Bacillus Calmette-Guerin (BCG) Vaccine and Morbidity From Malaria

Background

Studies undertaken in Guinea Bissau have shown that the protective effect of BCG vaccine against death in young children is greater than would be expected from its effect in preventing tuberculosis. How this might be achieved is not known but BCG could act by enhancing the immune response to common infections including malaria which is an important cause of death in young children in Bissau. There is some laboratory evidence to support this hypothesis. BCG prevents malaria in rodents, probably by activating their reticulo-endothelial system and Freund's adjuvant, an emulsion of tubercle bacilli in oil, prevents malaria in primates.

In most developing countries, BCG is given as soon as possible after birth and is an integral part of the immunisation programme so that a controlled trial which withheld neonatal BCG vaccination would not be ethical. However, it is uncertain how long the non-specific protective effect of BCG persists and whether this could be sustained by a further injection of BCG in the second year of life. A major trial to investigate this possibility is under way in Guinea Bissau with mortality as its main end-point. The current study has made use of this opportunity to investigate whether or not re-vaccination with BCG has a protective effect against malaria.

Objective

The primary objective of the study was to determine if re-vaccination with BCG reduced morbidity from malaria in young children.

Study area

The study was carried out within the area covered by the Bandim Health Project in Bissau, the capital of Guinea Bissau, West Africa. The study area, within the suburbs of Bissau, has a population of approximately 70,000 who are under demographic surveillance. Routine infant immunisations are given at 3 health centers in the area; babies who deliver in the national hospital are given BCG before they leave. Overall coverage with BCG is high (> 95%).

Study population

All infants resident in the study area were eligible for inclusion in the main BCG re-vaccination trial. Children were recruited to the malaria sub-study during the period January to October 2003 at the age of 19 months when they were due to receive BCG re-vaccination or to remain as a control. Separate, written, informed consent was obtained for inclusion in the malaria sub-study.

Study procedure.

At the age of 19 months, children whose parents gave consent were allocated randomly to receive BCG revaccination or to be a control and entered into the malaria study. Because BCG leaves a scar, it was not possible to undertake a double blinded study. However, the field staff who assessed the patients for clinical attacks of malaria were unaware of the hypothesis underlying the trial and blood films were read without knowledge of an individual's study group.

The prevalence of malaria in children in the two groups was determined through a combination of active and passive surveillance. Any child in the study who presented to one of the dispensaries in the study area or to the national hospital with an illness suggestive of malaria was investigated for possible malaria. If a child had a history of fever or a raised axillary temperature (> 37.5 degrees Celsius) blood films were obtained and the hemoglobin measured. Children found to have malaria were treated with chloroquine which was the first line treatment at the time of the study. In addition to passive surveillance, two cross sectional surveys of all children in the study were undertaken at the beginning and at the end of the malaria transmission season, during which blood films were collected and the hemoglobin measured. Quality control for the reading of blood films was provided by staff of the MRC Laboratories, The Gambia.

Some of the study children with clinical malaria were enrolled in a separate case control study of risks for malaria.

Trial end-point

The primary trial end-point was the incidence of malaria in children in the two study groups. Malaria was defined as the presence of (a) fever or a history of fever,(b) no other obvious cause for the fever and (c) the presence of asexual stage Plasmodium falciparum asexual parasitemia of any density. For a second definition, a parasite density of 5,000 parasites per ul or >, the estimated fever threshold for the study area, was required.

Sample size

On the basis of previous experience in the study area, it was estimated that the incidence of malaria in children in the control group would be about 0.3 attacks during the malaria transmission season. On this basis a study with 600 children in each arm would have 80% power at the 5% level of significance to detect a reduction in the incidence of malaria of 25% in the children who received BCG. A trial of this size would have 80% power to detect a reduction in the prevalence of parasitaemia of 22% at the final cross-sectional survey assuming a prevalence in the control arm of 35% asexual P. falciparum parasitemia, an assumption based on the results of a previous survey.

Monitoring

A Data Safety and Monitoring Board (DSMB) provided supervision for the study and approved the analytical plan before the trial code was broken. Monitoring visits were made by scientists from the UK and from The Gambia.