icon-    folder.gif   Conference Reports for NATAP  
  XIX International AIDS Conference
July 22-27, 2012
Washington, DC
Back grey_arrow_rt.gif
CD4 Count Under 700 Linked to All-Cause Mortality in Veterans Cohort
  XIX International AIDS Conference, July 22-27, 2012, Washington, DC

Mark Mascolini

A CD4 count under 700 independently predicted death from any cause in US veterans [1]. Other independent predictors included hepatitis virus coinfection and use of stavudine. Good antiretroviral adherence cut the death risk in this analysis.

Can antiretroviral-propelled CD4 gains predict death risk in virologically suppressed people taking antiretrovirals? VA researchers addressed that question while also weighing the relative impact of antiretroviral adherence and other factors on mortality. To do so they combed through VA Clinical Case Registry data to determine clinical follow-up time and death rates after a first prescription of antiretroviral therapy.

The investigators calculated annual area under the curve (AUC) averages for CD4 and CD8 counts and the fraction and annual AUC averages for lymphocytes negative for CD4 and CD8, which consisted of nearly equal proportions of B cells and CD3/CD19-negative T cells. The researchers also determined (1) annual prescription refill rates for individual antiretrovirals, (2) overall combination antiretroviral adherence, and (3) annual rates of viremia (a detectable viral load) after combination antiretroviral therapy (ART) began. The VA team used these data to build a Cox model with time-dependent variables calculating the risk of death from any cause while controlling for demographic, immunologic, and virologic baseline variables.

The analysis included 15,714 veterans with an antiretroviral-induced viral load below 400 copies. Each participant had a minimum follow-up of 1.5 years, and median follow-up extended to 3.3 years (interquartile range [IQR] 2.0 to 5.8). Median age when ART began was 47.6 years (IQR 41.3 to 53.8). Median age for all participants was 44 years in 1996 and 56 years in 2009. While 52% of these people were African American, 48% were European American.

Baseline CD4 count was 260 (IQR 121 to 417), baseline CD8 count 1116 (IQR 782 to 1548), and baseline CD4-/CD8- count 367 (IQR 246 to 543). Baseline viral load was 20,000 copies. Median year of first ART was 2001 (IQR 1998 to 2004).

During follow-up, 1139 veterans died. Unadjusted mortality according to the year ART began fell progressively from 29.6 per 1000 patient-years in 1996 to 14.1 per 1000 patient-years in 2008. CD4 counts rose during the first 10 years of ART and reached a median of 601. Median CD8 counts remained elevated (above 800).

Veterans who reached the highest CD4 quintile (above 700) had mortality rates similar to those of a gender-, race-, and age-matched sample of the US population in 2007 (10.7 per 1000 patient-years). But the likelihood of boosting the CD4 count above 700 depended on pretreatment CD4 count. Fewer than half of veterans starting ART with a CD4 count of 350 to 499 lifted their count above 700 through 10 years of treatment.

The researchers devised two multivariate models, one without CD4-negative lymphocyte subpopulations and one with those subpopulations. Results were largely consistent with both models. In the model without CD4-negative subpopulations, failure to reach a CD4 count of at least 700 independently raised the risk of death, even for people with who reached counts of 510 to 699:

Hazard ratio for death compared with reaching 700 or more CD4s:

-- 510 to 699 CD4s: HR 1.39

-- 380 to 509 CD4s: HR 1.45

-- 240 to 379 CD4s: HR 1.62

-- Under 240 CD4s: HR 2.60

Other independent predictors of death were older age (HR 1.91 per decade), HCV coinfection (HR 1.68), HBV coinfection (HR 1.22), and use of stavudine (HR 1.32). Every 10% better antiretroviral adherence lowered the death risk 5% in this model. Gender, race, history of AIDS, and pretreatment viral load or viremic episodes did not predict death.

In the model that did include CD4-negative subpopulations, a CD8-cell AUC below 550 raised the death risk by one third (HR 1.33), and a CD4-/CD8- AUC below 300 raised that risk by more than half (HR 1.57).

The VA investigators proposed that "'normalization' of CD4 counts to levels above 500 cells/uL on [combination ART] may not be sufficient to minimize the risk of death."

"Given that more than half of patients may progress to CD4 counts less than 500 cells/uL within a year after HIV seroconversion," the researchers add, "our observations suggest that immediate ART initiation will likely translate into a mortality benefit by maintaining CD4 counts well above 700 cells/uL."

Like most US veterans studies, this one involved a largely male population with free access to antiretrovirals and HIV care. So results of this study may not apply to HIV-positive women or to the other populations of HIV-positive men. In addition, the researchers were unable to factor in the impact of some important variables that affect mortality, such as smoking.

An earlier non-VA study in five US cohorts focused on 365 people who kept their viral load below 1000 copies for at least 4 years [2]. This analysis found that 44% of participants who began treatment with a CD4 count below 100 and 25% who started with 100 to 200 CD4s could not push their CD4 count above even 500 through an average follow-up of 7.5 years [2].

Just before the International AIDS Conference, French researchers published findings on a 24,037-person analysis of Africans starting their first antiretroviral combination [3]. Compared with those who reached a CD4 count above 500, those who reached counts between 350 and 499 had a 70% higher risk of death (adjusted HR 1.70, 95% CI 1.26 to 2.30) and those who reached counts between 200 and 349 had more than a doubled risk of death (adjusted HR 2.56, 95% CI 1.93 to 3.38).


1. Drechsler H, Zhang S, Holodniy M, Bedimo R. Immune reconstitution on HAART defines survival in US veterans. XIX International AIDS Conference. July 22-27, 2012. Abstract MOPE113.

2. Kelley CF, Kitchen CM, Hunt PW, et al. Incomplete peripheral CD4+ cell count restoration in HIV-infected patients receiving long-term antiretroviral treatment. Clin Infect Dis. 2009;48:787-794.

3. Maman D, Pujades-Rodriguez M, Nicholas S, et al. Response to antiretroviral therapy: improved survival associated with CD4 above 500 cells/ml. AIDS. 2012;26:1393-1398.