53 HIV: management

Highly active antiretroviral therapy (HAART) has produced dramatic improvements in the prognosis of HIV infection with ↓ rates of mortality and opportunist infections (OIs).

Important principles to consider in those under regular review are as follows.

Patient’s commitment to starting treatment is essential. Pre-treatment adherence education is vital as adherence must be >95% to obtain both maximum magnitude and duration of antiviral effect. HAART rarely needs starting as an emergency.

Drug resistance should be assayed pre-treatment, ideally at diagnosis, because of possible resistant viral transmission.

Prime indication for treatment is severe symptoms (which may indicate risk of more rapid progression to symptomatic disease). These may resolve with HAART. Treatment duration is undefined and its influence on subsequent clinical course is uncertain.

In the asymptomatic patient, the likelihood of developing clinical AIDS over a 3 year period can be predicted from a combination of viral load (VL) and CD4 count (Fig. 53.1). CD4 counts give the best indication of OI risk but high VLs are associated with more rapid rates of CD4 ↓.

HAART should be offered:

Other factors, e.g. older age (↑ speed of development of immuno-deficiency) should also be considered. In pregnancy, treatment may be given to ↓ risk of mother-to-child transmission when such therapy may not be indicated in the non-pregnant state.

Starting HAART at very low CD4 counts ↑ drug side-effects, although with nevirapine hepatotoxicity is ↑ by CD4 counts >250cells/μL. In patients with active OIs commencement of HAART should not be deferred for a long time. Where there are significant drug interactions such as in TB/HIV co-infection HAART can be delayed if CD4 >200 cells/μL until completion or simplification of treatment. Once started, treatment (except when given for fetal protection) should be continued indefinitely.

Initial treatment influences the pattern of resistance mutations. If it fails, the range of drugs available for 2nd-line therapy is restricted. Therefore the following are important aims for the first treatment.

Drug interations between antiretrovirals and other drugs the patient is taking or is likely to take must be considered, all drugs should be checked against a reliable database.

Drug combinations in the treatment naive should take account of resistance testing, co-pathologies, risk factors for diabetes and cardiovascular disease, lifestyle, and informed patient choice.

Essential for successful viral suppression. Stress importance prior to starting treatment and continually reinforce as adherence may diminish with time.

Sub-therapeutic drug levels select for HIV-resistant mutations arising from error-prone viral replication. Some regimens have a low genetic barrier to resistance. Investing in strategies that improve adherence is more cost effective than managing the consequences of poor compliance. Such strategies will also minimize the risk of transmission of drug-resistant virus which may adversely affect HAART response in the newly infected.

A multifactorial approach should be adopted, taking account of the patient’s perceptions of the benefits as well as the practicalities of treatment. The individual’s commitment to taking drugs should be assessed before starting therapy and at regular intervals. Concerns about drug side-effects should be explored. Motivational techniques can help patients to strengthen their intentions and adherence behaviour. Psychosocial aspects including relationships, alcohol and drug use, housing, employment, and immigration status should be considered, with appropriate professional involvement. The following measures may improve adherence:

There is no ideal and accurate method of monitoring adherence. Self-reporting, pill counting, self-completed questionnaires, and drug levels have all been used with varying success.

Current classes of antiretroviral drugs inhibit the virus at different stages of its cellular lifecycle (Fig. 53.2):

Triple combinations are the standard of care. The drug classes have differing side-effects, drug interactions, and impact on co-pathologies. Individual drugs, within classes, may have significantly better convenience, tolerability, or side-effect profiles than others, e.g. atazanavir (PI) ↓ effect on lipids and once-daily therapy.

Currently most treatment-naive patients are started on NNRTI regimens. Efavirenz and nevirapine have different side-effect profiles but are of equivalent potency. Nevirapine is associated with skin rash and hepatic reactions (↑ risk in women with CD4 >250 cells/μL and men with CD4 >400 cells/μL), and efavirenz with neuropsychological side-effects and potential teratogenicity.

Side-effects are very common with HAART, although most resolve spontaneously after a few weeks. Patients should be asked to report side-effects and be given written information about serious complications, e.g. abacavir hypersensitivity reactions. Consideration should be given to providing anti-emetics and anti-diarrhoeal agents for the common early GI symptoms. Sedatives may be needed for severe EFV-associated insomnia/vivid dreams.

Both CD4 and VL should be checked and repeated at 4 weeks and 12 weeks into therapy and then at 3 monthly intervals if virological control is achieved.

Adherence reinforcement should be undertaken at each visit. Risk of acquisition of co-infections and STIs should be assessed and tested for when appropriate.

