1. Prophylaxis Against Pneumocystis Pneumonia

Patients with CD4 counts below 200 cells/mcL, a CD4 lymphocyte percentage below 14%, or oral candidiasis should be offered primary prophylaxis for Pneumocystis pneumonia. Patients with a history of Pneumocystis pneumonia should receive secondary prophylaxis until their viral load is undetectable and they have maintained a CD4 count of 200 cells/mcL or more while receiving ART for longer than 3 months. Regimens for Pneumocystis prophylaxis are given in Table 33-6.

2. Prophylaxis Against M Avium Complex Infection

Prophylaxis against M avium complex is no longer recommended in most individuals who are initiating ART, including in those with CD4+ counts less than 50 cells/mcL. The incidence of M avium complex infection is very low among those on ART. In rare cases where individuals delay ART (eg, per patient preference) whose CD4 counts fall to below 50 cells/mcL, prophylaxis against M avium complex infection should be offered. Clarithromycin (500 mg orally twice daily) and azithromycin (1200 mg orally weekly) are the recommended regimens. The azithromycin regimen is generally preferred based on high adherence and low cost. Prophylaxis against M avium complex infection may be discontinued in patients who initiate ART.

3. Prophylaxis Against Toxoplasmosis

Toxoplasmosis prophylaxis is desirable in patients with a positive IgG toxoplasma serology and CD4 counts below 100 cells/mcL. Trimethoprim-sulfamethoxazole (one double-strength tablet daily) offers good protection against toxoplasmosis, as does a combination of pyrimethamine, 25 mg orally once a week, plus dapsone, 50 mg orally daily, plus leucovorin, 25 mg orally once a week. A glucose-6-phosphate dehydrogenase (G6PD) level should be checked prior to dapsone therapy, and a methemoglobin level should be checked at 1 month. Prophylaxis can be discontinued when the CD4 cells have increased to greater than 200 cells/mcL for more than 3 months.

1. Nucleoside and Nucleotide Reverse Transcriptase Inhibitors

There are currently six NRTI agents available (counting TDF and TAF as separate agents) for use. The choice of which agent to use depends primarily on the patient's prior treatment experience, results of resistance testing, medication side effects, other underlying conditions, and convenience of formulation. However, most clinicians use fixed-dose combinations (see Table 33-8. Fixed-Dose Antiretroviral Combinations (Listed in Alphabetical Order by Brand Name)) of including either emtricitabine/TDF, emtricitabine/TAF, or abacavir/lamivudine (ABC/lamivudine), all of which can be given once a day. Abacavir should be given only to HLA-B^* 5701-negative persons due to a risk of hypersensitivity in those who are HLA-B^* 5701-positive. In patients with viral loads greater than 100,000 copies/mL, ABC/lamivudine was less effective than emtricitabine/TDF when combined with efavirenz or ritonavir-boosted atazanavir. However, ABC/lamivudine appears to be equally efficacious as emtricitabine/TDF in patients with viral loads greater than 100,000 copies/mL when combined with dolutegravir. In some studies, abacavir increased risks of MI, and therefore should be avoided in patients at high risk for CVD. Zidovudine is now rarely used due to toxicity (eg anemia or neutropenia). Of the available agents, zidovudine is the most likely to cause anemia or neutropenia. Emtricitabine, TDF, TAF, and lamivudine have activity against hepatitis B. TDF, TAF, emtricitabine, abacavir, and lamivudine can be administered once daily. Information specific to each medication is given below and in Table 33-7. Antiretroviral Therapy Agents by Class (Listed in Alphabetical Order Within Classes).

