Kinetics of Decline of Maternal Measles Virus-Neutralizing Antibodies in Sera of Infants in France in 2006

The optimal age for measles vaccination is an important healthissue, since maternal antibodies may neutralize the vaccineantigen before a specific immune response develops, while delayingvaccination may increase the risk of complicated diseases ininfants. However, measles vaccination impacts the duration ofprotection afforded by transplacental transfer of maternal antibodies:vaccination-induced maternal antibodies disappear faster thandisease-induced antibodies. In order to maintain protectionagainst measles in infants, it is important to monitor the dynamicsof this phenomenon in vaccinated populations. To assess thecurrent situation in France, a multicenter, prospective seroepidemiologicalstudy was conducted in seven French hospitals between October2005 and January 2007. Maternal measles antibody concentrationsfrom 348 infants 0 to 15 months old were measured using theplaque reduction neutralization assay. Geometric mean concentrationsand the percentage of infants with maternal measles antibodyconcentrations above the protection threshold ( $≥$ 120 mIU/ml) wereassessed according to age. Results show that after more than20 years of routine measles vaccination in France, maternalmeasles-neutralizing antibodies decrease dramatically in Frenchinfants by 6 months of age, from 1,740 mIU/ml for infants 0to 1 month old to 223 mIU/ml for infants 5 to 6 months old,and that 90% of infants are not protected against measles after6 months of age. Infant protection against measles could beoptimized both by increasing herd immunity through an increasedvaccine coverage and by lowering the age of routine vaccinationfrom 12 to 9 months.

INTRODUCTION

Top

INTRODUCTION

Large-scale measles vaccination has led to a dramatic decreasein measles incidence and deaths from measles worldwide (54).In France, the estimated number of measles cases dropped from300,000 in 1985 to less than 10,400 cases in 2003 (3, 48), andmortality has declined from 30 deaths per year in the 1980sto less than 10 per year in recent years (3, 48). However, despitethe success in controlling measles, the disease has not beeneliminated in Europe; outbreaks still occur (2, 3, 13, 18, 29,34, 35, 45, 48, 50, 52, 53), and measles can still representa serious health threat, especially in infants under 1 yearof age (7, 18).

The optimal age for infant measles vaccination is an importanthealth issue, since maternal antibodies may neutralize the vaccineantigen before a specific immune response develops, while delayingvaccination may increase the risk of complicated diseases ininfants. However, the introduction of measles vaccine in a countryparticularly impacts the duration of protection afforded bymaternal antibodies: levels of vaccine-induced maternal antibodiesare lower, and they disappear faster than disease-induced antibodies.As vaccine coverage of a population increases, measles viruscirculation declines, and more infants are born with vaccine-inducedmaternal antibodies. These infants will be protected for a shorterperiod of time than those in the prevaccine era. To maintainprotection against measles in infants, it is thus importantto monitor the dynamics of this phenomenon in vaccinated populations.

Although the measles vaccine was licensed in France in 1966,it was not included in the vaccination schedule until 1983,as a bivalent vaccine in combination with rubella, and thenin 1986, as a trivalent measles-mumps-rubella (MMR) vaccinefor infants 12 to 15 months old. In 1996, recommendations wereextended to include a second dose of MMR vaccine for children11 to 13 years old. This second dose does not constitute a booster,with long-term immunity acquired following the first vaccination.It constitutes a catch-up for infants who did not seroconvert,for one of several antigens, with the first vaccination (10).

In 1997, the recommended age range for the second dose was loweredto 3 to 6 years (42). In 2005, within the framework of the WHOcongenital measles and rubella elimination program, the recommendationwas modified again, and the age of vaccination was lowered to12 months for the first dose of MMR vaccine, with the seconddose given between 13 and 24 months of age (11). For childrenattending day care centers or traveling to countries where measlesis highly endemic, the recommendation for the first dose is9 months of age, with the second dose given between 12 and 15months of age. Monovalent measles vaccine can be used in infants6 to 8 months old who have been in contact with measles cases(10, 11, 16). Lowering the recommended age of vaccination wasalso prompted by several reports showing a rapid decline inmaternal measles antibody concentrations in other well-vaccinatedpopulations (12, 24, 32, 40, 44).

In France, measles vaccine coverage increased significantlyonly after the official recommendation in 1983 and the subsequentavailability of the combined MMR vaccine in 1986. Measles vaccinecoverage of infants (with a single dose at 24 months of age)increased from 19% in 1979 to 35% in 1983, 57% in 1987, and80% in 1994 and then remained relatively unchanged until 2004,when it reached 87% (4, 10).

