Infectious Disease Postdoc/Clinician
Department of Pediatrics, University Hospital, Syracuse NY
Send response to journal:
Now that the Montagnier graph has been fixed, it appears that the interpretation has also been "fixed".
This graph looks at a single data point. You cannot conclude anything about the predictive value of viral load.
It is well known that the viral load remains relatively constant throughout asymptomatic infection. However the level of this load is predictive of the rate of decline (NOT the decline itself, the RATE of decline). I.e. in the graph shown, if one were to draw a second viral load line in red at a higher level, the corresponding CD4 count would drop faster (say down to 200 by year 6 instead of year 9).
In case anyone says this is mere hand-waving, I refer them to the work by Mellors et al , where this is made clear. I did refer the Perth Group to this paper but they have either chosen to ignore it, or are simply incapable of understanding it.
Figure 1 of Ref  very clearly states that the SLOPE of CD4 T cell loss (i.e. the rate of decline) corresponds to the level of steady-state viral load. As Mellors et al say:
" Figure 1 shows a monotonic relation between HIV-1 RNA concentrations and decline in CD4 (+) lymphocyte counts-the higher the HIV -1 RNA concentration, the greater the rate of decline in CD4+ lymphocyte count."
Unlike the Montagnier conference presentation, this work looks at 1,604 patients with HIV infection. Montagnier was simply presenting the untreated disease process, apples and oranges. And besides, you can't simply disagree with someone and then present evidence to fit that (unless of course you're the Perth Group). You must look at the evidence and THEN come to a conclusion.
Fundamentally, the CD4 count is a measure of how far the train is from the end of the tracks (AIDS) and the viral load is a measure of how fast the train is travelling.
Please note that the viral load measurements used in this paper were NOT derived from PCR. However an FDA-approved RT-PCR kit was used as well and correlated with a coefficient of 0.93 with their own methods.
Nick Bennett firstname.lastname@example.org
1. One free online paper, of several reporting the same results is at http://www.annals.org/cgi/content/full/126/12/946 Mellors et al Ann Intern Med. 1997 Jun 15;126(12):946-54. "Plasma viral load and CD4+ lymphocytes as prognostic markers of HIV-1 infection."
Competing interests: None declared