2. Nonnucleoside Reverse Transcriptase Inhibitors

NNRTIs inhibit reverse transcriptase at a site different from that of the nucleoside and nucleotide agents described above. The major advantage of NNRTIs is that four of them (efavirenz, rilpivirine, doravirine, and nevirapine) have potencies comparable to that of PIs (next section), at least for patients with viral loads under 100,000 copies/mL-with lower pill burden and fewer side effects. However, they have lower barrier to resistance when compared to PIs and integrase inhibitors. Unlike PIs, they do not cause lipodystrophy and do not seem to cause weight gain; patients with cholesterol and triglyceride elevations who are switched from a PI to an NNRTI may have improvement in their lipids. The resistance patterns of NNRTIs are distinct from those of the PIs. Because these agents may cause alterations in the clearance of PIs, dose modifications may be necessary when these two classes of medications are administered concomitantly. There is a high degree of cross-resistance among the "first-generation" NNRTIs, such that resistance to one medication in this class uniformly predicts resistance to other medications. However, the "second-generation" NNRTIs etravirine, rilpivirine, and doravirine can have consistent antiviral activity in patients with prior exposure and resistance to nevirapine or efavirenz, although genotypic analysis is needed first in these contexts. In particular, the K103N pathologic variant does not have an impact on etravirine, doravirine, or rilpivirine. There is no therapeutic reason for using more than one NNRTI at the same time.

3. Protease Inhibitors

Four PIs-ritonavir, lopinavir (in combination with ritonavir), atazanavir, and darunavir-are still available (with others used rarely). Darunavir boosted with ritonavir or cobicistat is the most commonly used PI in the United States. PIs are potent suppressors of HIV replication and are administered as part of a combination regimen.

All of the PIs-to differing degrees-are metabolized by the cytochrome P450 system, and each can inhibit and induce various P450 isoenzymes. Therefore, medication interactions are common and difficult to predict. Clinicians should consult the product inserts before prescribing PIs with other medications. Medications that are known to induce the P450 system, such as rifampin, should be avoided.

The fact that the PIs are dependent on metabolism through the cytochrome P450 system has led to the use of ritonavir to boost the medication levels of other PIs, allowing use of lower doses and simpler dosing schedules of these PIs. A second boosting agent, cobicistat, is coformulated with the PI atazanavir (Evotaz) and darunavir (Prezcobix and Symtuza). Similar to ritonavir, cobicistat also inhibits liver enzymes that metabolize other HIV medications.

All PIs have been linked to varying degrees of metabolic side effects, including elevated cholesterol levels, elevated triglyceride levels, insulin resistance, diabetes mellitus, and changes in body fat composition (eg, abdominal obesity). The lipid abnormalities and body habitus changes are referred to as lipodystrophy. Although lipodystrophy is commonly associated with PIs, it has been seen also in people with HIV who have never been treated with these agents. In particular, the lipoatrophy effects seen in patients receiving ART appear to be more related to the nucleoside toxicity and in particular to the thymidine analogs (ie, zidovudine).

Of the different manifestations of lipodystrophy, the dyslipidemias that occur are of particular concern because of the likelihood that increased levels of cholesterol and triglycerides will result in increased prevalence of heart disease. All patients taking PIs should have fasting serum cholesterol, LDL cholesterol, and triglyceride levels assessed. Clinicians should assess for CHD (see Part 30) and consider initiating dietary changes or medication therapy (or both). PIs inhibit statin metabolism. Lovastatin and simvastatin should be avoided. In general, the least interaction is with pravastatin (20 mg daily orally). Atorvastatin (10 mg daily orally) or rosuvastatin (5 mg/day orally initially; maximum 10 mg/day) may also be used cautiously. Patients with persistently elevated fasting serum triglyceride levels of 500 mg/dL or more who do not respond to dietary intervention should be treated with one of the statin medications, followed by gemfibrozil (600 mg twice daily prior to the morning and evening meals). PIs are associated with abnormalities in cardiac conduction, especially prolongation of the PR interval.

