Whether you are failing your first or fourth anti-HIV regimen, there are several strategies available to you - changing therapy , continuing a failing regimen, or stopping treatment for a set period of time.

Your strategy may depend on what effective drug combinations you have available to you. If you have taken several combinations, you may be cross-resistant to all current non-nucleoside reverse transcriptase inhibitors (NNRTIs) and most of the nucleoside reverse transcriptase inhibitors (NRTIs), and have some cross-resistance to protease inhibitors (PIs). Your choice of drugs may also be limited by side-effects.

In the process of changing treatment, you need to identify any factors which have contributed to the failure of previous treatments - as well as resistance, low drug levels and adherence to your regimen may often play a significant role. Taking your treatment as prescribed is the area you control completely. If you have difficulties taking your drugs as required, or you miss dosages, discuss with your doctor when and why non-adherence to therapy occurs. This will help you maximise adherence with your next regimen.

See What level of adherence is necessary? in Anti-HIV therapy: Adherence for further discussion of this issue.

Below are several strategies to help you determine which drugs are likely to be most effective for you.

Treatment history to select treatment

You and your doctor may be able to reach an informed conclusion about which of your drugs is failing by taking account of your previous treatment history and the relative risk of developing resistance to different drugs.

If you have taken any drugs as monotherapy in the past, these drugs are obvious candidates for resistance, especially AZT (zidovudine, Retrovir). If you have taken dual combinations, or added another drug to AZT, these drugs might also be candidates for resistance. The amount of time you have spent on particular drugs also needs to be taken into account.

Resistance to some drugs (such as 3TC [lamivudine, Epivir] and the NNRTIs) emerges very quickly.

On the other hand, resistance to ddI (didanosine, Videx / VidexEC) and d4T (stavudine, Zerit) emerges very slowly, and resistance to ddC (zalcitabine, Hivid) is also fairly slow to emerge. However, in 1 to 2% of people who take these drugs, a mutation can emerge which causes cross-resistance to all the currently available NRTIs.

See Resistance testing to select treatment in Anti-HIV therapy: Resistance for further details.

Resistance testing to select treatment

The first possibility is using resistance testing to define which drugs might still work.

Two types of resistance test are available. Genotypic tests look for specific changes, or mutations, in the virus which may predict resistance to a drug. Phenotypic tests measure the amount of drug which is needed to control viral replication. As the virus grows more resistant, the amount of drug needed to control replication will increase. This is called loss of sensitivity or susceptibility.

In people with lots of drug experience, resistance tests have shown rather limited benefits. Though this may be due to the relative lack of effective treatment options, the chief advantage of resistance testing in these circumstances probably lies in ruling out drugs which will definitely not work (especially in the absence of a detailed medical history).

Nevertheless, several studies have shown that the presence of specific genotypic mutations can predict the response to future treatment. An international panel of resistance experts has agreed that it would be prudent to use this form of resistance testing if current treatment is failing.

See Testing for resistance in Anti-HIV therapy: Resistance for further details of this test and other tests for resistance.

Therapeutic drug monitoring

There are big variations between individuals in the drug levels achieved when a group of people take the same dose of a drug, so it may not be helpful to report that an individual has developed a tenfold loss of sensitivity according to a standard scale, unless we also know whether that person was getting average drug levels to begin with. Any attempt at boosting may be unsuccessful unless we know how much drug is really needed to overcome resistance, rather than just making a guess based on average drug levels in previous clinical trials.

Companies which make resistance tests are trying to bring together the results of resistance testing and therapeutic drug monitoring to provide doctors with a single package of information that can predict the optimal therapy for an individual.

See Anti-HIV therapy: Testing drug levels for further information.

Treatment interruption

For many people who have experienced virological failure of antiretroviral therapy that is not first- or second-line, stopping therapy until enough new drugs become available to make a potent combination may appear to be an attractive option. However, people who are in this position are likely to be the group of patients at greatest risk from trying this strategy. This is because people who have exhausted most available options probably started treatment with very low CD4 cell counts.

One of the few consistent findings from studies of treatment interruptions is that if you stop treatment, your CD4 count eventually falls back to its pre-treatment level. Additionally, two major studies from 2003 looking at treatment interruption prior to salvage therapy came up with somewhat conflicting results, suggesting both that a structured treatment interruption in treatment-experienced people before changing therapy may be harmful, and that a shorter interruption (no longer than two months) may be the safer option (Katlama 2003a; Lawrence 2003).

