HAART's Improving Control of HIV Has Bottom-Line Limit

LONDON, Aug. 4 -- Improvements in the number, type, and quality of medications used in highly active antiretroviral therapy (HAART) have resulted in lower viral loads and higher immune cell counts.

But the improved virological response has not translated into a decline in mortality or progression to AIDS among those on HAART, according to a study in the Aug. 5 issue of The Lancet.

When HAART was introduced a decade ago, there was an immediate and sharp decline in the death rate from AIDS, as those infected with HIV were introduced to the new treatment approach.

Since then, in a search for even better therapies, researchers have modified HAART dramatically, culminating in the approval last month of Atripla, a single-pill combination of the three top antiretroviral drugs, that is taken once daily.

But all of that appears to have had little impact on progression to AIDS or death among people on HAART, found an analysis of 22,217 treatment-naive patients from North America and Europe who started on therapy between 1995 and 2003.

The Antiretroviral Therapy Cohort Collaboration found:

• In 1995-96, 58% of patients starting HAART achieved a viral load of less than 500 copies of viral RNA per milliliter of blood within six months of starting treatment.
• In 2002-03, that proportion had risen to 83%.
• The median CD4 cell count when starting HAART in 1995-96 was 170 cells per microliter of blood. That increased to a peak of 269 cells per microliter in 1998, but fell again to 200 cells in 2002-03.

At the same time, the risk of progression to AIDS or death did not change greatly.

Taking 1998 as a reference year, the researchers found:

• The hazard ratio for AIDS was 1.07 in 1995-96, rose to 1.30 in 1997, and then dropped again 1.07 in 1999. Since then, it has risen steadily, peaking at 1.35 in 2002-03.
• The hazard ratio for death was 0.87 in 1995-96 and 0.96 in 2002-03, but in no year was it significantly different from 1998.

The researchers analyzed the data on a year-by-year basis, but 1995-96 and 2002-03 were grouped together because fewer people started HAART in those years.

One possible reason for the discrepancy, the researchers suggested, is that the make-up of the patient community is changing, as is the nature of the first AIDS-defining disease.

In 1995-96, 16% of patients starting HAART were women, a proportion that has climbed steadily, reaching 32% in 2002-03. Also, in 1995-96, 56% of patients were men who had sex with men, while in 2002-03, the largest transmission risk group was heterosexual, at 47%.

At the same time, tuberculosis has become much more common as an AIDS-defining illness: in 1995-96 the hazard ratio for tuberculosis was 0.73, while in 2002-03 it was 2.94, significantly higher than in the reference year of 1998.

By contrast, other AIDS-defining conditions have not changed significantly from 1998.

In one example, the researchers noted that the Swiss HIV Cohort Study has seen a steady increase in patients from sub-Saharan Africa -- patients who were younger, more likely to be women and to have been infected heterosexually, and to have TB as a first AIDS -- defining illness.

In the first decade of HAART, physicians and patients faced several challenges, including large pill burdens, worrisome side effects, and drug resistance, said Gregory Dore, Ph.D., and David Cooper, M.D., of the University of New South Wales in Sydney, Australia, in an accompanying comment article.

While most of those were overcome, they added, the shifting demographics -- including immigration from areas where both HIV and TB are raging -- are likely responsible for the discrepancy between better response to therapy and no apparent improvement in AIDS or death.