Malaria chemoprophylaxis and the serologic response to measles and diphtheria-tetanus-whole-cell pertussis vaccines

The study was conducted from May through December in 1975 in six villages; all were located in the Guinean savanna and were hyper- and holo-endemic for malaria, depending on transmission season [25]. Before the study began (February-March, during the low transmission season), a 52% Plasmodium falciparum parasitaemia prevalence was found in 150 children (25 per site) <6 years of age, with no major differences between the sites; during this pre-study investigation, antibodies to P. falciparum were detected by indirect haemagglutination (IHA) in 100 percent of children tested from five of the six villages (25 children per village). Burkinabe clinicians in the nearest dispensaries and hospitals stated that the study area was endemic for measles (cases and deaths occurred during the study), diphtheria, tetanus, and pertussis, but the incidence was unknown; routine data had not been collected from the study villages because the EPI had not yet begun [26]. Hence, previous vaccination of children was extremely unlikely and was confirmed by interrogation of individual families. There was no malaria prophylaxis; treatment for fevers and other illness was obtained from traditional healers and in dispensaries within five km of the villages. P. falciparum resistance to 4-aminoquinolones was unknown in the area. Exclusion criteria for participation in the study included acute or chronic severe illness and the presence of detectable pre-vaccination measles antibodies. Children were weighed with a calibrated hanging scale. The length of very young children was measured with them lying down on a calibrated board and, for older children, standing, according to Centers for Disease Control and Prevention (CDC) nutrition programme guidelines.

Children, aged 4 to 74 months (N = 996) living in the 6 malarious villages were assigned to 2 groups of 3 villages each, depending on whether they lived in villages east or west of Bobo-Dioulasso. One group received amodiaquine prophylaxis, CH+ [N = 488], and the other received no prophylaxis, CH- [N = 508]. Virtually all children within the same village received either measles [N = 482], or DTP vaccine [N = 514] (Figure 1). The study villages had differing populations of target-aged children. Hence, village and subject separation into measles versus DTP groups was based on the estimated number of children needed for each vaccine, and was dependent on the initial calculation of sample size (see Statistical Analysis). There was no blinding of study participants or researchers.

CH+ children received a single prophylactic dose of amodiaquine hydrochloride suspension or tablets every 2 weeks for 7 months beginning in May and June, the start of the transmission season; those <12 months of age received 100 mg, those 12 to 47 months received 200 mg, and those 48 to 73 months old received 300 mg. Both amodiaquine and chloroquine are 4-aminoquinolines. Amodiaquine is the prodrug for the active ingredient desethylamodiaquine (DAQ). DAQ has a terminal half life of one to three weeks and is schizonticidal in very low concentrations [27, 28]. While amodiaquine lost favour because of its association with agranulocytosis, the drug is now being re-evaluated. Because amodiaquine has not been used for over 2 decades, it has somewhat greater efficacy than chloroquine for P. falciparum resistant to chloroquine [29, 30].

Seroconversion rates were measured in all CH+ and CH- children for whom paired sera were available. In addition, a CH+ group having strict compliance to drug ingestion was analyzed to examine more carefully the effect of malaria or drug use on serconversion and seroresponse. For this study, criteria for strict compliance included: receipt of ≥75% of the doses, no two consecutive doses missing, no doses missed in the month prior to vaccination(s), and no doses missed between the second vaccination and the last blood draw. At study completion, during the beginning of the low transmission season, families of children in the CH+ group were given a short-term supply of amodiaquine prophylaxis and instructions for home treatment in case of a rebound malaria attack. Follow-up visits occurred every one to two months in all villages for six months after the study.

All children were vaccinated with either measles vaccine or the first dose of DTP at month 5 (October or November, peak malaria transmission) and the second DTP dose one month later. Two doses of DTP were given during the study (rather than the standard initial series of three doses) to increase the likelihood that any effect from chemoprophylaxis would be discerned; a third dose was given at the end of the study but bloods were not drawn after this third vaccination. Vaccinations and amodiaquine were administered on the same day. The manufacturer and source of the licensed vaccines used were the Dow, Lirugen (Schwarz strain) measles vaccine, Lot No. 185806 AA and Merrill-National DTP, adsorbed, USP, vaccine, filling number 1036 DM, bulk Lot No. 1832. Measles and DTP vaccines were injected intramuscularly (0.5 ml) via a hydraulic (pressurized), injection device, the Ped-O-Jet^® injector. Following study completion, all children in participating villages received the vaccine that they did not receive during the study. Prior to use, the measles and DTP vaccines were tested for potency and met standard requirements.