A third study showed that in a group of 23 patients with very low baseline CD4 cell counts who stopped treatment for an average of 24 weeks, two thirds developed an AIDS-defining illness after resuming treatment, but only one achieved a viral load reduction of more than 1 log₁₀ upon resumption of treatment, despite reversion of all resistance mutations to wild type in 13 patients (Ghosn 2005).

Taken together, these findings suggest that stopping treatment is not a long-term option for people with advanced treatment failure who may previously have experienced either a low CD4 nadir (below 200 cells/mm³) and/or AIDS-defining illnesses, and provides a questionable short-term advantage prior to starting salvage therapy. In this group of people, the major consequence of a long period off treatment could be the re-emergence of old infections, and the possibility of new OIs as the immune system declines.

See Treatment interruption before salvage therapy in Anti-HIV therapy: Structured treatment interruption for more details.

Optimising background therapy

For people who are considering salvage regimens, using resistance testing to optimise your background therapy - the backbone of drugs used in addition to the new drug or drug class being used - has been shown to be extremely useful in several recent trials, including TORO-1 and 2. These were the studies of T-20 (enfuvirtide, Fuzeon), which added the fusion inhibitor to a background regimen chosen after resistance testing in highly treatment-experienced patients.

The investigators used a measure called a phenotypic susceptibility score (PSS) to figure out how potent the background therapy was.

In TORO-1, viral load response according to PSS ranged from 2.3 log₁₀ in T-20 recipients with a PSS of 3 to 4 to 0.2₁₀ log in optimised background regimen recipients with a PSS score of 1 or less.

Multivariate analysis showed that PSS was a significant factor in the success of the regimen, accounting for 0.33 log¹⁰ per active drug. (Lalezari 2002).

See the section on Phenotypic or genotypic susceptibility score (PSS/GSS) in Resistance testing to select treatment in Anti-HIV therapy: Resistance for more on this concept.

Continuing with 'failing' treatment

If you have detectable viral load and few immediate options for new treatment, continuing with existing treatment may have its advantages, not least because remaining on therapy - even failing therapy - has clinical benefits. A recent study from the United States has shown that individuals with detectable viral load below 20,000 copies/ml have a comparable risk of disease progression to individuals with viral load below 400 copies/ml over a median of 2.4 years (6 vs. 7%; Raffanti 2003)

Several years ago it was established by Veronica Miller of Goethe University in Frankfurt that people taking antiretroviral therapy have a significantly lower risk of disease progression than those not taking antiretroviral therapy. A person with a CD4 cell count of 100 cells/mm³ and a viral load of 50,000 copies/ml taking PI-containing antiretroviral therapy is almost 50% less likely to develop an AIDS-defining illness than someone with similar counts not on antiretroviral therapy (Miller 2000).

This benefits do not last forever, however. Between 2000 and 2002, Steven Deeks of San Francisco General Hospital reported several times on what had happened to 302 patients in his clinic who had detectable viral load since 1996 whilst taking PI-based HAART. After four years of detectable viraemia, patients had a 41% probability of developing an AIDS-defining illness; the risk after two years was 18%. (Deeks 2000, 2002)

If you are fortunate enough to have a detectable but low viral load (below 10,000 copies/ml), then remaining on the same regimen may well have immunological advantages comparable or even better than being undetectable. A majority of individuals in the EuroSIDA cohort experiencing low-level virologic failure of their anti-HIV treatment did not experience large increases in viral load over a 12 month period. This study looked at 488 patients who still had viral load between 1,000 copies/ml and 10,000 copies/ml more than six months after starting their first HAART regimen. The majority of patients had already received nucleoside analogues before commencing a triple drug regimen. After 12 months follow-up, 62.5% of the group remained on the failing regimen, and after 24 months 35.6% remained on the failing regimen. The median viral load increase was 0.024 log₁₀ per month, the equivalent of a viral load increase from 1000 to 2,000 copies/ml, or from 10,000 to 20,000 copies/ml, in the course of a year.

The median CD4 cell count fell by 0.53 cells/mm³ per month. In one third of patients the CD4 cell count increased by at least 1 cell/mm³ per month, and in a further third it increased by at least 2 cells/mm³ per month, suggesting that immune reconstitution continued, or was not compromised, in the majority of patients despite ongoing viral replication above the limits of detection.

Surprisingly, the researchers found that people receiving NNRTI-containing triple combinations were significantly less likely to have switched treatment within twelve months of treatment failure compared to people taking protease inhibitor-containing regimens. People taking regimens containing more than three agents were significantly more likely to switch within twelve months than people taking a protease inhibitor with two NRTIs.