The present study is the first to assess the seroepidemiologyof measles in infants in France, a country where the populationis well vaccinated against measles. The main objective of thisstudy was to determine the kinetics of the decline of passivelytransferred maternal measles antibodies in infants 0 to 15 monthsold. The second objective was to define the proportion of infantsprotected against measles (with neutralizing serum antibodyconcentrations of $≥$ 120 mIU/ml).

MATERIALS AND METHODS

Top

MATERIALS AND METHODS

Study design. This prospective multicenter study was conducted between 6 October2005 and 31 January 2007 in seven hospitals located in six regionsthroughout France, as follows: Normandie, Bretagne, Pays deLoire, the greater Paris area, Rhônes-Alpes, and Aquitaine.

Study population. Seven hospitals consecutively recruited infants, who consultedor were hospitalized in a pediatric department and/or a pediatricemergency service, who met the following inclusion criteria:between 0 and 15 months of age, born at term (mother with $≥$ 37weeks amenorrhea), with a birth weight of $≥$ 2.8 kg and a medicalcondition requiring blood sampling, and with informed consentfor the present study given by at least one parent.

Infants were excluded from the study if they met any of thefollowing critieria: previous vaccination against measles ora clinical history of measles, contact with a measles case inthe previous 3 weeks, a known or suspected immunodeficiency,and previous treatment with immunoglobulins or a blood transfusion.Infants were also excluded if their mother received a bloodtransfusion during pregnancy or had been living in metropolitanFrance for less than 3 years. In addition, if collection ofan additional 0.5 ml of blood posed a medical risk or if aninfant was enrolled in any other study, the infant was excludedfrom this study.

The study was performed in accordance with the Helsinki Declaration,the good epidemiology practice guidelines of the Associationof French-speaking Epidemiologists, and the procedures establishedby French law. It was approved by the Consultative Committeefor the Protection of Persons Participating in Biomedical Researchof Saint-Germain-en-Laye. Informed consent was given by at leastone of the parents of all infants participating in the study.Children enrolled in the study were listed in the National Fileof Persons Participating in Biomedical Research without DirectIndividual Benefit.

Sampling. For each infant, a 0.5-ml blood sample was collected in a drytube for determination of measles antibody concentration inaddition to the volume collected for the infant's medical management.

The additional 0.5-ml blood sample was processed by the localhospital laboratory. Blood was centrifuged at 3,000 rpm for10 to 15 min, and serum was separated, placed in a tube labeledwith the patient's identification code (tag), and stored at–20°C until the end of the study. Coded serum sampleswere shipped at the end of the study to the Centre for Infections,Health Protection Agency, London, United Kingdom, for analysis.

PRN assay to determine measles-neutralizing serum antibodies. Measles antibodies were measured by a plaque reduction neutralization(PRN) assay as described previously (8). In brief, each serumdilution was mixed with an equal volume of challenge virus (Edmonstonwild-type strain) containing a standard virus dose. After incubation,to allow virus neutralization to proceed, the virus-serum mixtureswere inoculated onto Vero cells and incubated for an additional5 to 7 days. The highest dilution of serum giving a 50% reductionin plaque count was defined as the endpoint concentration. Resultswere transformed into mIU/ml by comparing sample endpoint concentrationswith the concentration of the second international standardfor measles antibody containing 5,000 mIU/ml (National Instituteof Biological Standards and Control, South Mimms, United Kingdom),which was tested in parallel with each batch of test samples.

The seroconversion threshold was defined as the detection limitof the technique. The detection limit varied slightly betweenassay runs, ranging from 27 to 63 (average 45.3) mIU/ml in thenine assay runs performed to test the samples from this study.The protection threshold was defined as a concentration of $≥$ 120mIU/ml (6). Antibody concentrations were reported in mIU/ml,and the geometric mean concentration (GMC), extreme values (minimumand maximum range), and 95% confidence intervals (CI) were calculated.Only antibody concentrations over the limit of detection weretaken into account for the calculation of GMCs.

Data collection. A case report form was compiled for each participant and includedidentification of the infant and investigator; the place anddate of inclusion; the date of birth, gestational age, birthweight, and sex of the infant; and the date of and reason forblood sampling for the infant's medical management. Informationabout the mother, including her date of birth and her measleshistory (natural infection or vaccination), was obtained duringthe interview with the parents and was noted in the case reportform.

