Measles susceptibility in maternal-infant dyads—Bamako, Mali

Measles is endemic in Africa; measles mortality is highest among infants. Infant measles antibody titer at birth is related to maternal immune status. Older mothers are likelier to have had measles infection, which provides higher antibody titers than vaccine-induced immunity. We investigated the relationship between maternal age and measles susceptibility in mother-infant pairs in Mali through six months of infancy. We measured serum measles antibodies in 340 mother-infant pairs by plaque reduction neutralization test (PRNT) and calculated the proportion of mothers with protective titers (>120 mIU/mL) at delivery and the proportion of infants with protective titers at birth, and at three and six months of age. We explored associations between maternal age and measles antibodies in mothers and infants at the timepoints noted. Ten percent of Malian newborns were susceptible to measles; by six months nearly all were. Maternal and infant antibody titers were highly correlated. At delivery, 11% of mothers and 10% of newborns were susceptible to measles. By three and six months, infant susceptibility increased to 72% and 98%, respectively. Infants born to younger mothers were most susceptible at birth and three months. Time to susceptibility was 6.6 weeks in infants born to mothers with measles titer >120–<430 mIU/mL versus 15.4 weeks when mothers had titers ≥430 mIU/mL. Maternal and newborn seroprotective status were positively correlated. Improved strategies are needed to protect susceptible infants from measles infection and death. Increasing measles immunization coverage in vaccine eligible populations, including nonimmune reproductive-aged women and older children should be considered. For the original influenza vaccination trial, 4,425 mothers were recruited in the third trimester of pregnancy from referral and community health centers in urban Bamako, Mali, between September 2011 and April 2013; 4,193 of them were randomly assigned (1:1) to study intervention (trivalent inactivated influenza or quadrivalent meningococcal vaccine). Exclusion criteria, enrollment, and additional study methods have been described in detail elsewhere [27]. This study, including the measles antibody analysis in stored specimens, was approved by the Institutional Review Board of the University of Maryland School of Medicine; the ethics committee of the Faculté de Médecine, Pharmacie et Odonto-Stomatologie of Mali; the Ministry of Health of Mali; and the US Centers for Disease Control and Prevention. Mothers provided informed consent to blood draw for themselves and for their infants and for specimen storage and future use. Serum samples obtained at the local health care centers were transported to the Center for Vaccine Development in Bamako, Mali, then shipped to University of Maryland, Baltimore, USA, and stored at −80 °C. Information on study population for the parent influenza study had been collected via hand completed scannable forms (TeleForm, version 10.5.1; Autonomy Cooperation PLC) and populated a Microsoft Access database. A subsample of eligible mother-infant pairs (n = 340) was randomly selected from the maternal influenza study to investigate seroprotection status (circulating measles antibody levels) as a risk factor for infant susceptibility to measles. Mother-infant pairs were eligible for measles antibody analysis if the mother was 15–39 years old at the time of (singleton) birth and had completed all three scheduled visits at birth, three months, and six months. Mothers in the original study were stratified into five age groups: 15–19, 20–24, 25–29, 30–34, and 35–39 years. An equal number of mother-infant pairs (n = 68) in each stratum achieved 90% power to detect a linear trend in the proportion of infants with protective maternal measles titer across the five age groups using a Cochran-Armitage test and a significance level of 0.05. The sample size was estimated using PASS v 14.09 (NCSS, Kaysville, Utah, USA) assuming equally spaced proportions from 0.70 to 0.90. Measles plaque reduction neutralization tests (PRNT) were performed, as previously described, with modifications [28]. Heat inactivated (30 min at 56 °C), serially diluted serum samples were incubated with an equal volume containing 30 plaque-forming units (pfu) of measles virus (Edmonston strain, ATCC, Manassas, VA) for 1.5 h at 37 °C, 5% CO2. The serum-virus mix was transferred onto ∼90% confluent Vero cell monolayers (ATCC) seeded in 24-well plates (in duplicate) and incubated for 1 h. After incubation the liquid was removed; cells were overlaid with 1% carboxymethylcellulose (1 mL, Sigma-Aldrich, St. Louis, MO) and incubated for 5 days at 37 °C, 5% CO2. On day 5, 0.5 mL of neutral red stain (Sigma-Aldrich, St. Louis, MO) diluted in Temin’s Modified Eagles Medium supplemented with 4% fetal bovine serum (Thermo Scientific, Waltham, MA) was added and incubated overnight at 37 °C, 5% CO2. The stained overlay was aspirated, fixed with 0.5 mL of 10% formalin solution (Millipore, Darmstadt, Germany) and allowed to dry overnight (covered). Each well was scanned using the Immunospot S6 Macro Analyzer. Cytopathic plaques were counted with Biospot Pro version 7.0 software (Cellular Technology Limited, Cleveland, OH). PRN titers were calculated using the Reed-Muench method as the inverse of the dilution that inhibited 50% plaques compared with no serum control and reported in mIU/mL based on the 3rd International WHO (97/648) reference standard (NIBSC, Hertfordshire, UK). Date of birth or enrollment age was used to calculate maternal year of birth. Measles cases and MCV delivered via RI and SIA in Mali by year were plotted against maternal year of birth as a proxy for measles exposures [20]. PRNT > 120 mIU/mL were considered protective [29]. Maternal measles immune status was further classified into three categories, based on median titer at delivery: seronegative (≤120 mIU/mL), moderately positive (>120–430 mIU/mL), and strongly positive (>430 mIU/ mL). The Cochran-Armitage test for trend assessed whether the proportion of infants protected against measles at birth increased linearly with mother’s age. The Kruskal-Wallis test tested for differences in antibody titers among the five maternal age groups at birth (mothers and infants), three months, and six months. A loess smoother with span value = 0.7 was used to graph the relationship between maternal and infant titers. In an exploratory analysis, a linear mixed effect model was fit to the data to estimate the waning of infant measles titer over time. The model included a random intercept, time (weeks), the maternal measles immune status (seronegative, moderately positive, strongly positive), and the interaction between time and strength of response. The population average time, in weeks, to susceptibility was calculated from the estimated average decay of infants of mothers with moderately positive and strongly positive immune status. Analyses were conducted in SAS (version 9; SAS Institute) and R software (version 4.0.2; R Foundation); p-values ≤0.05 were considered statistically significant.