Up to 35% of patients taking PIs have sub-optimal drug concentrations with ~50% developing virological failure. A fixed drug dose may not be appropriate for all patients. Measuring drug levels to determine therapeutic dose may help promote durable viral suppression and ↓ resistance. Dose–response and concentration–response relationships have been identified for PIs and NNRTIs (some data for NRTIs).

Only of value if highly adherent (>95%) and at steady-state conditions (after at least 14 days of therapy). Blood samples should be obtained at the end of the dosing interval, as close to minimum concentration (C_min) as possible, to enable comparison with product monograph concentrations. TDM can be of benefit in the following.

See Table 53.1

A mechanism by which NRTIs may cause myopathy, peripheral neuro-pathy, hepatic steatosis, lactic acidosis, and, in infants, neurological disease. The link is strongest with lactic acidosis which has been reported in infants born to mothers receiving AZT or 3TC + AZT. D4T + DDI has been associated with lactic acidosis in pregnancy and should be avoided/switched. Caused by inhibition of mitochondrial γ DNA polymerase, the enzyme responsible for DNA synthesis, but effects on other mitochondrial enzymes may contribute. D4T and DDI inhibit γ DNA polymerase, whereas AZT inhibits other mitochondrial enzymes. Hence toxicity induced by D4T improves on switching to AZT or ABC.

Clinical significance of isolated hyperlactataemia (venous lactate between 2.5 and 5.0mmol/L) unknown. Routine measurement of venous lactate and anion gap in asymptomatic patients not recommended.

Characterized by arterial pH <7.35 and venous lactate >5mmol/L (sample taken without tourniquet into tube containing fluoride-oxalate, transported immediately on ice to laboratory). Occurs most frequently with D4T and in ♀. Usually develops after several months of treatment. Main features are nausea, vomiting, weight loss, fatigue, abdominal pain, tender hepatomegaly, and respiratory failure. Laboratory findings:

Diagnosis must be considered in any patient presenting with nausea, vomiting, abdominal pain, and abnormal LFTs. Essential to discontinue antiretroviral medication and exclude other causes. Supportive therapy required with fluid replacement, oxygen therapy, and, if necessary, assisted ventilation and haemodialysis. Benefit from carnitine, thiamine, co-enzyme Q, and riboflavin is limited.

Risk of drug-induced peripheral neuropathy ↑ with HIV disease progression. Reported with NRTIs; more frequent with D4T and DDI. Presents with distal symmetric polyneuropathy (DSP), which may be difficult to differentiate from HIV-related DSP but tends to be painful, more sudden, and progressive. Discontinuation of the offending NRTI may result in improvement, but symptoms may deteriorate for several weeks. Pain relief may be obtained with acetyl-L-carnitine, tricyclic antidepressants, anticonvulsants (gabapentin and lamotrigine), and recombinant human nerve growth factor.

Reported with all classes of antiretrovirals. ↑ in ♂ and those with other predisposing risk factors, e.g. excessive alcohol consumption, HBV and HCV infections. Abnormal LFTs graded from 1 (ALT 2–3 × upper limit of normal) to 4 (ALT >10 × upper limit of normal). Minor abnormalities do not require intervention apart from monitoring.

Abnormal LFTs found in those on antiretroviral treatment:

Careful history including alcohol intake. Screen for HBV and HCV. Important to measure LFTs before starting HAART. Asymptomatic ↑ of ALT (especially grade 1–3) does not normally require any action apart from close monitoring, exclusion, and treatment of underlying aggravating factors. Isolated hyperbilirubinaemia 2° to IDV and ATV not clinically significant.

Specific action required with:

Most commonly implicated NRTIs are DDI (up to 7%) and D4T. They should not be combined. Possibly caused by mitochondrial toxicity or direct toxic effect of the pancreas. Risk ↑ in♀, especially with CD4 <200cell/μL, excessive alcohol use, and nutritional deficiencies.

Presents with acute abdomen ± nausea and vomiting. Differentiate from other causes of acute abdomen, e.g. cholecystitis and intestinal obstruction. Serum amylase usually ↑; may be normal (can also be ↑ in other causes of acute abdomen). DDI can ↑ salivary amylase, usually associated with sicca syndrome. Diagnostic accuracy higher if serum lipase also ↑. Ultrasound and CT scan of the abdomen may help to establish diagnosis, extent of disease, and complications, e.g. pancreatic abscess, pseudocyst.

Monitor circulatory, renal, and liver functions (in a high dependency unit if necessary). Support with analgesia, fluid, and nutrition. Stop anti-retrovirals or substitute the implicated drug with TDF or a non-NRTI regimen.