4. Integrase Inhibitors

Integrase inhibitors slow HIV replication by blocking the HIV integrase enzyme needed for the virus to multiply. They are now the preferred regimens for initiating therapy because of the combination of efficacy, ease of administration, and low incidence of side effects. Five integrase inhibitors are currently available: raltegravir; elvitegravir; dolutegravir; bictegravir; and cabotegravir, which is given via injections along with injections of rilpivirine every 4 or 8 weeks. Clinical trials of available integrase inhibitors reveal a consistent pattern of more rapid decline in viral load compared with more standard PI/r or NNRTI-based regimens. Integrase inhibitors are effective (when combined with other active medications) in the treatment of people with HIV with documented resistance to each of the three main classes of antiretroviral medications (nucleoside analogs, PIs, NNRTIs). Avoid administering oral integrase inhibitors with antacids or other medications with divalent cations (Ca²⁺ , Mg²⁺ , Al²⁺ , Fe²⁺ ) because chelation of the integrase inhibitor by the cation reduces absorption. When these medications must be taken with integrase inhibitors, they should either be taken together with food or the integrase inhibitor taken 2 hours before divalent cations. Integrase inhibitors have been associated with weight gain, with or without tenofovir alafenamide.

5. Entry and Fusion Inhibitors

6. Attachment Inhibitor

7. Capsid Inhibitor

8. Constructing an Initial Regimen

Guidelines for starting ART are shown in Table 33-9. Recommended and Alternative Initial Antiretroviral Therapy Regimens (Listed in Alphabetical Order Within Categories). The regimens with the strongest evidence all contain integrase inhibitors. This reflects their high efficacy, high barrier to resistance, tolerability, low pill burden, and safety profile. The two best-tolerated, high barrier to resistance integrase inhibitors are bictegravir and dolutegravir and so they form the backbone of the recommended regimens. Integrase inhibitors are typically given with a backbone of two NRTIs (although see discussion of two-drug therapy below). With respect to NRTIs, the most commonly used combination is TDF or TAF with emtricitabine or lamivudine. From an efficacy standpoint, there is no difference between TDF or TAF; for patients on a boosted regimen (ie, one including cobicistat or ritonavir), those with renal dysfunction or osteoporosis or osteopenia (or risk for these conditions) should receive TAF. However, TAF is associated with increased weight gain and dyslipidemia, so TDF may be considered in individuals with obesity or dyslipidemia. Also, choosing between TDF and TAF may depend on convenience: which one is coformulated with other desired partner drugs. Emtricitabine and lamivudine are essentially the same from the point of view of efficacy and side effects. The next most commonly used two-drug NRTI backbone is abacavir and lamivudine (coformulated as part of Epzicom or with dolutegravir in Triumeq). Given the need to perform B^* 5701 allele and hepatitis B testing, abacavir-containing regimens are not appropriate for rapid ART initiation.

The only two-drug regimen approved for initial ART is dolutegravir plus lamivudine, which is available as a single pill for once-a-day treatment (Dovato, Table 33-8. Fixed-Dose Antiretroviral Combinations (Listed in Alphabetical Order by Brand Name)). (All other recommended first-line initial regimens contain three medications, sometimes with a fourth agent as a booster.) This two-drug regimen is not recommended for patients with high HIV viral load (greater than 500,000 copies/mL), or for patients with HBV coinfection (because of development of resistance by the hepatitis B virus to lamivudine when used as monotherapy with possibility of severe flares of hepatitis), or for patients for whom the results from HIV resistance or HBV testing are not yet available. There is also concern about its use in patients with CD4 cell counts less than 200/mL. Dolutegravir/rilpivirine (Juluca) and injectable cabotegravir/rilpivirine (Cabenuva) are approved only in the setting of prior virologic suppression without prior treatment failure or drug resistance, although there are promising observational data regarding its use among patients who are not virologically suppressed but do not have resistance to its components and cannot take oral ART.

Studies have shown dolutegravir/abacavir/lamivudine to be superior to efavirenz/TDF/emtricitabine and have shown dolutegravir to be superior to ritonavir-boosted darunavir (both combined with either abacavir/lamivudine or TDF/emtricitabine). A network meta-analysis adjusting for NRTI backbone found that dolutegravir had superior efficacy in suppressing viral load compared with regimens containing ritonavir-boosted atazanavir, ritonavir-boosted darunavir, efavirenz, or ritonavir-boosted lopinavir. Discontinuation due to adverse events was also statistically lower with the dolutegravir regimens.

