Malaria accounts for an estimated 1 to 3 million deaths each year, with the majority occurring in children under five years of age in sub-Saharan Africa 1. Vaccine-preventable diseases cause an estimated 1 to 2 million deaths in African children 2. The WHO's Expanded Program on Immunization (EPI) is targeted at malarious areas, emphasizing the need to understand the effect of malaria and antimalaria drug use on vaccine immunogenicity and efficacy. Accordingly, a study that began in 1975 has been fully analysed following great increasing recent interest in the important topic of malaria chemoprophylaxis and, in particular, intermittent preventive (malaria) therapy of infants (IPTi) 34567.

Acute malaria has been associated with a decreased response to tetanus toxoids, and meningococcal polysaccharide, Hib conjugate, and whole cell vaccines for typhoid fever 8910. Asymptomatic parasitaemia has been associated with a decreased response to the newer acellular pertussis and meningococcal vaccines, suggesting a benefit from malaria prophylaxis prior to vaccination 111213. Other studies have shown that asymptomatic parasitaemia or anti-malarial drug administration does not inhibit vaccine response to various live, attenuated, whole-cell killed, and toxoid vaccines 4514151617181920. No human studies have looked at the association between parasitaemia and the serologic response to whole-cell pertussis vaccine, a product still used in many vaccination programmes, particularly in developing countries. Antimalarials may also depress vaccine response as illustrated by the immunodepressive effect of 4-aminoquinolones1321222324.

The study aimed to determine the effect of malaria chemoprophylaxis on vaccine seroconversion or seroresponse to live, attenuated measles vaccine, diphtheria and tetanus toxoids and whole-cell pertussis (DTP) vaccines.

The groups were similar at baseline with regard to age, sex, and nutritional status, except for a slight excess of males in the CH- group for DTP (Table 1). Twenty percent of the children were <12 months of age (N = 202). Figure 1 shows the distribution flow of children receiving or not receiving chemoprophylaxis by vaccine type. Although the vaccine was administered as combined DTP, the number of children with serological data that were evaluated differed for the three DTP assays as defined above. The final number of children analysed reflects the availability of paired sera, or loss due to moves or death (Figure 1). When comparing the compliant CH+ children to the non-compliant CH+ children, there were no significant differences in sex, age, or weight-for-height Z-score for those children receiving measles vaccine (P = 0.33, 0.56, 0.52 respectively) or in sex for children receiving DTP (P = 0.22). For the DTP group, age and weight-for-height Z-score was significantly less in the noncompliant CH+ group (P < 0.01, P = 0.02, respectively).

Table 2 shows that seroconversion rates to measles vaccine and seroresponse to diphtheria and tetanus vaccines were not significantly different in the CH+ and CH- groups, both when all children were included and when non-compliant CH+ children were excluded from the analysis (P > 0.05). When all children were included in this analysis, there was a lower rate of seroconversion to diphtheria and tetanus in the CH+ group, but this difference was not statistically significant. Percent seroresponse to pertussis was greater in the CH- group (P < 0.01). In this cluster analysis, there was adjustment for the random effect of village. When the analysis was done without adjusting for the random effect of village, there was a significantly greater rate of seroconversion to diphtheria and tetanus in the CH- group. When non-compliant children were excluded, the difference was no longer significant for diphtheria and tetanus. Percent seroresponse to pertussis remained greater in the CH- group (P < 0.01).

For measles, pre-vaccination measles titers for all children were <1:10 (lowest detectable titer) (Figure 2); GMR was not significantly different in the CH+ vs. CH- group (P = 0.44). For all three antigens GMR was higher in the CH- group, but this difference was statistically significant only for pertussis (P < 0.01).

Multivariate logistic regression analysis indicated that for tetanus, a lower pre-vaccination GMT was positively associated with seroresponse (P = 0.02); for pertussis, a lower pre-vaccination GMT (P < 0.01) and younger age (P = 0.04) were positively associated with seroresponse. There was no significant difference in pre-vaccination pertussis titres between the CH+ and CH- group (P 0.22) when looking at all children regardless of compliance; however, when excluding non-compliant children, pre-vaccination pertussis titres were higher in CH+ children (P <0.01). While pre-vaccination pertussis titres were higher in compliant CH+ children (log₂ of GMT = 6.19, N = 37) compared to non-compliant CH+ children (log₂ of GMT = 5.80, N = 84), this difference was not significantly different (P = 0.80). Chemoprophylaxis was not associated with seroresponse for any of the vaccines. Thus, especially for pertussis, the lower vaccine response rate observed in the CH+ group appears to be due, in part, to the greater proportion of subjects with high pre-immunization titers.

Malaria antibody titers were significantly lower in the CH+ group compared to the CH- group following seven months of prophylaxis. GMTs for children receiving measles vaccine were: CH+, 196 (N = 128) vs. CH-, 1089 (N = 219) (P < 0.01) using IFA and CH+, 285 (N = 60) vs CH-, 1990 (N = 64) using ELISA (P < 0.01) and for children receiving DTP vaccine: CH+, 109 (N = 132) vs CH-, 178 (N = 178) (P < 0.01) using IFA and CH+, 86 (N = 63) vs CH-,153 (N = 30) (P = 0.01) using ELISA. Only 13 percent of all CH+ children with detectable malaria titers prior to chemoprophylaxis had undetectable titers post-chemoprophylaxis (N = 159). This indicates that chemoprophylaxis given to young children for five to seven months after previous exposure to malaria was not adequate to eliminate malaria antibodies.

