CMV viral load remains useful predictor of HIV disease risk

Changes in cytomegalovirus (CMV) viral load are an important prognostic marker of HIV disease progression in the HAART era according to new data from the Royal Free and University College School of Medicine in London, which were presented at the Ninth Retroviruses Conference in Seattle today.

Prior to the introduction of HAART, study of CMV viral load had established CMV viraemia as a predictor of both future risk of CMV disease, and of HIV disease progression. The current analysis was designed to investigate whether this relationship remained despite the widespread use of effective antiretroviral therapy. (CMV viraemia is a term used to describe a detectable CMV viral load, which is itself a sign that CMV is active, and replicating within the body).

Jane Deayton and colleagues recruited patients who had a CD4 count below 100 during the period between January 1997 and December 2000. All received CMV viral load testing, using a PCR-based method developed in-house, at every subsequent clinic visit regardless of whether their CD4 rose above the 100 mark. This level is a marker for increased risk of clinical CMV disease in immunocompromised individuals. This commonly occurs in the form of retinitis, a progressive condition of the eye which can result in loss of vision without effective treatment.

Three hundred and ninety people were followed for a median of 36 months. Seventy-three percent had been exposed to antiretrovirals at baseline, and 92% received HAART. Eighty per cent of the cohort were male, and a quarter were Black Africans. At baseline, median CD4 was 77, and median HIV viral load was close to five logs, or 100,000 copies.

Progression to a new AIDS-defining event occurred in 37% of those with CMV viraemia and in 19% of those whose CMV status was negative. Whilst CD4 count was an independent predictor of HIV disease progression, CMV status predicted this risk more effectively. Similarly, in relation to the risk of death, CMV status was a better prognostic marker than either CD4 or HIV viral load, by multivariate analysis.

CMV has long been proposed as a possible co-factor in HIV disease, and these new data appear to support this hypothesis. An alternative explanation is that CMV viraemia may be a surrogate for incomplete immune function in people receiving HAART. How to respond to CMV viraemia in people who have low CD4 counts despite taking HAART remains an open question. In closing in Seattle, Dr Deayton proposed the need to evaluate the role of CMV treatment in this context.