The emergence of generic antiretroviral medications is also likely to affect prescribing choices when equally effective regimens are available at different costs. There are generic versions available for abacavir, atazanavir, efavirenz, lamivudine, and TDF. But how this will affect the costs patients pay can be very difficult to determine because of complicated copay rules.

For patients who cannot take an integrase inhibitor, alternative regimens are recommended (Table 33-9. Recommended and Alternative Initial Antiretroviral Therapy Regimens (Listed in Alphabetical Order Within Categories)).

Resistance testing should be performed prior to starting ART. Of people with newly diagnosed infections in some urban areas of the United States, 8-10% have transmitted drug resistance (most commonly to NNRTIs followed by NRTIs).

The most important determinant of treatment efficacy is adherence to the regimen. Therefore, it is vitally important that the regimen chosen be one to which the patient can easily adhere (Figure 33-7). In general, patients are more adherent if their medication regimens (Table 33-9. Recommended and Alternative Initial Antiretroviral Therapy Regimens (Listed in Alphabetical Order Within Categories)) offer complete therapy in one pill that needs to be taken only once a day, do not require special timing with regard to meals, can be taken at the same time as other medications, do not require refrigeration or special preparation, and do not have bothersome side effects. Given the high level of effectiveness of recommended regimens, patients for whom the viral load does not fully suppress are likely to be encountering adherence challenges. Pharmacists and other specially trained clinicians can be very effective in helping patients improve their adherence by taking the time to understand why patients miss their medications and to problem solve (eg, take medicine at same time every day, keep a supply in the car or at work in case they forget). Demonstration projects are examining long-acting cabotegravir and rilpivirine as injectables in people challenged by adherence or viremia, with high virologic suppression achieved in one demonstration project in San Francisco, with 97.5% virologic suppression projected. For certain populations (eg, unstably housed individuals), specially tailored programs may be beneficial.

Figure 33-7. Approach to Monitoring Initial and Subsequent Antiretroviral Therapy

Approach to monitoring initial and subsequent antiretroviral therapy.

1. Goals of Monitoring Art

Monitoring of ART (Figure 33-7) has two goals: evaluate for toxicity and measure efficacy using objective markers to determine whether to maintain or change regimens. Laboratory evaluation for toxicity depends on the specific medications in the combination but generally should be done approximately every 3-6 months once a patient is on a stable regimen. Patients who are intolerant of their initial regimen should be changed to one of the other initial recommended or alternative regimens in Table 33-9. Recommended and Alternative Initial Antiretroviral Therapy Regimens (Listed in Alphabetical Order Within Categories). The second aspect of monitoring is to measure HIV viral load, the objective marker of efficacy. The HIV viral load should be repeated 2-4 weeks after the initiation or change of antiretroviral regimen and every 3-6 months thereafter in clinically stable patients. With integrase regimens, a two-log₁₀ reduction is expected within 2 weeks of starting therapy, and approximately 80% of patients will have undetectable HIV viral load at 1 month. All patients should have undetectable viral loads by 3 months; if not, the usual problem is nonadherence (see below). CD4+ counts are most useful in determining the immunologic response to ART, although the response can vary significantly, and determining whether opportunistic infection prophylaxis can be discontinued. CD4+ counts should be monitored approximately every 3-6 months in individuals newly initiating ART. For those who have consistently suppressed HIV viral loads over the first 1-2 years of therapy with a CD4+ count greater than 300 cells/mcL, monitoring can occur yearly, and is optional in those with CD4+ counts greater than 500 cells/mcL.

2. The Challenge of Medication Adherence

In a patient who is adherent to an integrase inhibitor regimen, viral load should drop by 100-fold within 2 weeks. For patients in whom viral loads do not decrease adequately, or who have viral rebound after suppression, the major question facing the clinician is whether the patient is nonadherent or has resistance to the regimen, or both. Patients who are having trouble adhering to their treatment should receive adherence counseling. In patients who are adherent or who have missed enough doses to make resistance possible, resistance testing should be performed. Based on the results of resistance testing, if there is no treatment-emergent resistance, and the patient is tolerating the regimen well, the patient should continue the regimen with assessment of potential barriers to adherence (ie, mental health, substance use, medication coverage, or housing challenges). If resistance has emerged, the patient should be switched to a high-barrier-to-resistance regimen with at least two, but ideally three, active agents (ie, three-drug dolutegravir, bictegravir, or boosted darunavir-based regimen).