Skin reactions are common. Most frequently found with NNRTIs (NVP 16%, EFV 4%) and ABC 8%. NVP-induced rash typically occurs in first 2 weeks of therapy. An induction dose of half the maintenance dose for 2 weeks minimizes this risk. Typically maculopapular, affecting the trunk. Systemic symptoms occur, with more severe reactions seen especially with ABC. Toxic epidermal necrolysis and Stevens–Johnson syndrome reported in 0.5%.

Mild skin rash does not require intervention and will often settle spontaneously (with continuation of antiretrovirals). Antihistamines may be needed for symptomatic relief. More severe reactions need drug switching, e.g. EFV may replace NVP.

Occurs in ~4%, usually in the first 6 weeks of therapy (94%)—median 11 days. Normally presents with ≥2 of following features: GI tract (nausea, vomiting, diarrhoea, pain), headache, fever, malaise, maculopapular or urticarial rash, abnormal LFTs, myalgia, dyspnoea, cough, respiratory distress, and eosinophilia. Once suspected, ABC should be stopped promptly and supportive treatment instituted. Symptoms, except rash, resolve in 24–48 hours.

ABC should not be used again as mortality from re-challenge is 4%.

( Chapter 47, Metabolic disorders p. 539.)

CD4 count ↑ on HAART mainly due to CD4 memory cells in first 4 months then followed by ↑ naive cells associated with ↓ CD4 activation markers due to ↓ viral replication. Initial phase of CD8 ↑ followed by a second phase of ↓. Most studies demonstrate ↓ HIV-specific immune responses and ↓ CD8 responses towards HIV. This contrasts with ↑ immune responses against other pathogens. ↓ incidence of OIs in patients who have higher CD4 counts from HAART. HIV damages thymus and lymphoid tissue at an early stage and may ↓ immune recovery. ↓ immune response may be due in part to failure of the thymus, as demonstrated by ↓ thymus emigrants measured in peripheral blood. The lower the CD4 nadir, the slower and less complete immune reconstitution is likely to be.

Inflammatory response induced by HAART occurring at sites of clinical and subclinical disease, usually seen in patients with CD4 count <100cells/μL at the initiation of therapy. The enhanced immunity is the likely mechanism converting a subclinical infection to an apparent symptomatic one due to the expansion of CD4 memory cells. Examples of IRIS include the following.

Treatment is usually started when the CD4 level ↓, ideally below 350cells/μL and certainly before it reaches 200cells/μL. The aim of treatment is to suppress viral replication, measured by the viral load (VL), and this should later be followed by ↑ in CD4 count. A combination of three antiretroviral drugs is usually used termed HAART (highly active antiretroviral therapy).

The HIV virus can develop resistance by mutating so that it can replicate despite the antiretroviral treatment. This most commonly arises when medication is not taken reliably. If a person is infected with a drug-resistant strain of HIV virus, their treatment options will be ↓ (some times severely).

Yes. It is important to practice safe sex to avoid superinfection (becoming infected with another strain of HIV) which may adversely affect the immune system and carry drug resistance.

Yes. It is very important that you take your anti-HIV drugs regularly. If doses are missed, the virus may not be suppressed and there is ↑ risk of viral mutations and drug resistance.

No. An undetectable VL does not mean that there is no virus in the blood. It just means that there are too few particles to be detected by the test. There is still a risk of transmission with a low VL and so you must continue to practice safe sex.

Current therapy has had a dramatic effect in improving the wellbeing and life expectancy of people with HIV. Effective therapy makes HIV a chronic rather than a life-threatening infection. Therefore it is likely that you may not develop AIDS. Some people never need treatment, but if it is started it must be taken consistently and probably for life.

May be intrinsic, e.g. HIV-2 resistance to NNRTIs, or acquired as a result of mutations in viral proteins targeted by antiretroviral agents. Two factors drive mutations: high rate of viral replication (10^8–10 virions produced daily) and error-prone reverse transcription (1 base pair substitution, deletion, insertion, recombination for every genome transcription). Examples of resistance mutations are shown in Table 53.2

1° mutation predates antiretroviral treatment which selects for it. 2° mutation develops during HAART and may be additive to 1° mutation. HAART ↓ development of resistance by suppressing viral replication and thus generation of new variants. It can also suppress existing mutants if they are not resistant to all drugs in the regimen. However, resistance emerges if drug levels are insufficient to block viral replication but high enough to exert a positive selective pressure on these mutants. Even with undetectable VL, low-level replication may allow resistance to develop. Drug resistance has been demonstrated in up to 25% of patients on HAART. Compensatory (2°) mutation reverses ↓ viral fitness resulting from other mutations. Some mutations induce resistance to certain agents while simultaneously producing hypersusceptibility to others (e.g. M184V ↑ sensitivity to AZT).