The researchers suggest that the main reason for this observation is likely to be the greater tolerability of NNRTI-containing regimen rather than any effect of such regimens on the replicative capacity of the virus. (Cozzi-Lepri 2003)

A recent study from France followed 28 people on HAART or dual NRTI therapy with a stable viral load under 10,000 copies/ml for 18 months and compared them with 29 people on HAART with a stable viral load below 20 copies/ml, as well as 14 people on HAART with viral loads above 10,000 copies/ml. The median CD4 count was 341/mm³ in those with a stable detectable low viral load, compared with 53/mm³ in those with a high detectable viral load and 479/mm³ in those who were undetectable. Interestingly, they found that both CD4 and CD8 responses to HIV were significantly more frequent and intense in those with a low detectable viral load compared to those who were undetectable. Another 18 months later, just over half (54%) of those with stable detectable viral loads below 10,000 copies/ml kept their viral loads at this low level. (Katlama 2003b).

Another recent study found similar results in those with low detectable viral load. Researchers from Chicago followed 25 people for up to five years with a stable viral load under 10,000 copies/ml. Fourteen of them were on HAART and eleven were on not-currently-recommended dual NRTI therapy. They compared them to people on HAART with stable undetectable (less than 200 copies/ml) viral load. They found that the median change in CD4 cells each year was a gain of 48.7 cells in those who were undetectable; a gain of 22.2 cells/mm³ in those who were on HAART and had a viral load between 200 and 10,000 copies/ml; and a loss of 10.9 cells/mm³ in those on dual NRTIs with had a viral load between 200 and 10,000 copies/ml. Although some drug resistance was found in most of the people in the study, the majority retained susceptibility to drugs in two or more classes of antiretrovirals. Additionally, they found that both clinical and immunological benefits were maintained - even with low level viral replication (Tenorio 2003).

Another recent French study using slightly different criteria to define low-level viraemia (between 500 and 5000 copies/ml), also found benefits of remaining on PI-based antiretroviral therapy after failure. The investigators analysed the records of 3740 HIV-positive individuals who started a PI-containing HAART regimen between January 1997 and December 1998 at 68 treatment centres across France. All the patients initially achieved an undetectable viral load on antiretroviral therapy (below 500 copies/ml). In the year after inclusion in the study, 2636 patients (71%) had a persistently undetectable viral load, with 387 individuals (10%) experiencing a low rebound in their viral load (between 500 and 5000 copies/ml), and 713 patients (19%) a high rebound (above 5000 copies/ml).

Unsurprisingly, the investigators found that the patients with the most sustained viral suppression had the largest median increase in CD4 cell count (163 cells/mm³ and that patients with high viral load rebounds, the lowest median CD4 cell gains (59 cells/mm3). The median increase in CD4 cell counts for patients with a low viral load rebound was 120 cells/mm³. However, the 15-month probability of clinical failure was 2.4% amongst patients with a persistently undetectable viral load, 2.2% for individuals whose viral load rebounded to up to 5000 copies/ml, and 5.8% in patients with higher sustained rebound in viral load.

The corresponding 15-month probabilities for immunological failure were 37.2% for those with a sustained undetectable viral load, 39.7% for patients with a persistently low viral load, and 51.9% for individuals with a persistently high viral load, suggesting that a viral load below 5000 copies/ml perhaps can not be considered to be failure, at least in the short-term. The French investigators concluded among patients with low level rebound who are receiving PI-containing HAART, who have difficulties with treatment adherence, and who have previously been exposed to multiple antiretroviral regimens, delays in treatment switches may spare some therapeutic options for future use (Abgrall)

This lack of viral rebound and preservation of CD4 cell gains may be attributed to a reduced ability of HIV to replicate after the development of certain drug resistance mutations; a concept known as viral fitness.

See Viral fitness, drug resistance and the immune system in Anti-HIV therapy: Restoring the immune system for a discussion of this topic.

See also Immune recovery in Anti-HIV therapy: Choosing your treatment strategy for further discussion.

Nucleoside and nucleotide reverse transcriptase inhibitor-sparing regimens

A large proportion of treatment-experienced people have already acquired cross-resistance to current NRTIs / NtRTIs. Given also that long-term toxicities have been linked to some of the nucleoside analogues, there is emerging interest in NRTI / NtRTI-sparing regimens for salvage therapy. Currently little data is available and this strategy is not recommended by clinical guidelines.

See Options for simplifying therapy in Anti-HIV therapy: Choosing a combination for more on NRTI / NtRTI-sparing HAART.

