EMPIRICAL - Empirical treatment against cytomegalovirus and tuberculosis in severe pneumonia in HIV-infected infants: a randomized controlled clinical trial
Pneumonia is the main cause of death in HIV-infected children. A significant number of undiagnosed or poorly treated HIV-infected children present to health services with severe pneumonia. WHO guidelines to treat severe pneumonia in HIV-infected infants include empirical treatment against common bacteria plus Pneumocystis jirovecii. Although this approach has contributed to reduce overall case fatality rates, mortality in this particularly vulnerable group remains unacceptably high. Autopsy studies in Africa have shown that cytomegalovirus infection and tuberculosis are important underdiagnosed and undertreated causes of death, each accounting for up to 20% of mortalities in HIV-infected infants. Un-controlled studies from South Africa have shown that ganciclovir can successfully treat cytomegalovirus pneumonia, but there have been no clinical trials demonstrating efficacy in HIV-infected African children to date. The prevalence of unrecognized tuberculosis in HIV-infected infants with severe pneumonia has been estimated to be around 18%. However, this is likely an underestimate, as >50% of children who die of tuberculosis had not been diagnosed ante-mortem. Despite the high mortality associated with tuberculosis and cytomegalovirus, in many resource-limited settings there is no diagnosis in place and available diagnostic tools perform poorly. Our objective is to evaluate whether empirical treatment against cytomegalovirus and tuberculosis improves survival of HIV-infected infants with severe pneumonia. A randomized factorial clinical trial will be conducted in six sub-Saharan African countries to evaluate the safety and efficacy of empirical treatment against cytomegalovirus and tuberculosis in HIV-infected infants aged 1 month-12 months admitted to hospital with severe pneumonia. Study end-points include 15-day and 12-month survival. HIV-infected infants will receive standard of care (SoC) pneumonia treatment including antibiotics, cotrimoxazole and prednisolone. Infants with presumptive tuberculosis will receive tuberculosis treatment and will be randomized to receive either SoC plus valganciclovir or SoC. Patients with a diagnosis of non-presumptive tuberculosis will be randomized to receive SoC plus valganciclovir and tuberculosis treatment in a 2x2 factorial randomization. Causes of death among study participants will be studied using verbal autopsies and minimally invasive autopsies. A drug-to-drug interaction sub-study of antiretrovirals/rifampicin will be assessed in a sub-sample of participants. This project has a high relevance related to the EDCTP call topic because it focuses on the management optimization of three of the diseases included in the call: HIV, tuberculosis and lower respiratory infections, which often overlap. We propose an innovative empirical management approach that could result in a decrease of mortality in this highly vulnerable population group.