The most immediate measure of the effectiveness of antiretroviral therapy is the viral load test. The lower a person's viral load, the less likely they are to progress to AIDS.

The effectiveness of anti-HIV treatments is usually measured in terms of the percentage of people who achieve viral loads below 50 and 400 copies/ml after a set period of time. For example, a trial will find out how many people on a particular combination are below 50 copies/ml after six months of treatment.

A number of factors impact on the effectiveness of a particular combination in any one trial including: viral load and CD4 count of participants; adherence to the combination by participants; statistical analysis applied to the results; side-effects; previous use of anti-HIV drugs, and resistance. Thus, it is not considered scientific to compare results between trials.

Results to date

An emerging trend in the data suggests HIV is not suppressed to undetectable levels in a significant minority of people, six to 18 months after staring HAART.

A review study, which deemed missing data as treatment failure, found that 57-95% of people on trials had viral loads below 400 copies/ml after six months of anti-HIV treatment. On average, four out of five people were below 400 copies/ml after six months. Combinations reviewed in this study included one PI and two nucleosides, two PIs and two nucleosides, a non-nucleoside and two nucleosides, and three nucleosides including abacavir.

When the ultra-sensitive viral load measure was used, between 30-78% were below 50 copies/ml after six months. On average, 52% had sustained viral load suppression.

Figures from the Royal Free Hospital in London have cast further doubt over the ability of treatment to suppress HIV in a real world setting, despite clear clinical benefits. At commencement of therapy, median viral load was 5.13 log and median CD4 count was 186. Sixty-one percent had never taken anti-HIV drugs and those who had taken antiretroviral therapy had not generally been heavily treated. In accordance with the review study findings discussed above, 65% of individuals had suppressed viral load to below 400 copies/ml after six months of HAART. However, 12.4% of those with viral suppression had two viral load results above 400 in the following six months and the proportion of patients experiencing treatment failure increased to 27.4% at 12 months, with 42.2% having at least one viral load result above 400 during this time (Mocroft 2000).

A review of 336 patients who started therapy at the Goethe University Clinic in Frankfurt found that treatment failure was largely associated with interruption of treatment due to toxicity or co-morbidity rather than genuine virological failure of the regimen. The risk of viral breakthrough above 500 copies, or switch to new drugs after a viral load above 50 copies, after 3.3 years was 5.2% (Phillips)

A study of 556 patients at the Royal Free Hospital found that 44% of patients had modified their first HAART regimen after a median of 14.4 months on treatment. Less than one third of those had modified their treatment due to virological failure; the overwhelming reason for changing treatment was due to poor adherence or toxicity (Mocroft 2001).

In those who can tolerate HAART however, the effects appear to be durable, with no increase in the risk of treatment failure as time goes on. The Royal Free Hospital in London reported that of 237 treatment-naﶥ patients who started triple therapy and subsequently achieved a viral load below 50 copies/ml, only 5.4% experienced a viral load rebound above 50 copies/ml during the follow-up period, an average rebound rate of 1 person per 26 person years of follow-up. The majority of patients had been followed for at least two and a half years, and the maximum follow up period was 4.4 years. No cases of viral rebound were observed in patients with three or more years' experience of HAART. The authors of the study noted that if current rates of virologic failure persisted, the patients' current combinations could be expected to last for 20 years - if they could be tolerated without side effects (Lampe 2002).

A number of groups have reported long-term follow-up from clinical trials which began up to six years ago, showing that the majority of patients who initiated treatment continue to have undetectable viral load on the same regimen.

US investigators have reported on 33 patients recruited to a study of AZT/3TC/indinavir at two centres. Of the 33 study members, 16 (48%) discontinued treatment with the trial combination before six years of follow-up. Data were obtained for twelve patients who stopped taking the study combination before six years. Of these, seven changed therapy because of an increase in their viral load, two because of the use of contraindicated medications to treat AIDS-defining illness, one because of side-effects, and two individuals switched medication after requesting it. At year six, 53% of study members for whom data were available had a viral load below 500 copies/mL, and 47% had a viral load below 50 copies/mL. A total of ten patients experienced virological failure, six in year one, three in year three and one in year six. By year six the median CD4 increase from baseline was 268 cells/mm3. Body fat changes of abnormal blood lipids were present in the majority of patients who remained on study medication to the end of six-years of follow-up. Of these 17 patients, ten (59%) were classified as having lipodystrophy.

Five year follow-up of 100 patients who began treatment with lopinavir/ritonavir in an initial dose ranging study shows that 67% had viral load below 400 copies/ml by intent to treat analysis. Nine of 32 discontinuations were due to adverse events, with the remainder due loss to follow up or non-compliance, and the only patient with two consecutive viral load measurements above 400 copies had interrupted treatment (Hicks 2003).

A team from San Francisco has reported that after more than four years on treatment, patients with viral load below 1,000 copies were continuing to experience CD4 cell increases in the fourth year of a HAART regimen, with a median increase between year 3 and 4 of +64 cells/mm³. Interestingly, this is the average annual CD4 cell decline observed in untreated people with HIV, suggesting that after the initial period of strong immune reconstitution in the first six months of therapy, the rate of CD4 reconstitution will slow and that individuals with lower baseline CD4 cell counts will take longer to return to normal, age-dependent adult CD4 cell levels (Hunt 2002).

For further details of the link between viral load and disease progression, see Predicting prognosis with viral load in Viral load, CD4 cell counts and other tests: Viral load. For further discussion of immune recovery due to treatment, see Immune recovery in Anti-HIV therapy: Choosing your treatment strategy.

References

AVANTI Steering Committee. Analysis of HIV-1 clinical trials: statistical magic? Lancet 353(9169): 2061-2064, 1999.

Hicks C et al. 5 year results of lopinavir/ritonavir-based therapy in antiretroviral-naﶥ HIV-infected adults. Ninth European AIDS Conference, Warsaw, abstract 7.3/16, 2003.

Hunt P et al. Continued CD4+ T-cell count gains in patients with 4 years of suppressed viral load on HAART. Ninth Conference on Retroviruses and Opportunistic Infections, Seattle, abstract 480, 2002.

Lampe F et al. Viral rebound after suppression with HAART: experience from 237 people with viral load <50 copies/ml followed for up to 4.4 years. Ninth Conference on Retroviruses and Opportunistic Infections, Seattle, abstract 536, 2002.

Mocroft A et al. Immunological, virological and clinical response to highly active antiretroviral therapy treatment regimens in a complete clinic population. AIDS 14(11): 1545-1552, 2000.

Mocroft A et al. Reasons for modification and discontinuation of antiretrovirals: results from a single treatment centre. AIDS 15: 185-194, 2001.

Phillips AN et al. Durability of HIV-1 viral suppression over 3.3 years with multi-drug antiretroviral therapy in previously drug-naﶥ individuals. AIDS 15: 2379-2384, 2001.