We found a mortality of 6.6% with 8.8% LTFU, giving an overall mortality rate of 1.9 per 100 child-years among Asian children who received any level of ART in this regional cohort. The 5-year survival rate from the time of ART initiation was 91.7% (95%CI 90.0-93.2%). Mortality was highest in the first three months after ART initiation and fell with time on ART, and was frequently caused by bacterial infections, Pneumocystis carinii pneumonia, and tuberculosis.
In comparison to our results, the KIDS-ART-LINC collaboration reported 6.9% mortality after ART initiation and 10.3% LTFU after two years in West Africa.19
Reduction in mortality with the availability of ART has been shown in several studies from developed countries1-4, 20-22
as well as from resource-limited settings.19, 23,24
Before ART became available to children in this region, survival of HIV-infected children with advanced stage disease was low. In a study of 353 symptomatic HIV-infected children cared for at Chiang Mai University Hospital, Thailand between 1989 and 1994, 42% of them died during their first hospital admission.25
In a study at a tertiary-care hospital in northeastern Thailand between 1989 and 1997, among 90 HIV-infected children with AIDS, 50% of them died with 30% LTFU. The 1-year survival rate from the time of first symptom was 75.3% (95% CI 65.8%-84.7%).26
In our study, the predictors of mortality in ART-experienced children included WHO clinical stage 4, low CD4% (as a time-dependent variable), and low weight-for-age Z score. Children who died in the first three months of ART had very low baseline weight-for-age z-score (i.e., median -5.1), although not significantly lower than those who died during other periods. Malnutrition can play a critical role in HIV infection, as it impairs both humoral and cell-mediated immunity, while infection itself also contributes to malnutrition.27,28
The type of initial ART regimen (i.e., cART vs. mono- or dual-NRTI) was not associated with increased mortality. This may be related to subsequently switching from mono- or dual-NRTI regimens to cART during the follow-up period; as only 0.5% children who were previously initiated on mono- or dual-NRTIs remained on those regimens. Unlike studies from sub-Saharan Africa, we did not find baseline severe anemia as an independent risk factor for mortality.19
Excess early mortality in the first three months on ART has been reported both in children and adults.19,24,29-30
In a prospective cohort study of 135 HIV-infected Kenyan children initiating cART between August 2004 and December 2008 and followed for a median of 21 months, 20 (13%) children died; the overall mortality was 8.4 per 100 child-years.29
However, the mortality rate was 46 per 100 child-years in the first four months and dropped to 1.0 per 100 child-years from four to 24 months.29
A Medecins sans Frontières study including 3936 children <5 years of age in Africa and Asia reported 6.3% mortality at a median follow-up time of 10.5 months. The survival rate at one year was 93% (95% CI 92-94%), and 55% of deaths and LTFU occurred during the first three months of ART.30
This study has several limitations. The data were both retrospectively and prospectively collected, therefore specific causes of death and outcomes of children LTFU were not fully ascertainable. In addition, although there were at least three patients with suspected immune reconstitution syndrome, we were unable to assess the impact of immune reconstitution on mortality. The lack of data on adherence and viral load are other factors that limited our analysis. The children in our cohort were seen at regional pediatric referral centers, so results may not be generalizable to children receiving care in primary care centers or rural areas. Their acuity was high, as they frequently presented late to care and may have been transferred from centers less equipped to manage complicated opportunistic infections. However, some of these factors are challenging to control in the analysis of observational databases from a resource-limited setting. As the largest analysis of survival and mortality in HIV-infected children across Asia, these findings contribute important evidence on which to consider future management strategies.
Survival of HIV-infected children in Asia has improved substantially as ART has become more widely available in the region. The findings that advanced clinical and immunologic disease predicted mortality emphasize the urgent need for early identification of HIV-infected infants and children and access to cART before disease progression. The frequent association of low weight-for-age with poor outcomes in pediatric cohorts further highlights the importance of integrating nutritional interventions into ART programs to reduce early mortality.