Once ART has been initiated, it is not advisable to stop the therapy. So-called drug holidays or structured treatment interruptions are not recommended because they have been shown to increase risk of AIDS-related complications, increase CD4 declines, and increase morbidity from non-AIDS-related complications (eg, MIs and liver failure).

3. The Challenge of Medication Resistance

Resistance to HIV-1 medication has been documented for all currently available antiretrovirals. Although HIV medication resistance was common in the past, high-level resistance has been declining in the past few years, likely related to better tolerated, easier to use, and more efficacious antiretroviral agents. Resistance also occurs in patients who are ART-naïve but who are infected with a medication-resistant strain-called "primary" or "transmitted" medication resistance. Cohort studies of ART-naïve patients entering care in North America and Western Europe show that roughly 8-10% of people with a recent infection have a medication-resistant strain of HIV-1.

In addition to being part of a standard baseline evaluation, resistance testing is recommended for patients who are receiving ART and have suboptimal viral suppression (typically a viral load greater than 500 copies/mL is needed for resistance testing be performed). Both genotypic and phenotypic resistance tests are commercially available, and in randomized controlled studies, genotypic resistance testing has been shown to result in improved short-term virologic outcomes compared to making treatment choices without resistance testing. Furthermore, multiple retrospective studies have conclusively demonstrated that resistance tests provide prognostic information about virologic response to newly initiated therapy that cannot be gleaned from standard clinical information (ie, treatment history, examination, CD4 count, and viral load tests).

Because of the complexity of resistance tests, many clinicians require expert interpretation of results. In the case of genotypic assays, results may show that the pathologic variants that are selected for during ART are medication-specific or contribute to broad cross-resistance to multiple medications within a therapeutic class. An example of a medication-specific pathologic variant for the reverse transcriptase inhibitors would be the M184V pathologic variant that is selected for by lamivudine or emtricitabine therapy-this pathologic variant causes resistance only to those two medications. Conversely, the thymidine analog pathologic variants ("TAMs") of M41L, D67N, K70R, L210W, T215Y/F, and T219Q/K/E are selected for by either prior zidovudine or stavudine therapy, but cause resistance to all the medications in the class and often extend to the nucleotide inhibitor tenofovir when three or more of these TAMs are present. The combination of M184V and K65R is important to know as it causes resistance to all NRTIs except zidovudine and would typically necessitate change to an NRTI-sparing regimen. The most common pathologic variants associated with medication resistance and cross-resistance patterns for NRTIs, NNRTIs, PIs, and integrase inhibitors can be found at http://hivdb.stanford.edu (see specific references below). Phenotypic resistance testing can provide complementary data and is usually provided with an interpretation about susceptibility to the virus.

Resistance results may be misleading if a patient is not taking antiretroviral medications at the time of testing because the dominant virus is likely the wild-type, even if there are resistant viruses in the body that can become dominant with the selective pressure of antivirals. Thus, resistance results do not replace a careful history of what medications a patient has taken in the past and for how long. Also, the results of resistance testing should be viewed cumulatively-ie, if resistance is reported to an agent on one test, it should be presumed to be present thereafter even if subsequent tests do not give the same result.

Stanford University HIV Drug Resistance Database Home Page (http://hivdb.stanford.edu/) provides Genotypic Resistance Interpretation Algorithm, HIVdb Program, version 9.0, February 22, 2021 (http://hivdb.stanford.edu/hivdb/by-mutations/); Genotype-Phenotype Correlations (http://hivdb.stanford.edu/pages/genotype-phenotype.html); Genotype-Treatment Correlations (http://hivdb.stanford.edu/pages/genotype-rx.html); Genotype-Clinical Outcome Correlations (http://hivdb.stanford.edu/pages/genotype-clinical.html).