Tool to calculate prognosis for patients starting HIV therapy updated

The Antiretroviral Therapy (ART) Cohort Collaboration, an international team of investigators, have updated their prognostic model predicting the risk of progression to AIDS or death in individuals starting antiretroviral therapy. Younger patients starting HIV therapy with a CD4 cell count above 350 cells/mm³ had the lowest probability of progression to AIDS or death in the updated model, underlining the importance of timely diagnosis of HIV infection. The model, which now makes predictions for five years after HIV therapy is started, takes into account factors such as age, CD4 cell count, viral load and mode of HIV transmission. The investigators findings are published in the May 31^st edition of AIDS and the prognostic model will soon be available online at www.art-cohort-collaboration.org.

The finding that patients with CD4 cell counts above 350 cells/mm³ when they started HIV treatment had the best prognosis will add weight to the arguments of some clinicians that current HIV treatment guidelines (that recommend the initiation of HIV treatment at a CD4 cell count of 200 cells/mm³) are too cautions. The benefits of HIV therapy, would they suggest, be greater if treatment was initiated earlier, when a patient had a CD4 cell count of 350 cells/mm³. HIV drugs are now sufficiently safe, potent, and convenient to make such a treatment strategy viable, they argue.

The Antiretroviral Therapy (ART) Cohort Collaboration is an international project involving investigators from HIV cohort studies across North America and Europe. It was established in 2000 to monitor the progression of HIV disease in treatment-naïve individuals commencing potent antiretroviral therapy. The investigators have previously published three-year prognostic models.

Data from over 20,000 patients commencing HIV therapy was studied by the investigators to update their models to provide five year estimates of the risk of progression to AIDS or death. They also studied data from 16,000 patients who had completed six months of HIV therapy, enabling them to predict five-year prognosis on the basis of treatment response after six months.

Patients had a median age of 36 years. Median CD4 cell count was 224 cells/mm³, and median viral load was approximately 80,000 copies/ml.

After six months of treatment, median CD4 cell count had increased to 345 cells/mm³, and median viral load had fallen to 200 copies/ml.

During almost 62,000 person-years of follow-up, 1,005 individuals died and 1,303 developed AIDS.

Risks of progression to AIDS or death from the start of HIV treatment and after six months of HIV therapy were calculated. Five variables were used to estimate prognosis from the start of HIV therapy: age; CD4 cell count; viral load; mode of HIV transmission (injecting drug use versus non-injecting drug use); and an AIDS diagnosis before the commencement of HIV therapy.

Variables used to calculate prognosis after six months of HIV treatment were age; six month CD4 cell count; six month viral load; AIDS at baseline; and injecting drug use versus non-injecting drug use.

The first model found that patients aged under 30 who started HIV therapy with a CD4 cell count above 350 cells/mm³, a viral load below 100,000 copies/ml, had not progressed to AIDS before starting HIV treatment and were not injecting drug users, had the lowest risk of five-year progression to AIDS or death with a probability of 5.6%. The probability of death alone was 1.8%.

However, there was a 77% risk of progression of AIDS or death within five years for patients aged over 50 when they started HIV treatment, who were injecting drug users, and who had a baseline CD4 cell count of below 25 cells/mm³ and a viral load above 100,000 copies/ml and who already had an AIDS diagnosis. When the risk of death was analysed separately for these patients, it was 65%.

Overall, patients in the entire study population had a 12% probability of progression to AIDS or death within five years of starting HIV treatment, and a 5% of probability of progression to death.

Attention was then turned by the investigators to six-month response to HIV treatment and prognosis.

They found that the lowest risk of progression to AIDS or death was 4% at five years and was in patients aged under 50 when they started HIV treatment, not infected with HIV due to injecting drug use, and who had a good response to HIV therapy, their six month CD4 cell count increasing to over 350 cells/mm³ and their viral load falling to below 500 copies/ml. The risk of death alone was 1%.

However, there was an almost certain (99%) probability of AIDS or death within five years for those aged over 50, who were injecting drug users, with a very poor response after six months of HIV treatment, their CD4 cell count being below 25 cells/mm³ and viral load above 100,000 copies/ml. The risk of death within five years was 96%.

“The most strongly prognostic factor from the start of HAART was CD4 cell count at baseline”, write the investigators. They add, “in analyses from 6 months, baseline CD4 cell count was no longer prognostic, but both 6-month CD4 cell count and 6-month viral load were.”

Late diagnosis of HIV was identified by the investigators as a probable factor in poor prognosis. They observe, “approximately half of patients started HAART either with a CD4 cell count of less than 200 cells/mm³ or with an AIDS diagnosis…the start of HAART below this level is probably caused by late presentation and diagnosis of HIV.” The investigators express particular concern as “early diagnosis and timely treatment are of great importance to prevent clinical progression and to prevent spread of the infection.”

The investigators did not have information on the specific causes of death. There is evidence from other studies that non-AIDS defining illnesses are now accounting for an increasing proportion of deaths in HIV-infected individuals.