icon-    folder.gif   Conference Reports for NATAP  
  21st Conference on Retroviruses and
Opportunistic Infections
Boston, MA March 3 - 6, 2014
Back grey_arrow_rt.gif
Cumulative Viral Load Predicts All-Cause and AIDS
Death But Not Non-AIDS Death
  CROI 2014, March 3-6, 2014, Boston
Mark Mascolini
Cumulative viral load--a sum of detectable viral loads since a person started antiretroviral therapy (ART)--predicted all-cause mortality and AIDS mortality better than pre-ART viral load or most recent viral load--and independently of most recent CD4 count--in a 33,563-person ART Cohort Collaboration (ART-CC) analysis [1]. Cumulative viral load did not predict non-AIDS death.
Over the past few years, studies have suggested the predictive value of measuring cumulative viral load [2,3]. To explore the impact of cumulative viral load in a larger cohort, researchers analyzed data from 17 of 21 ART-CC cohorts, which enroll antiretroviral-naive people across Europe and North America. This analysis involved people who began ART between January 2000 and December 2009. The investigators aimed to assess the impact of cumulative viral load on all-cause mortality, AIDS-related mortality, and non-AIDS mortality.
Cumulative viral load is based on a measure called viremia copy-years, which tally all detectable viral loads in a person for a given period--just as smoking pack-years reflect cumulative time smoking. People with the lowest viremia copy-years are those who reached an undetectable viral load quickly after starting treatment and kept their viral load suppressed. Progressively higher copy-years mean a person had a detectable viral load for progressively longer times, either while taking a failing regimen or during periods when not taking antiretrovirals for any reason. In an earlier report, these investigators offered an example: "10,000 copy-years equals having 10,000 copies of the virus every day for 1 year or 1000 copies of the virus every day for 10 years" [2]. Multivariable Cox models to calculate the impact of viremia copy-years on mortality adjusted for age, sex, HIV transmission risk, entry CD4 count, time-updated (most recent) CD4 count, and cohort.
The 33,563 people studied included 23,803 men (71%) and 9760 women. People infected during heterosexual sex accounted for 46% of the study group, men who have sex with men for 34%, and injection drug users for 11%. Six months after starting ART, 82% of people had a viral load below 500 copies, 7% had a load between 500 and 9999 copies, 7% had between 10,000 and 100,000 copies, and 4% had more than 100,000 copies.
Study participants had an average 13 viral load measures, and follow-up after starting ART averaged 3.2 years. During that time, 1341 people died, including 503 (38%) who died of AIDS, 614 (46%) who died of non-AIDS causes, and 224 (17%) with an unknown cause of death.
The first analysis used a Cox model adjusted for death risk factors listed above. Compared with people who had under 10,000 viremia copy-years, those in each higher copy-year bracket had an independently higher risk of all-cause mortality, at the following adjusted hazard ratios (aHR) and 95% confidence intervals:
Risk of any death compared with <10,000 copy-years adjusted for other prognostic factors:
-- 10,000 to 100,000 copy-years: aHR 1.23 (1.08-1.39)
-- 100,000 to 350,000 copy-years: aHR 1.75 (1.45-2.11)
-- Over 350,000 copy-years: aHR 2.00 (1.52-2.63)
When the Cox model additionally adjusted for all viral load measures, having 100,000 or more viremia copy-years still raised the risk of all-cause mortality, though the association for 10,000 to 100,000 copy-years stopped just short of statistical significance:
Risk of any death compared with <10,000 copy-years also adjusted for all other viral load measures:
-- 10,000 to 100,000 copy-years: aHR 1.14 (0.98-1.33)
-- 100,000 to 350,000 copy-years: aHR 1.56 (1.24-1.96)
-- Over 350,000 copy-years: aHR 1.74 (1.27-2.39)
Repeating these analyses for AIDS-related death, the researchers found that each higher viremia copy-year bracket independently raised chances of an AIDS death compared with under 10,000 copy-years:
Risk of AIDS death compared with <10,000 copy-years adjusted for other prognostic factors:
-- 10,000 to 100,000 copy-years: aHR 1.40 (1.13-1.74)
-- 100,000 to 350,000 copy-years: aHR 2.08 (1.51-2.86)
-- Over 350,000 copy-years: aHR 3.40 (2.23-5.20)
Risk of AIDS death compared with <10,000 copy-years also adjusted for all other viral load measures:
-- 10,000 to 100,000 copy-years: aHR 1.45 (1.12-1.89)
-- 100,000 to 350,000 copy-years: aHR 2.18 (1.46-3.26)
-- Over 350,000 copy-years: aHR 3.64 (2.17-6.10)
Notably, all these associations are independent of pretreatment CD4 count and latest CD4 count.
When the researchers looked only at non-AIDS death, higher viremia-copy year brackets generally did not raise the death risk, with one exception: Compared with under 10,000 viremia copy-years, 100,000 to 350,000 raised the risk of non-AIDS death 41% (aHR 1.41, 95% CI 1.07 to 1.87) in the analysis not additionally adjusted for other viral load measures.
In contrast with viremia copy-years, pre-ART viral load proved a less robust predictor of all-cause mortality and AIDS mortality. Compared with a pre-ART viral below 100,000 copies, a load above 100,000 copies raised the risk of death from any cause or death from AIDS in the Cox model that adjusted for other death risk factors, but not in the model that also adjusted for all other viral load measures. Latest viral load (500 to 9999 copies, 10,000 to 100,000 copies, and more than 100,000 copies, compared with under 500 copies) predicted death better than pre-ART viral load but not as strongly as viremia copy-years.
The investigators proposed that "cumulative viral load measures such as viremia copy-years may have independent and enhanced prognostic value for mortality relative to cross-sectional [snap-shot] measures." Although the researchers did not make this point, the results imply that keeping viral load undetectable consistently after starting ART offers the best chance of longer survival.
1. Mugavero MJ, Westfall AO, Gill MJ, et al. Cumulative viral load predicts all-cause and AIDS-related mortality after initiation of ART. CROI 2014. Conference on Retroviruses and Opportunistic Infections. March 3-6, 2014. Boston. Abstract 565.
2. Cole SR, Napravnik S, Mugavero MJ, Lau B, Eron JJ Jr, Saag MS. Copy-years viremia as a measure of cumulative human immunodeficiency virus viral burden. Am J Epidemiol. 2010;171:198-205.
3. Mugavero MJ, Napravnik S, Cole SR, et al. Viremia copy-years predicts mortality among treatment-naive HIV-infected patients initiating antiretroviral therapy. Clin Infect Dis. 2011;53:927-935.