Predictors of mortality among HIV-positive children in resource-limited settings

The first large-scale meta-analysis of prognostic markers among HIV-positive children living in resource-limited settings has provided valuable insight in how to improve care for HIV-infected children. The research, published in the January 2^nd issue of AIDS, confirms the predictive value of CD4 percentages and CD4 counts and reasserts the need to provide affordable and feasible lab tests in the developing world. The study also identifies growth measures as useful predictors and stresses the importance of addressing non-HIV related issues such as malnutrition and anaemia in HIV-positive children.

At the end of 2006, nearly 90% of the estimated 2.3 million children living with HIV worldwide were from sub-Saharan Africa. With no anti-HIV therapy, African children born with HIV have a median survival age of two years, compared with eight to ten years for children born in Europe of the US.

Anti-HIV treatment in children faces considerable gaps in knowledge, including when to start therapy in countries where access to treatment is limited. The WHO paediatric treatment guidelines for resource-limited settings are largely based on data from the HIV Paediatric Prognostic Collaborative Study (HPPMCS), a meta-analysis of studies performed in developed world settings. Similar data from resource-limited settings is lacking.

3Cs4kids in resource-limited settings

To begin addressing this gap, the Cross Continents Collaboration for Kids (3Cs4kids) study undertook a large-scale analysis of mortality predictors among children in resource-limited settings. The study team, which included researchers from the MRC Clinical Trials Unit, performed a meta-analysis of ten studies (nine African and one Brazilian) that included individual longitudinal data from 2510 children. Researchers evaluated the value of selected laboratory tests and growth markers in predicting short-term risk of death.

The researchers included the following markers: CD4 percentage (CD4%), CD4 cell count, total lymphocyte count (TLC), haemoglobin, weight-for-age, height-for-age and BMI-for-age. Results were adjusted for use of cotrimoxazole (Septrin), which is estimated to reduce risk of death among HIV-positive children by 43%.

Median age of the children at first measurement was 4.0 years, and median follow up per child was 12.7 months. Median CD4% was 15% and median weight-for-age was a z-score of -1.9 when compared to the UK reference for HIV-negative children.

Overall, there were 357 deaths among 3769 child-years-at-risk. After adjusting for age, cotrimoxazole prophylaxis and study effects, all markers individually predicted 12-month mortality. Statistical analysis revealed that CD4 cell count and CD4% were the strongest predictors, followed by weight-for-age, height-for-age, haemoglobin, BMI-for-age and TLC.

In multivariable analyses, weight-for-age and haemoglobin remained strongly predictive of death, even after controlling for CD4% or CD4 count. Conversely, there was a steady increase in risk with decreasing CD4% or CD4 count at any specific weight-for-age and haemoglobin value. However, both CD4% and CD4 count were less predictive at lower weight-for-age values.

Comparing 3Cs4kids to HPPMCS

While results from the 3Cs4kids study generally supported those from HPPMCS, there were several notable differences. The twelve-month risk of death estimated by CD4%, CD4 cell count and TLC was generally higher with 3Cs4kids than HPPMCS at any given marker value and age. Moreover, the increase in risk with decreasing marker value was more gradual and occurred at higher thresholds in 3Cs4kids.

Current WHO guidelines recommend that the decision to start therapy be guided by clinical stage or monitoring CD4 percentage and count or TLC. However the researchers note, “The steep rise and threshold effect in mortality risk with falling marker values observed in HPPMCS, which influenced the choice of marker thresholds for ART initiation in the WHO guidelines, was less pronounced in 3Cs4kids. Consequently, both CD4% and count were less effective in discriminating between low and high mortality levels for children in resource-limited settings.“

As well, in contrast with the industrialized setting of HPPMCS where TLC was highly prognostic, TLC was a poor predictor of death among children in resource-limited settings. TLC retained only a small independent effect after adjusting for CD4% or CD4 count, and TLC had no additional prognostic value if CD4 count was known.

Importance of nutrition

The 3Cs4kids study also underscored the significance of nutritional support and prevention and treatment of anaemia in HIV-positive children. The risk of death remained high among young children, especially those age one to two years, who were malnourished or anaemic, even at high CD4% or counts. At older ages, CD4% or CD4 count was better at identifying low-risk children if weight-for-age and haemoglobin were included in the assessment.

While conceding that defining thresholds for these markers is difficult due to their complex inter-relationship, the researchers conclude that “for effective care of HIV-infected children in resource-limited settings, prevention and treatment of malnutrition and anaemia need to be integrated within routine clinical management.”

Need for better lab tests

“The strong effect of CD4% and CD4 cell count over and above other markers underlines the importance of access to low-cost laboratory monitoring,” write the study authors. They also note that while both CD4% and CD4 count were similarly prognostic, CD4 count is difficult to interpret in young children due to the dramatic decline in early in life. “The financing of currently available CD4% technology and the development of more affordable and feasible technology for resource-limited settings is therefore a priority,” they urge.

Informing guidelines

In agreement with current guidelines, the 3Cs4kids study supports the use of CD4% and CD4 count as predictors of short-term risk of death. The researchers add that any African trials of treatment initiation strategies should include growth markers and haemoglobin and that the timing of nutritional of nutritional care and antiretroviral treatment should be clarified.