Advent of HAART Associated With Shift in Causes of Death in HIV-Infected Individuals

June 1, 2002

SEATTLE-The proportions of deaths in HIV-infected individuals caused by non-AIDS-related diseases have increased since the advent of highly active antiretroviral therapy (HAART), according to a study presented at the 9th Conference on Retroviruses and Opportunistic Infections (abstract 14). Mitchell I. Wolfe, MD, MPH, medical epidemiologist, Division of HIV/AIDS Prevention, Centers for Disease Control and Prevention (CDC), presented the data.

SEATTLE—The proportions of deaths in HIV-infected individuals caused by non-AIDS-related diseases have increased since the advent of highly active antiretroviral therapy (HAART), according to a study presented at the 9th Conference on Retroviruses and Opportunistic Infections (abstract 14). Mitchell I. Wolfe, MD, MPH, medical epidemiologist, Division of HIV/AIDS Prevention, Centers for Disease Control and Prevention (CDC), presented the data.

Dr. Wolfe and his colleagues compared proportions of death from various causes for 4,870 HIV-infected individuals dying in the pre-HAART era (1992-1995) and 2,318 HIV-infected individuals dying in the HAART era (1996-2000). The data were obtained from an ongoing national cohort study that has followed more than 54,000 HIV-infected individuals since 1990. The deaths are based on causes of death from the death certificate or significant diagnoses at death based on medical record review.

After the advent of HAART, there was a significant decrease in the proportion of deaths caused by tuberculosis (-59%), nontuberculosis mycobacterial disease (-30%), Pneumocystis carinii pneumonia (-30%), and toxoplasmosis (-30%).

In contrast, there was a significant increase in the proportion of deaths caused by liver disease (70%), non-Hodgkin’s lymphoma (43%), cachexia/wasting (22%), kidney disease (27%), and sepsis (10%). There was also a nonsignificant 80% increase in the proportion of deaths caused by ischemic heart disease.

The observed decrease in proportion of deaths from AIDS-related causes is consistent with published reports and national data from death certificates. In 1995, there were over 45,000 reported deaths with HIV infection listed, compared with just over 15,000 such deaths in 1999. "This decrease is largely attributable to decreases in deaths due to HIV-related causes," Dr. Wolfe said.

The observed increase in proportion of deaths from non-AIDS-related causes did not, in most cases, reflect an increase in the absolute incidence of death from these causes, Dr. Wolfe cautioned.

Nonetheless, the data support closer scrutiny of two factors that may now have a relatively greater influence on the survival of HIV-infected patients: adverse effects of antiretroviral drugs and prolongation of life, which opens the window for other causes of death.

"There have been some reports in the literature of adverse events possibly associated with HAART; these include reports of renal disease, liver disease, heart disease, lipodystrophy, and diabetes," Dr. Wolfe said. "Furthermore, people may be living long enough to acquire other diagnoses at death due to aging or lifestyle factors . . . for example, hepatitis C due to intravenous drug use, or cirrhosis due to alcohol use."

The findings highlight several areas for future research, Dr. Wolfe said. It will be important to pin down the absolute incidences of death from specific causes in HIV-infected individuals in order to gain a better understanding of why proportional causes of death have changed. Ongoing surveillance will also be paramount in this population. "We should continue to monitor trends to detect additional evidence of changing diagnoses at death. The lack of long-term outcome data from clinical trials means that observational cohort data may be useful to detect changes in cause of death or adverse events from therapy," he said.

Dr. Wolfe and his colleagues are extending their research by comparing diagnoses at death in HIV-infected individuals in the early and recent HAART periods. The findings also have clinical implications with respect to health promotion and routine preventive care in HIV-infected individuals, he added.