Many mutations ↓ viral fitness. However, resistance mutations confer a selective advantage by ↓ susceptibility to antiviral agents, thereby enabling the mutant quasi-species to proliferate under treatment with those agents. Resistance mutations are described using a number referring to the affected codon (group of three nucleotides coding for an amino acid). A letter may be added after the number to denote the amino acid in the mutant, e.g. 74V (Valine). This may be coupled with another letter preceding the number to show the wild-type amino acid, e.g. L74V (Leucine → Valine). Resistance develops rapidly (within weeks of commencing treatment) if only a single mutation is required e.g. M184V (3TC, FTC) and K103N (NVP). It evolves more gradually if multiple mutations are required, e.g. AZT, ABC, or PIs. Additional requirements may include a compensatory mutation, e.g. 30N (NFV).

K65R mutation (selected by TDF, ABC and DDI) confers resistance to TDF, ABC, and 3TC and ↑ susceptibility to AZT and D4T. This mutation develops rapidly when regimens combining TDF with two of ABC, DDI or 3TC are given to the treatment naive. Coexistence of K65R and M184V ↑ resistance to ABC and DDI but retains susceptibility to TDF, AZT, and D4T. Regimens with TDF must include AZT or a PI/NNRTI. Multiple mutations may interact. Resulting resistance patterns can be predicted by matching with resistance profile databases. This is provided by commercial resistance tests (e.g. Virtual Phenotype^TM).

Databases of resistance profiles are available at:

When treatment that selected for resistant quasi-species is discontinued, wild-type virus usually becomes predominant within 2 months. Drug-resistant mutants occasionally remain dominant, e.g. 41L (zidovudine), but usually cease to be detectable by standard assay. However, they may still persist as minority quasi-species, e.g. 90M (PI), or latent integrated proviral DNA (archived resistance). Therefore standard assays may not exclude drug resistance if carried out >1 month after stopping a failing regimen. Interpretation of resistance mutations must take into account previous treatment history including evidence of viral persistence.

Standard resistance assays require VL of ≥1000copies/mL and cannot detect minority species. Expert advice is needed to interpret the results.

Viral genes are sequenced to identify key mutations known to confer (alone or with others) resistance. Current methodology only detects viral mutants comprising at least 20–30% of the total population. Analysis is based on known correlation between genotype and phenotype from previous studies. Results are normally available in 2–4 weeks.

Viral cell cultures are set up with increasing concentrations of anti-retroviral drugs to determine IC₅₀, the concentration of drug required to inhibit viral replication by 50%. Cut-off value indicates by what factor the IC₅₀ of an HIV isolate can be ↑ while still being classified as susceptible (when compared with a wild-type control). IC₅₀ above this value indicates resistance. Usually takes longer than genotyping.

P-glycoprotein (P-gp) and multidrug-resistance associated protein 1 (MRP1) are human cell membrane constituents known as efflux pumps. They are found in the lining of intestine, renal tubules, biliary canaliculi, capillaries in brain, testes, placenta, stem cells, lymphocytes, and macrophages. Their function is to protect tissues by actively transporting foreign substance out of cells. Cell membrane expression of P-gp/MRP1 and resulting activity of efflux pump vary depending on genetic polymorphism and induction or inhibition by various factors including hiv infection (↑ P-gp in advanced stage). PIs may be subject to efflux action resulting in ↓ absorption (intestinal P-gp) or ↓ levels in CD4 cells leading to ↓ response to treatment with ↑ likelihood of resistance mutations.

Now recommended soon after diagnosis of HIV infection. It is particularly required in the following situations:

Studies suggest that HAART achieves better viral suppression when guided by resistance testing (with expert interpretation).

Treatment switches may be required for intolerance, side-effects, metabolic disorders, or virological failure (defined as viral rebound or failure to achieve initial viral suppression).

In patients where new three-drug options (that are likely to fully suppress viral replication) are available, switches should be considered when there have been two or more consecutive viral loads >400copies/mL having excluded other explanations (e.g. intercurrent infection). Also important to exclude poor adherence or factors leading to ↓ drug levels before switching. Viral resistance testing should be done. Patients should always be switched onto at least 2 active drugs and where possible 3. Single drug switches can be made for drug-related problems (e.g. side-effects) if there is satisfactory viral suppression.

