Heart attacks, non-AIDS cancers, more common in Botswana than in US patients on HIV treatment

Incidence rates of non-AIDS-defining events (NADEs) in HIV-infected individuals on ART were 50% higher in a clinical trial cohort in Botswana - compared to a routine care (observational) cohort in the United States (US) - with especially high rates of heart disease and cancer, according to CW Wester and colleagues in a retrospective analysis published in the advance online edition of AIDS.

Liver disease, however, was only seen in the US cohort (4.0 per 1000 person-years).

 Antiretroviral treatment (ART) has significantly reduced HIV-related death and disease over the past two decades in resource-rich settings. However, those on long-term ART have experienced continue to experience some negative health outcomes. These include heart, kidney, liver or non-AIDS-defining cancer illnesses, classified as non-AIDS-defining events.

The causes of NADEs are multiple and can include older age (which may also indicate longer exposure to antiretroviral therapy); smoking or alcohol use; long-term viral replication (lengthy suppression of the immune system); co-infection with chronic diseases (for example, hepatitis); as well as long-term ART complications (for example, diabetes).

US and Western European data have shown that the six-month cumulative risk of death from NADEs was three times greater, compared to AIDS-defining events, in HIV-infected individuals with high CD4 cell counts. Heart and liver disease and non-AIDS-associated cancers were the most common causes.

 While the majority of those on ART live in sub-Saharan Africa, few studies have looked at the effect of NADEs there. 

The authors retrospectively compared crude and standardised (based on the US cohort by age and gender) NADEs rates in HIV-infected individuals on ART in two geographically distinct, adult, urban populations: a randomised clinical trial population in Gaborone, Botswana, and an observational cohort in Nashville, Tennessee, US, with data from 1 December 2002 to 31 December 2007, and 1 January 2003 to 31 December 2007, respectively.

Of the 650 people in the Botswana cohort, 65% were female, median age 33.3 years, with a median CD4 cell count when starting ART of 199 cells/mm³. In contrast, the US cohort of 1129 was predominantly male (75%), older (median age 40.1 years), with a higher median CD4 cell count at the start of ART (243 cells/mm³).

Overall, crude incidence rates for NADEs were similar in the two cohorts: 10.00 per 1000 person-years (95% CI: 6.3-15.9, p=0.20) in Botswana, compared to 12.4 per 1000 person-years (95% CI: 8.4-18.4, p=0.20) in the United States.

 However, when standardised to the US cohort (with an older and mostly male population), the rates were 50% higher in the Botswana cohort compared to the US cohort: 18.7 per 1000 person-years (95% CI: 8.3-33.1, p=0.20) and 12.4 per 1000 person-years (95% CI: 8.4-18.4, p=0.20), respectively.

Standardised rates of heart disease and non-AIDS-defining cancers were higher in the Botswana population: 8.4 per 1000 person-years (95% CI: 2.4-18.4, p=0.20) and 8.0 per 1000 person-years (95% CI:1.3-20.8, p=0.015), compared to 5.0 per 1000 person-years (95% CI: 2.7-9.2, p=0.20) and 0.5 per 1000 person years (95% CI: 0.1-3.5, p=0.015).

Hodgkin’s lymphoma was the primary non-AIDS-defining cancer in Botswana, a finding that suggests it may be a significant cause of non-AIDS-defining mortality in the region, note the authors. With improved diagnostic capacity, they add, cancer registers should be established to better describe cancer disease patterns.

Heart disease was the most common non-AIDS-defining event in both settings, possibly because of ART or the chronic inflammatory state that is present in spite of viral suppression, note the authors.

Regardless of the differences in treatment regimens and socio-demographic factors in the cohorts, deaths linked to heart disease will increasingly be associated with non-AIDS-defining mortality, they add.

Kidney NADEs rates were similar in both groups: 3.0 per 1000 person-years (95% CI: 1.3-6.6, p=0.92) and 2.4 per 1000 person-years (95% CI: 0.0-7.2, p=0.92) in the US and Botswana, respectively. The authors expressed surprise at this outcome. Recent findings showed individuals of African descent infected with HIV to have a twenty-fold greater risk in developing HIV-associated nephropathy (kidney disease) than those of non-African descent.

The authors acknowledge the considerable challenges and limitations in undertaking this comparative analysis. These include, among others, differences in visit schedules, laboratory and clinical monitoring, ART regimens, death information (oral reports in Bostwana and registry reviews and clinician assessments in the US), and diagnostic capacity, as well as differences in the definition of virological failure. The authors add that trying to control for all possible confounders makes no sense.

Despite these limitations, the authors conclude: “NADEs appear to be as significant a problem in our sub-Saharan African setting and the monitoring, prevention and treatment of NADEs should be a critical component of care in resource-limited settings.”