment' above.)

●Most patients who have selected for drug-resistant virus will require a change in their regimen to achieve sustained viral suppression. To determine which agents are most likely to result in virologic suppression, some patients may need additional testing (eg, drug resistance phenotype and tropism testing). (See 'Additional information for patients with highly resistant virus' above.)

ACKNOWLEDGMENTS — The UpToDate editorial staff acknowledges Meena Lagnese, MD, who contributed to an earlier version of this topic review.

UpToDate also gratefully acknowledges John G Bartlett, MD (deceased), who contributed as Section Editor on earlier versions of this topic and was a founding Editor-in-Chief for UpToDate in Infectious Diseases.