Cotrimoxazole prophylaxis reducing child deaths by preventing bacterial lung infections

Cotrimoxazole prophylaxis in children seems to reduce the risk of illness or death primarily by reducing the number of lower respiratory tract infections rather than by an effect on PCP pneumonia, according to an analysis of the CHAP trial published in the January 2^nd edition of the journal AIDS.

The benefits of cotrimoxazole (CTX) prophylaxis in adults and children living with AIDS has been confirmed in several African countries including Cote d’Ivoire, South Africa, Malawi, Zambia, and Uganda. In particular the CHAP study, conducted by the UK’s Medical Research Council in Zambia and Malawi, showed that cotrimoxazole prophylaxis reduced the risk of death in children with HIV by half over two years of follow-up.

Nevertheless there are outstanding scientific questions about the mechanisms of action of CTX prophylaxis in adults and children living with AIDS. Researchers in the CHAP study could not explain why cotrimoxazole prophylaxis reduced illness and death, given that only one case of PCP-related pneumonia could be detected, and so carried out a more detailed analysis of causes of death, hospitalisation and antibiotic use to see if they could detect where cotrimoxazole might be having a greater effect on morbidity and mortality.

Causes of hospital deaths were assigned by the attending paediatrician. If the child died before seeing the paediatrician, this was done by clinical officers based on information from hospital notes and nursing staff. All hospital deaths were reviewed and a primary cause of death was assigned by two paediatricians blinded to the treatment group.

The mortality review was based on case report forms, death certificates, hospital notes, post-mortem findings, and laboratory data. Agreement between reported and reviewed causes was relatively high. Hospital admission case report forms were validated against hospital notes by an independent trial manager and trained data monitor.

Investigations such as nasopharyngeal aspirates (NPAs) for suspected respiratory infections, blood cultures, chest X-rays, thin/thick films for malaria parasites, cerebro-spinal fluid (CSF), urine, and other samples for microcopy and culture were carried out on CLWA admitted to hospital whenever possible. Post-mortems were carried out in only twelve cases.

Tuberculosis (TB) was diagnosed by examination for acid-fast bacilli in NPAs, sputa or gastric washings. All NPAs were examined for Pneumocystis jiroveci using immunofluorescence and polymerase chain reaction (PCR). Serious bacterial infections were defined as presumptive or definitive bacterial infections (septicemia, meningitis, septic arthritis, pneumonia, empyema) requiring hospitalization.

Laboratory analyses of 13 blood cultures, 34 malaria films, and five CSF samples were carried out during the month preceding death in 40 children who were hospitalized and subsequently died. By contrast, in 132 children who were hospitalized but survived admission, 98 blood cultures, 166 malaria films, and five CSF were analysed in the month preceding or during 187 hospital admissions.

Overall, 28 and 42 % of CLWA on CTX and placebo, respectively, died. This included 15 and 21 % of hospital deaths in CTX and placebo, respectively, and 28 and 42 % of deaths outside hospital in CTX and placebo, respectively. The number of days spent in hospital was significantly higher in the placebo group by comparison with the CTX group (1043 versus 881 days; P = 0.03).

There were 91 hospital deaths, the leading causes of which were serious bacterial infections (49.5 %), diarrhoea (13.2 %), and malnutrition (7.7%). Pneumonia or empyema accounted for 29 and 39 % of hospital deaths in CTX and placebo, respectively. There was no evidence that mortality reductions in the CTX group varied by the cause of disease.

When hospital admission by cause was considered, pneumonia/empyema was the most predominant, either singly or concurrently with TB, malaria, and malnutrition. Staphylococcus aureus and Salmonella species were the major bacterial isolates identified from blood cultures. The majority of malaria (51/55) and TB (43/47) diagnoses were presumptive and not laboratory-confirmed. P. jiroveci was rare with 1/115 and 0/73 NPAs being positive by immunofluorescence and PCR, respectively.

Serious bacterial infections and malnutrition admission rates were reduced in the CTX group by comparison with the placebo group. These differences were not significant probably as result of the small sample sizes. However, by two years, the cumulative probability of first admission to or death in hospital with serious bacterial infections was significantly reduced in the CTX group.

There were no differences between CTX and placebo group when age, height-for-age, or CD4 cell percentages among deaths or hospital admissions were considered.

There was a reduction in the total number of days on antibiotics in the CTX group by comparison with the placebo group (7401 versus 7952 days). When specific antibiotics were considered, the reductions were significant for penicillin, chloramphenicol, ciprofloxacillin, cloxacillin, erythromycin, and gentamycin. In the children who died outside of hospital, antibiotic use was again significantly lower in the CTX group.

The data of Mulenga et al provide additional evidence for the protective effect of CTX prophylaxis in CLWA irrespective of the CD4 count and against a high background of antibiotic resistance in Zambia. CTX prophylaxis reduced bacterial infections, hospital admissions and days of hospitalisation, and antibiotic use in CLWA.