Venous blood samples were kept refrigerated and within 1 to 3 days serum was separated and kept at -20°C prior to shipment to the CDC, Atlanta, Georgia, USA, in dry ice. Serologic testing for the measles vaccine was performed in 1977 at the then Virology Division, Bureau of Laboratories, CDC. Antibody response to the DTP vaccine was performed in 1977 at the then Division of Bacterial Products, Bureau of Biologics, Food and Drug Administration, in Bethesda, MD.

Measles antibody was assessed by haemagglutination inhibition [31‐33]. Seroconversion was defined by a rise in titer to ≥1:20 from an initial titer of <1:10 (the lowest detectable titer). Diphtheria and tetanus antibody titers were measured by passive microhemagglutination, using tanned sheep red blood cells [34, 35]. Individuals showing a >4-fold rise in antibodies to uncoated sheep red blood cells were not included in the analysis. Pertussis titers were measured by microagglutination using killed cells of Bordetella pertussis strains 134 and 165 [36]. The titer is the log₂ of the reciprocal of the highest final serum dilution resulting in detectable agglutination. When sufficient serum was available, the lowest final serum dilution tested was 1:8; by convention, samples negative at a 1:8 dilution were assigned the titer log₂ = 2. If the serum sample volume was low, higher initial dilutions were used. When such sera were negative at the lowest dilution tested, the titer was reported as <lowest dilution tested. Because the actual end-point was not known, titers for these sera were not included in the calculation of geometric mean titer (GMT) or geometric mean fold rise in titer (GMR). For some individuals, it was possible to verify a ≥4-fold rise even if the actual endpoint was not known for both sera. A ≥4-fold rise for DTP antigens was considered positive. For DTP, in cases where the titer decreased from pre- to post-vaccination, a ΔGMT = 0 was used in the calculation of GMR.

P. falciparum IgG antibodies were measured by the Immunofluorescence Assay (IFA) [37] and by the Enzyme-Linked Immunosorbent Assay (ELISA) [38] following collection of blood 7 months after the children were CH+ or CH- status.

Sample size was determined initially by a method comparing two proportions. For measles, 215 subjects were required for each group in order to have 90% assurance of significant results to detect this 10% difference in response rates. Similarly, for DTP vaccines, assuming 70% seroconversion for the test group and 60% for the control group, 387 subjects were needed for each group. SAS software, version 9.00 (SAS Institute, USA), was used. Weight-for-height Z-score was calculated using an anthro-system (version 1.02, WHO-CDC, Switzerland).

The primary outcome was rate of seroconversion or seroresponse in CH+ and CH- individuals. As secondary outcomes, geometric mean titers (GMT) and mean fold rise in titer (GMR) were measured for measles, DTP, and malaria antibodies for CH+ and CH- individuals for the strict compliance group.

Study population characteristics at vaccination were compared for the CH+ and CH- children using the Chi-squared test and Student's pooled t-test. The Chi-squared test was used for comparison of seroconversion to measles vaccine and seroresponse to DTP vaccinations; the Student's pooled t-test was used for pre- and post-vaccination GMT and GMR. A univariate logistic analysis was performed to assess effects of sex, age (> or < 24 months), prophylaxis, weight-for-height Z-score (> or < median Z-score), and pre-vaccination GMT (> or < median GMT) on seroconversion or seroresponse. Multivariate logistic regression was performed on those factors found to be independently associated with seroresponse. Analyses for seroconversion, GMTs, GMRs, and logistic regression were adjusted for village effect.

The study protocol was approved by the Burkina Faso (Upper Volta) Ministry of Health and the Institutional Review Board at the CDC. Verbal permission for the study was obtained from the village chiefs, their "council of elders," and each participating family, as was the custom for working in Burkinabe villages.