Affordable Access

Publisher Website

Inflammatory biomarkers prior to antiretroviral therapy as prognostic markers of 12-month mortality in South Africa and Uganda.

  • Siedner, Mark J1, 2, 3, 4
  • Bwana, Mwebesa Bosco3
  • Asiimwe, Stephen3
  • Musinguzi, Nicholas3
  • Castillo-Mancilla, Jose5
  • Amanyire, Gideon3
  • Tracy, Russell P6
  • Bangsberg, David R7, 8
  • Orrell, Catherine9
  • Haberer, Jessica E1, 2
  • 1 Harvard Medical School, Boston, MA.
  • 2 Massachusetts General Hospital, Boston, MA.
  • 3 Mbarara University of Science and Technology, Mbarara, Uganda. , (Uganda)
  • 4 Africa Health Research Institute, Kwa-Zulu Natal, South Africa. , (South Africa)
  • 5 University of Colorado, Denver, CO.
  • 6 University of Vermont, Burlington, VT.
  • 7 Brigham and Women's Hospital, Boston, MA.
  • 8 Oregon Health Sciences University-Portland State University School of Public Health, Portland, OR.
  • 9 University of Cape Town, Cape Town, South Africa. , (South Africa)
Published Article
AIDS (London, England)
Publication Date
Jul 02, 2019
DOI: 10.1097/QAD.0000000000002305
PMID: 31274541


To determine the utility of biomarkers of immune activation, systemic inflammation, and coagulopathy prior to antiretroviral therapy to predict mortality during the first year of antiretroviral therapy (ART) in sub-Saharan Africa DESIGN:: Prospective, observational cohort METHODS:: We measured soluble CD14, interleukin-6, and D-dimer in non-pregnant individuals initiating ART in South Africa and Uganda in the Measuring Early Treatment Adherence (META) Study. We used survival analysis methods to estimate their association with 12-month mortality, and fit receiver operator curves (ROC) to assess the prognostic value of each biomarker. Six-hundred sixty individuals were enrolled and had pre-treatment biomarkers measured. Approximately 60% were female, with a median CD4 of 187 (IQR 111-425) and approximately half were enrolled each from South Africa and Uganda. We observed 34 deaths for a crude mortality of 5.3 deaths/100 person-years (95%CI 3.8-7.4), which ranged from 0/100py to 13.7/100py in the lowest and highest tertile of pre-treatment sCD14, respectively. In Cox models, all three biomarkers were strongly predictive of the hazard of death (AHR 3-6, all P < 0.01). In multivariable models including biomarkers, both pre-treatment CD4 count and pre-treatment viral load became borderline or non-significantly associated with mortality. The c-statistic for area under ROC was higher for all three biomarkers than for CD4 count (P < 0.01). Biomarkers of immune activation, systemic inflammation, and coagulopathy prior to ART initiation are strongly predictive of early death on treatment after adjustment for CD4 count. Such biomarkers might serve as important prognostic indicators for patient triage in this population.

Report this publication


Seen <100 times