Immune stimulation

People who have low CD4 cell counts despite control of viral load may consider augmenting their antiretroviral regimen with interleukin-2 (IL-2; Proleukin) or another immune stimulant. IL-2 is an experimental treatment for HIV disease, although it is a recommended salvage therapy option in France. It boosts CD4 cell levels in people with HIV but its clinical benefits are yet to be established. See Interleukin-2 (IL-2) - overview in Drugs used by people with HIV for details.

Choosing the combination

United States and United Kingdom guidelines recommend that people on a failing triple combination should change at least two drugs, and preferably all three. Adding only one or two drugs can have benefits, especially in the short-term, but experts believe this strategy encourages less than optimal therapy and lead to drug resistance in the medium term.

See Choosing a second-line regimen and Choosing a salvage regimen in Anti-HIV therapy: Changing treatment for details.

Reference

Abgrall S et al. Clinical and immunological outcome in patients with human immunodeficiency virus infection, according to virologic efficacy in the year after virus undetectability, during antiretroviral therapy. Clinical Infectious Diseases 37 (on-line edition), 2003.

Cozzi-Lepri A et al. Changes in viral load in people with virological failure who remain on the same HAART regimen. Antiviral Therapy 8: 127-136, 2003.

Deeks SG. Sustained CD4+ T cell response after virologic failure of protease inhibitor-based regimens in patients with human immunodeficiency virus infection. Journal of Infectious Diseases 181(3):946-953, 2000.

Deeks SG et al. Duration and predictors of CD4 T-cell gains in patients who continue combination therapy despite detectable plasma viraemia. AIDS 16(2): 201-207, 2002.

Deeks S et al. Continued reverse transcriptase inhibitor therapy is sufficient to maintain short-term partial suppression of multi-drug resistant viremia. Tenth Conference on Retroviruses and Opportunistic Infections, Boston, abstract 640, 2003.

Ghosn J et al. No benefit of a structured treatment interruption based on genotypic resistance in heavily pretreated HIV-infected patients. AIDS 19: 1643-1647, 2005.

Katlama C et al. Intensification of stable background therapy with abacavir in antiretroviral therapy experienced patients: 48-week data from a randomized, double-blind trial. HIV Med 2(1):27-34, 2001.

Katlama C et al. Long-term Benefit of Treatment Interruption in Salvage Therapy (GIGHAART ANRS 097). Tenth Conference on Retroviruses and Opportunistic Infections, Boston, abstract 68, 2003a.

Katlama C et al. Characterization of a subset of patients with persistent low viral replication under combined antiretroviral therapy. Antiviral Therapy 8 (Suppl 1) abstract 623, 2003b.

Lawrence J et al. Structured treatment interruption in patients with multidrug-resistant human immunodeficiency virus. New England Journal of Medicine 349 (9): 837-846, 2003.

Lalezari J et al. Enfuvirtide (T-20) in combination with an optimised background (OB) regimen vs OB alone: week 24 response among categories of treatment experience and baseline (BL) HIV antiretroviral (ARV) resistance. 42nd ICAAC, San Diego, abstract H-1074, 2002.

Miller V et al. The impact of protease inhibitor-containing highly active antiretroviral therapy on progression of HIV disease and its relationship to CD4 and viral load. AIDS 14(14): 2129-2136, 2000.

Molina J et al. Didanosine in treatment-experienced HIV-infected patients: results from a randomised double-blind study (A1454-176 Jaguar). 43rd ICAAC, abstract H-447, Chicago, September 14 17th, 2003.

Raffanti SR et al. Moderate viraemia and HIV disease progression. Second International AIDS Society Conference on HIV Treatment and Pathogenesis, Paris, abstract LB35, 2003.

Rozenbaum W et al. Treatment (TRT) intensification with abacavir (ABC) in HIV-infected adults with previous 3TC/ZDT antiretroviral TRT - 48 week results (CNA3009). Seventh European Conference on Clinical Aspects and Treatment of HIV-Infection, Lisbon, abstract 466, 1999.

Squires K et al. Tenofovir DF: a 48-week final analysis from a phase III randomized, double blind placebo controlled study in antiretroviral experienced patients. Ninth Conference on Retroviruses and Opportunistic Infections, Seattle, abstract P 413-W, 2002.

Tenorio AR et al. HIV-1 infected antiretroviral-treated patients with prolonged partial viral suppression: clinical, virologic, and immunologic course. JAIDS 34 (5): 491-496, 2003.