No adverse events were recorded after chemoprophylaxis, blood sample collection, or vaccination other than a few children with 1–3 mm nodules on their arms after receiving the vaccine by injector and one child who developed a cellulitis where the venopuncture occurred: this child recovered with systemic antibiotic treatment.

Proposed mechanisms for malaria-associated immunodepression include impaired macrophage function 394041, altered cytokine production 3942, a depletion of T or B cells 43, impaired dendritic cells424445, elevated nitric oxide production 46 and elevated prostaglandin E during febrile malaria47. Clinical evidence includes an association of malaria with increased susceptibility to bacterial infections 48, reactivation of viral infections 4950, a low prevalence of autoimmune disease in endemic areas5152, and reports of decreased responses to vaccinations.

In contrast to asymptomatic parasitaemia, acute malaria impairs vaccine response891012171820. In vitro challenge studies in individuals with acute malaria have demonstrated a depression in the cellular immune response involving alterations in lymphoproliferation and cytokine responses 425354. The pyrogenic cytokine TNF-alpha is elevated in febrile malaria and may depress humoral immunity by impairing antigen handling by dendritic cells. T-cell levels, CD4 cells in particular, are depressed 55. IL-1, in addition to TNF-alpha, is elevated in acute illnesses 5657. Both promote increased T-cell adhesion to endothelium, which may lead to T-cell margination and sequestration and, thus, a decrease of functional T-cells 55. CD4 cells secrete cytokines that activate CD8 cells, B-cells and macrophages. In acute malaria, a depression of CD4 cells leads to depressed cellular and humoral immunity, impairing vaccine response.

The absence of association between malaria chemoprophylaxis and vaccine response in this study is consistent with findings from other chemoprophylaxis studies in malarious areas involving children with asymptomatic parasitaemia 451416171858. No prior studies have published data on the association between chemoprophylaxis and pertussis (killed, whole-cell) vaccine response in humans.

The agglutination test remains the test of choice for whole-cell pertussis vaccines. Although the antigen-specific ELISA tests can amplify the information provided by the agglutination test, the agglutination test is the only one that has been shown clinically to correlate with vaccine efficacy of whole-cell pertussis vaccines and has been used in relatively recent trials 5960. Although ELISA or cell-culture based methods are more widely used today than the passive haemagglutination method for tetanus and diphtheria antitoxins, passive haemagglutination remains acceptable for evaluations of immunized populations 61. Had they been available, the newer serological tests for measles, including neutralization testing, would have provided greater insight regarding clinical protection from disease.

Results of this study indicated that for the pertussis component, children <24 months of age had a better seroresponse. Vaccinating children <24 months of age will more effectively target the population in greatest need. Pertussis is most serious for very young infants and because complications leading to hospitalization, pneumonias, and death occur most often in those <24 months of age, the recommended age for initiation of pertussis immunization is generally two to three months.

Three doses of DTP vaccine comprise the usual primary series;thus, it would have been useful to assess seroconversion after a third DTP dose in addition to the response following the second dose. Technical and logistical considerations precluded this; additionally, there was concern regarding the possibility of decreased compliance with a longer study, as well as the potential to minimize any differences in the effect of chemoprophylaxis on seroconversion.

Malaria serologies demonstrated a significant difference in GMTs between the two treatment groups at the time of vaccination, adding some assurance that chemoprophylaxis decreased asymptomatic parasitaemia. Despite assumed effective chemoprophylaxis for five to seven months, virtually all compliant children had malaria antibodies; this probably reflected a durable antibody response to infections acquired before the study began. While not the primary study objective, fever prevalence data and blood smear records would have provided valuable insight on malaria prevention in the chemoprophylaxis group. Given efforts to administer intermittent preventive therapy of infants (IPTi) in conjunction with the vaccines included in the EPI, additional prospective studies are needed to establish more firmly the effect of antimalarials on response to childhood vaccinations 6.

Malaria chemoprophylaxis does not appear to enhance nor impair the antibody response to DTP and measles vaccines. There have been several changes over the 30 years since the study completion, including development of falciparum malaria resistance to 4-aminoquinolones throughout Africa, and establishment of the EPI (1977) and the Roll Back Malaria Partnership (1998). The continuing development and deployment of new antimalarial drugs and childhood vaccines mandates that the possible immunologic and protective interrelationships of these new products be investigated. Studies are in progress by the IPTi Consortium to address these issues http://www.ipti-malaria.org 7.

Dr. Breman was responsible for writing the protocol, carrying out the study in the field, data analysis and writing; Dr. Rosen for synthesis, data analysis and writing; Dr. Manclark for coordinating DTP serologies and doing the pertussis antibody tests; Dr. Meade for editing, assisting with analysis and interpretation of serologic data; Dr. Collins for supervising the malaria antibody testing; Dr. Lobel for assisting with the protocol, field work and expedition of the serologic analyses; Dr. Saliou for participating in field work and manuscript analysis; Ms. Roberts for data registration and preliminary analysis; Dr. Campaoré for serving as the responsible health officer in Burkina Faso; and Dr. Miller for analysis and interpretation of results.

The study was supported partly by a grant from the Malaria Unit at the World Health Organization. Amodiaquine was contributed by Warner Lambert Pharmaceuticals.