University of California, San Francisco Logo

University of California, San Francisco | About UCSF | Search UCSF | UCSF Medical Center

Prognosis of patients treated with cART from 36 months after initiation, according to current and previous CD4 cell count and plasma HIV-1 RNA measurements
Global Health Sciences Literature Digest
Published June 03, 2010
Journal Article

The Antiretroviral Therapy Cohort Collaboration (ART-CC). Prognosis of patients treated with cART from 36 months after initiation, according to current and previous CD4 cell count and plasma HIV-1 RNA measurements. AIDS. 2009 Oct 23;23(16):2199-208.

In Context

The CD4 cell count at baseline, and the HIV viral load (VL) after start of antiretroviral treatment (ART) have been shown to be useful predictors of disease and death, but are usually evaluated over short periods of follow-up.(1) The usefulness of the measurements in patients who are in treatment for many years has not been well studied.


To determine whether CD4 cell and VL measurements at baseline, six months and 36 months after starting treatment are useful in determining prognosis beyond three years of ART.


The Antiretroviral Therapy Cohort Collaboration (ART-CC), an international collaboration of cohort studies of HIV-infected patients from Europe and North America(2)


Combined analyses of data from multiple longitudinal treatment cohorts


Data were included of patients from 15 cohorts who were aged ≥16 years, were treatment naïve, had started ART after 1998, and had survived and were followed beyond three years on treatment. Availability of CD4 and VL measurements was mandatory before starting ART and at six months and 36 months following treatment initiation.


Data from 15 cohorts were used for analysis, which included Kaplan-Meier estimates and Cox proportional hazard modeling. All analyses were adjusted for gender, age, transmission group, and clinical AIDS at time of treatment initiation.


Of the 49,040 patients in the ART-CC data set, 14,208 met criteria for this study; 74% were male, median age was 38 years, and nearly half were men who had sex with men (MSM) or injection drug users (IDUs). At baseline, the median CD4 count was 210 cells/µL, the median HIV-1 RNA was 90,242 copies/mL. During the 32,689 person-years of follow-up after the first 36 months of treatment, there were 323 deaths (0.96 deaths per 100 persons years [p100py], 95% confidence interval [CI]: 0.86-1.08) and 345 patients with a new AIDS event (1.06 events p100py, 95% CI: 0.95-1.17). CD4 cell counts at each time point (0, 6, and 36 months) and VL at six months and 36 months were predictive of AIDS and of death during long-term follow-up, although the CD4 count was more strongly predictive of outcome. When adjustments were made in the analyses for measurements over time, 36-month measurements were most strongly associated with outcome during the follow-up period. However, the six-month VL and baseline CD4 counts were highly predictive of the 36-month values.


This study shows that HIV VL at treatment initiation is not prognostic of new AIDS events or death after 36 months, and that 36-month values for both VL and CD4 counts are the strongest predictors of longer-term outcomes. The strong association of baseline CD4 counts with six- and 36-month values confirms the importance of initiating ART before CD4 cells counts decline too far. Only 25% of patients who initiated ART at CD4 counts <200 cell/µL, had CD4 counts >500 cells/µL at 36 months, even while maintaining virologic suppression. Thus, those who start treatment at <200 cells/µL are unlikely to normalize their CD4 cell counts. The association of VL at six months with VL at 36 months and subsequent development of AIDS confirms the importance of achieving an early response to ART.

Quality rating

This was a high quality study; because, however, analyses excluded patients who did not follow-up or survive to at least 36 months, results are applicable only to those who exhibit some initial response to treatment. Some variables that might have been important to outcome were not collected in ART-CC participants, such as rates of medication adherence. Because the analyses were complex, not all data could be shown, such as changes in CD4 count and its relation to outcome.

Programmatic implications

The first six months on ART is a critical time. This study shows that patients who start therapy with CD4 counts <200 cell/µL are much less likely to do well over the long term and that achieving viral suppression within the first six months is important. Programs should try to identify HIV-infected patients early and start them on treatment before advanced immunological suppression,(3, 4) particularly because many patients in resource-limited countries do not seek treatment until their CD4 counts are well below 200 cells/µL. Monitoring should include regular testing of CD4 counts and VL at six months to make sure that viral suppression has been achieved. As data from this study are from North American and European cohorts, findings may or may not be applicable to resource-limited countries.


  1. Egger M, May M, Chene G, et al. Prognosis of HIV-1-infected patients starting highly active antiretroviral therapy: a collaborative analysis of prospective studies. Lancet 2002; 360:119-29.
  2. ART Cohort Collaboration. Defining prognosis in the era of potent antiretroviral therapy. Accessed May 13, 2010.
  3. When To Start Consortium. Timing of initiation of antiretroviral therapy in AIDS-free HIV-1-infected people: a collaborative analysis of 15 HIV cohort studies. Lancet 2009; 373:1352-63.
  4. Kitahata MM, Gange SJ, Abraham AG, et al. Effect of early versus deferred antiretroviral therapy for HIV on survival. N Engl J Med 2009; 360:1815-26.