Current issues of ACP Journal Club are published in Annals of Internal Medicine


HIV-1 RNA levels and CD4+ cell counts predicted disease progression in HIV infection

ACP J Club. 1998 Jan-Feb;128:13. doi:10.7326/ACPJC-1998-128-1-013

Source Citation

Mellors JW, Muñoz A, Giorgi JV, et al. Plasma viral load and CD4+ lymphocytes as prognostic markers of HIV-1 infection. Ann Intern Med. 1997 Jun 15;126:946-54.



To determine the predictive value of clinical and laboratory markers for disease progression and AIDS-related death in HIV infection.


Cohort study (Multicenter AIDS Cohort Study [MACS]).


4 university-based clinical centers in the United States.


From 1984 to 1985, the original MACS enrolled 4954 homosexual men who were free of clinical AIDS. Eligible patients in this study were 1813 men who were HIV-seropositive at baseline and 169 men who seroconverted to HIV within 18 months of enrollment. 1604 men (81%) for whom baseline CD4+ cell counts and HIV-1 RNA levels and follow-up were available. Median follow-up was 9.6 years for men who did not develop AIDS.

Assessment of prognostic factors

CD4+, CD3+, and CD8+ cell counts; HIV-1 RNA level; serum levels of β2-microglobulin and neopterin; and oral thrush or fever for ≥ 2 weeks.

Main outcome measures

Time to development of AIDS and AIDS-related death.

Main results

998 men developed AIDS, and 855 died of AIDS during follow-up. Disease progression was predicted by HIV-1 RNA levels, CD4+ cell counts, neopterin levels, and β2-microglobulin levels. The relative risks for AIDS and AIDS-related death correlated directly with HIV-1 RNA levels (Table). The addition of baseline CD4+ cell count to HIV-1 RNA level provided a better estimate of outcome (see Commentary). The predictors did not change when data were analyzed by center or use of antiretroviral therapy.


HIV-1 RNA level was the best single predictor of clinical outcome in HIV-1 infection, followed by CD4+ cell counts. Prognosis was most accurately defined by use of both viral load and CD4+ cell counts together.

Source of funding: National Institutes of Health.

For article reprint: Dr. J.W. Mellors, University of Pittsburgh, Graduate School of Public Health, 603 Parran Hall, 130 DeSoto Street, Pittsburgh, PA 15261, USA. FAX 412-383-7982.

Table. Progression to AIDS or AIDS-related death at 6 years*

Baseline HIV-1 RNA levels (copies/mL) AIDS AIDS-related death
% Relative risk (95% CI) % Relative risk (95% CI
≤ 500 5.4 (Reference standard) 0.9 (Reference standard)
501 to 3000 16.6 2.4 (1.4 to 4.1) 6.3 2.8 (1.4 to 5.6)
3001 to 10 000 31.7 4.3 (2.5 to 7.3) 18.1 5.0 (2.5 to 9.8)
10 001 to 30 000 55.2 7.5 (4.4 to 12.7) 34.9 9.8 (4.9 to 19.1)
> 30 000 80.0 12.8 (7.5 to 21.8) 69.5 18.1 (9.2 to 35.7)

* Adjusted for CD4+ cell count


In the past 18 months, the management of HIV infection has been transformed by the availability of HIV viral load assays and more potent antiretroviral therapies. An earlier report from Mellors and colleagues (1) on a subset of men showed that HIV-1 RNA load was an important prognostic marker. Uncertainty remained, however, about the best use of data on viral load in conjunction with CD4+ cell count.

This larger study has now shown that although plasma viral load is the strongest predictor of disease progression, the prognosis of HIV infection is more accurately described by the combined use of HIV viral load and CD4+ cell count. For example, among patients with a viral load ≤ 500 copies/mL, 3.6% had progressed to AIDS at 9 years if their CD4+ cell count was > 750 cells/mm3, compared with 22.1% if their CD4+ cell count was ≤ 750 cells/mm3. Similarly, if the viral load was > 30 000 copies/mL, the 9-year progression rate for a baseline CD4+ cell count of > 500, 351 to 500, 201 to 350, and ≤ 200 cells/mm3 was 76.3%, 94.4%, 92.9%, and 100%, respectively. It is important to note that although 60% of the cohort subsequently used antiretroviral therapy, pretreatment baseline viral loads remained predictive of disease progression independent of subsequent therapy.

A limitation of this study is that it does not describe a true inception cohort, given that most participants were already infected with HIV at enrollment. The duration of HIV infection at the time of the initial measurement of viral load was unknown. In addition, the lack of women and children and underrepresentation of ethnic groups limit the generalizability of the findings.

The findings do, however, underscore the importance of interpreting viral load results in the context of both clinical status and CD4+ cell counts. In addition to helping decide when to begin antiretroviral therapy, other studies have shown the value of viral load assays in assessing the efficacy of antiretroviral regimens (2) and in determining when to initiate changes in treatment before the onset of clinical failure (3). However, caution still must be exercised in the use of viral load: Different assay kits may yield different measurements of viral load, and viral load should not be measured during or close to the time of successful treatment of infection.

Philippa Easterbrook, MD
Imperial College School of MedicineLondon, England, UK


1. Mellors JW, Rinaldo CR Jr, Gupta P, et al. Prognosis in HIV-1 infection predicted by the quantity of virus in plasma. Science. 1996;272:1167-70. Published erratum appears in Science. 1997;275:14.

2. Hughes MD, Johnson VA, Hirsch MS, et al. Monitoring plasma HIV-1 RNA levels in addition to CD4+ lymphocyte count improves assessment of antiretroviral therapeutic response. ACTG 241 Protocol Virology SubstudyTeam. Ann Intern Med. 1997;126:929-38.

3. Saag MS, Holodniy M, Kuritzkes DR, et al. HIV viral load markers in clinical practice. Nat Med. 1996;2:625-9.