In the absence of resistance data, first treatment switches are influenced by initial therapy. First treatment failure options to consider:

Failing regimens may be continued if no viable treatment change possible. If viral load moderate and immune function stable, residual antiviral activity is likely to be beneficial.

The following may be considered:

Pathogenesis of HIV is complex, involving interactions between virus and immune system. Precise immune control of HIV infection is not fully understood. HIV infection is characterized by ↑ production of certain cytokines (e.g. IL-1, IL-6, tumour necrosis factor) and ↓ production of others (e.g. IL-2, IL-12, and interferon γ).

HAART partially reverses some immune abnormalities, but most patients, even with full viral suppression, lack effective HIV-specific responses. Especially common if treatment commenced in advanced disease. Immune therapy may improve these responses.

HIV infection results in gradual ↓ production and response to endo-genous IL-2. Synthesized by CD4 cells, it induces proliferation and differentiation of CD4 and CD8 cells. IL-2 given subcutaneously as an intermittent course produces ↑ CD4 count if given alone but more enhanced when combined with HAART. Viral load does not ↑, but long-term effects are unknown. Side-effects such as fever, tachycardia, hypotension, and respiratory failure, are typically dose dependent and can limit its use.

Endogenous antigens (structured treatment interruptions) or exogenous antigens (therapeutic vaccination) are other ways of stimulating immune responses. Remune (Th1 stimulant) and ALVAC vcp 1452 are examples of therapeutic vaccines which have shown immunological benefit.

Failure of HAART to restore HIV-specific immune responses may be related to loss of antigen presentation. Viral rebound following treatment interruption presents fresh HIV antigens to the immune system, facilitating rapid response by resting memory cells. Structured treatment interruption allows for emergence of wild-type virus, which is more responsive to anti-retrovirals. It may be considered as a salvage therapeutic intervention for multidrug resistance. Main drawbacks are a rapid rebound of virus, re-emergence of archived drug resistant virus, ↓ CD4 count, and development of an acute retroviral syndrome.

Acts by reducing cellular adenine (a nucleotide necessary for DNA synthesis) by inhibiting the enzymes needed for its production. It enhances antiretroviral activity (and toxicity) of adenosine analogues, such as DDI, and induces cellular kinases that phosphorylate NRTIs, ↑ their antiretroviral activity (and toxicity). Main side-effects are bone marrow suppression (dose dependent), pancreatitis, and liver toxicity. Optimum dose unknown (usually given as 500mg twice daily) and so far no major trial evidence of benefit.

A case–control study conducted by the US Centers for Disease Control (CDC) showed that zidovudine PEP given to those occupationally exposed to HIV was associated with an 80% ↓ in infection. Combination treatment, demonstrably more potent and less likely to be affected by viral resistance, is now recommended.

Therefore consider if contact with HIV likely through:

Average risk for HIV transmission after percutaneous exposure to HIV-infected blood in healthcare settings is ~3/1000 injuries. ↑ with large volumes of blood, deep injury, and high VL. After mucocutaneous exposure average risk is ~1/1000. No risk of HIV transmission if intact skin exposed to HIV-infected body fluids.

If HIV status of source is unknown, a designated doctor (not exposed worker) should obtain consent for HIV and other blood-borne virus testing.

Wash skin or exposed wound with soap and water, without scrubbing and antiseptics. Bleeding of puncture wounds should be encouraged. Exposed mucous membranes, including conjunctivae, should be liberally irrigated with water before and after removing any contact lenses.

Following a discussion of risks and benefits, PEP should be recommended to HCPs if they have had a significant occupational exposure to blood or other high-risk body fluid from someone either known to be HIV infected or considered to be at high risk of HIV infection. PEP should be commenced as soon as possible after the event and should be continued for 4 weeks. UK Department of Health guidance states that PEP may still be worth considering even if 2 weeks have elapsed following exposure. However, studies suggest that delays >72 hours may render PEP ineffective.

Limited data, but reports from Brazil and South Africa suggest that PEP may provide some protection. Use of PEP following sexual exposure to HIV is only recommended within 72 hours of exposure (as early as possible).

Risk–benefit assessment should be made considering risk of transmission according to the coital act (Box 53.1) and the likelihood of the source being HIV +ve ( Chapter 35, Prevalence p. 414). Other factors to consider include the possibility of pre-existing HIV infection, and the ability to adhere to/tolerate the proposed antiretroviral regimen.

PEP continued for 1 month. A negative antibody test 3 months after completing PEP confirms that infection has been avoided.