There may be no advantage to starting therapy at a CD4 count above 200, according to a review of several large cohort studies presented at the British HIV Association Autumn meeting in London last week
Professor Andrew Phillips of the Royal Free Hospital Medical School analysed four cohorts, including the 7,331 person Eurosida cohort, in order to assist the development of UK treatment guidelines. The data was gathered from clinic observations rather than clinical trials.
"Given the current evidence, there would appear to be no compelling case for starting therapy with a CD4 count above 200 unless the patient is particularly committed to starting therapy or is experiencing symptomatic illness", he concluded.
However, UK guidelines are likely to continue to recommend that treatment begin when the CD4 count falls below 350. Professor Brian Gazzard, Chairman of the British HIV Association, said "I wouldn't change the guidelines on the basis of this data alone. However, it is reassuring that you don't get progressively greater responses if you start therapy at 500 or above, indicating that 350 is a reasonable starting point".
Prof. Phillips reviewed four questions:
How will HIV disease progress if therapy is delayed?
"If we look at the delay between a CD4 count of 500 and a decline to 200 cells, the median delay is 4-5 years . In other words, half of individuals with CD4 counts which first fell below 500 in 1994 would only have reached 200 cells in 1999. Viral load had a slight independent effect on the rate of CD4 cell count decline".
Looking at the risk of developing lymphoma, which is not arrested or cured by HAART-related immune reconstitution, there was no evidence that HAART reduces the risk of developing lymphoma at CD4 counts greater than 200. The incidence in untreated individuals with CD4 counts greater than 200 was 1 per 300 person years in the Eurosida cohort; there was no significant difference in incidence compared with those on HAART.
Is the virological response better if HAART is started at a higher CD4 count?
To answer this question, Prof. Phillips looked at 560 treatment-naïve patients from the Frankfurt Clinic Cohort, and looked at rates of virological failure according to baseline CD4 count. The definition of virological failure used was failure to achieve viral load below 500 copies at week 24, or the time to virological rebound after achieving viral load below 500 copies at week 24.
Individuals who started treatment when their CD4 count was below 200 failed treatment more quickly than those with CD4 counts above 200. There was no statistically significant difference between time to treatment failure in those who started treatment with CD4 counts above 350 and those who started treatment with CD4 counts between 200 and 350.
This difference was also apparent in over 400 treatment-naïve patients from the Eurosida Cohort, and in 536 treatment-naïve patients from the Italian ICONA cohort. An individual's viral load at baseline had no independent effect on the durability of treatment in the Frankfurt and Eurosida cohorts, but it exerted a slight but statistically significant effect in the ICONA cohort.
Does starting therapy at a higher CD4 count reduce the risk of toxicity?
In the ICONA cohort, in which 805 individuals had a median treatment experience of 41 weeks, 34.9% (281) had stopped at least one drug. The most common reason was drug-related toxicity (20.9%), followed by non-adherence (7%). Only 4.6% stopped a drug due to virological failure. There was no significant relationship between baseline CD4 count and the subsequent risk of stopping a drug due to toxicity.
Can HAART adequately reconstitute immunity from a low baseline CD4 count?
In the Frankfurt Clinic Cohort, in which treatment-naïve individuals have been followed for two years after commencing treatment, the increase in absolute CD4 count has been approximately 200 cells, regardless of baseline CD4 count. Amongst those with CD4 counts below 20 at baseline, there has been a higher rate of subsequent opportunistic infections while the CD4 count remained below 20 cells (32 out of 123 person years) than while the CD4 count had risen to between 20 and 49 (4/44 person years) or 50-199 (6/178),
This indicates that the rises in CD4 count experienced were associated with reduced risk of AIDS-related diseases. So there does appear to be meaningful immune reconstitution in those starting therapy even at very low CD4 counts, as long as this count rises.