Castelnuovo B, Manabe Y, Kiragga A, et al. Cause-specific mortality and the contribution of immune reconstitution inflammatory syndrome in the first 3 years after antiretroviral therapy initiation in an urban African cohort. Clin Infect Dis. 2009 Sep 15;49(6):965-72.
Although the provision of antiretroviral therapy (ART) for HIV treatment in resource-constrained areas has resulted in decreased mortality,(1,2) early mortality rates among persons receiving ART appear to be higher than expected. The causes of early mortality have not been well described.(3,4,5) Speculation regarding the reasons include severe immunocompromise, high background prevalence rates of Mycobacterium tuberculosis and Cryptococcus neoformans, and unmasking and paradoxical immune reconstitution inflammatory syndrome (IRIS).
To determine cause-specific mortality and the contribution of IRIS to early mortality among persons receiving ART
ART-eligible but -naïve adults who initiated ART from April 2004 through April 2005 and who had regular attendance at the HIV care clinic (defined as two visits in the previous six months), stable residence within 20 km of the clinic, and willingness to be followed for two years
All cause mortality rate, cause-specific mortality, and risk factors for death within 36 months of ART initiation
Participants received trimethoprim-sulfamethoxazole prophylaxis and ART. Patients were required to have completed tuberculosis therapy before initiating ART. A complete blood count, CD4 test, HIV viral load test, and serum aspartate transaminase, and creatinine levels were measured every three months.
All deaths of cohort participants within the 36 months following initiation of ART were recorded. Verbal autopsies, medical records, and inpatient hospital records were used by one of the study physicians to determine the cause of death.
Deaths were classified as HIV-related (due to opportunistic illnesses [OIs]), ART toxicity, or other medical conditions. For all HIV-related deaths, a determination was made regarding whether the death was associated with IRIS which were classified as unmasking (new onset with atypical presentation) or paradoxical (worsening of a previously treated OI) within six months of initiating ART. The criteria were based upon modification of IRIS case definitions.(6,7) Two investigators reviewed the medical records of each suspected IRIS case independently. In the case of discrepant determinations, a third investigator made a final determination.
The cohort was 69% female and the median age was 38 years. Eighty-nine percent of patients met the WHO clinical criteria of stage 3 or 4 disease at the time ART was started.
Deaths totaled 99 (17%) during the study period; 80 (14%) died within the first 12 months and 73% of these patients died within the first month. Overall mortality was 12.2 deaths per 100 person-years at risk. Eighty-six percent of the 80 deaths in the first year were HIV-related and four of these were due to IRIS; all were unmasking of OIs. Three of the 80 deaths were due to ART toxicity.
There were 15 deaths in the second year and four (27%) were HIV-related and four were due to ART toxicity. There were four deaths in the third year and three of these were HIV-related.
Of the 76 HIV-related deaths, 18 were from central nervous system infections, 16 were due to tuberculosis, seven to Kaposi sarcoma, five to P. jiroveci pneumonia, four to chronic diarrhea, five to chronic anemia, and two were due to cervical cancer. There were 18 deaths due to undiagnosed infections. The two most common deadly pathogens were M. tuberculosis and C. neoformans.
Independent risk factors for death were CD4 cell count of <25 cells/µL, body mass index of <18, and hemoglobin levels of <8 g/dL. The probability of death was significantly different among the different CD4 cell count strata, with the lowest survival among patients who initiated ART with a CD4 cell count of <25 cells/µL. When baseline CD4 cell count is separated into quartiles, the same trend is observed, with a statistically significant association with death for patients in the lowest quartile.
Early mortality in this cohort is due to HIV-related conditions and immunosuppression and IRIS is not a major contributor.
Based on the Newcastle-Ottawa scale, this study was of very good quality. The case definition was adequate; ascertainment of outcome appears adequate and was clearly not present at the onset of the study. The follow-up period was appropriate. Unfortunately, it was not possible to assess representativeness of the cohort.
This study demonstrates that HIV-infection is the primary cause of early deaths in resource-constrained areas due to initiation of ART late in the course of disease. The limited contribution of IRIS to early mortality and the causes of deaths identified highlight the need for early diagnosis of HIV and initiation of ART prior to severe immunocompromise.
- Coetzee D, Hildebrand K, Boulle A, et al. Outcomes after two years of providing antiretroviral treatment in Khayelitsha, South Africa. AIDS 2004;18:887-95.
- Ivers LC, Kendrick D, Doucette K. Efficacy of antiretroviral therapy programs in resource-poor settings: a meta-analysis of the published literature. Clin Infect Dis 2005;41:217-24.
- Moh R, Danel C, Messou E, et al. Incidence and determinants of mortality and morbidity following early antiretroviral therapy initiation in HIV-infected adults in West Africa. AIDS 2007;21:2483-91.
- Braitstein P, Brinkhof MW, Dabis F, et al. Mortality of HIV-1-infected patients in the first year of antiretroviral therapy: comparison between low-income and high-income countries. Lancet 2006;367:817-24.
- Lawn SD, Harries AD, Anglaret X, Myer L, Wood R. Early mortality among adults accessing antiretroviral treatment programmes in sub-Saharan Africa. AIDS 2008;22:1897-908.
- Meintjes G, Lawn SD, Scano F, et al. Tuberculosis-associated immune reconstitution inflammatory syndrome: case definitions for use in resource-limited settings. Lancet Infect Dis 2008;8:516-23.
- French MA, Price P, Stone SF. Immune restoration disease after antiretroviral therapy. AIDS 2004;18:1615-27.