Statistical analysis. Quantitative variables were analyzed using a two-sided Studentt test or analysis of variance. Qualitative variables were studiedusing a two-sided chi-square test or a Cochran-Mantel-Haenszeltest. Data analysis was conducted using SAS version 8. A P valuebelow 0.05 was considered statistically significant.

The evaluable population was defined as the infant populationwith no major deviation from the protocol. Factors such as ahistory of vaccination or the absence of the information formor the data for gestational age or a blood sample were consideredmajor deviations. Infants with a birth weight between 2.6 and2.8 kg were taken into account in the evaluation.

Measles antibody concentrations were analyzed by 1-month agegroups. For statistical tests, 3-month age groups were usedif a sufficient sample size could not be obtained for each 1-monthage group. The relationship between measles antibody concentrationsand age was assessed using exponential nonlinear regressionanalysis weighted by age.

RESULTS

Top

RESULTS

A total of 353 infants were enrolled in the study. Five infantswere excluded due to previous measles vaccination (n = 2) ormissing data or blood samples (n = 3). Of the 348 evaluableinfants, 51.1% were males and 48.9% were females. Their mediangestational age at birth was 39 weeks (mean, 39.3 ± 0.2weeks), and the median birth weight was 3.330 kg (mean, 3.392± 0.424 kg; minimum to maximum, 2.610 kg to 5.050 kg).The median age at inclusion was 8.7 months (mean, 7.9 ±3.8 months; minimum to maximum, 0.03 to 15.57 months). The distributionaccording to age was as follows: 0 to 3 months, 13.8% (n = 48);3 to 6 months, 12.9% (n = 45); 6 to 9 months, 27.0% (n = 94);9 to 12 months, 33.0% (n = 115); and >12 months, 13.2% (n= 46).

The age of the mother was known for 347/348 infants. At delivery,the median age of the mothers was 29 years (mean, 28.9 years± 5.05 years; range, 17 to 45 years). Among the 348 mothers,113 (32.5%) said they had been vaccinated, and 141 (40.5%) reporteda history of natural infection.

Measles-neutralizing serum antibody concentrations in infants. A total of 40.2% of infants (140/348) had measles antibody concentrationsabove the detection limit. The proportion of infants with detectablemeasles antibodies decreased dramatically with age, from 95.8%in infants 0 to 3 months old to 39.4% in infants 6 to 9 monthsold and 10.9% in infants over 12 months.

The GMCs of measles antibodies for each 1-month age group weredetermined for the 140 serum samples with antibody concentrationsabove the detection level. Measles-neutralizing antibody GMCsdecreased rapidly from 1,740 mIU/ml for infants in the 0- to1-month age group to 223 mIU/ml (5- to 6-month age group) to65 mIU/ml (6- to 7-month age group). Among the 13- to 14-monthand the 14- to 15-month age groups, 11/12 and 6/6 children,respectively, had measles antibody concentrations under thelimit of detection. Consequently, these concentrations werenot taken into account for the calculation of GMCs (Fig. 1).When results for infants in the different 3-month age groupswere compared (Table 1), a statistically significant difference(P < 0.001) in GMCs was observed.

View larger version (16K):
[in this window]
[in a new window]

FIG. 1. Distribution of measles-neutralizing serum antibody concentrations (GMC) in infants 0 to 15 months old (1-month age groups) in France, 2006.

View this table:
[in this window]
[in a new window]

TABLE 1. Measles-neutralizing serum antibody concentrations by 3-month age groups in France in 2006

Proportion of infants with measles-neutralizing serum antibody concentrations at or above the seroprotection threshold ( $≥$ 120 mIU/ml). The proportion of infants with protective measles antibody concentrationsat or above 120 mIU/ml dropped from 100% in infants 0 to 1 monthold to 10% in infants 6 to 7 months old (Fig. 2). A statisticallysignificant difference in the proportion of infants with protectivemeasles antibody concentrations was observed between the 3-monthage groups (P < 0.001) (data not shown).

View larger version (17K):
[in this window]
[in a new window]

FIG. 2. Proportion of infants 0 to 15 months old (1-month age groups) with protective measles antibody concentrations according to age in France, 2006.

Measles-neutralizing antibody concentrations in infants according to the mother's year of birth. The GMC of measles-neutralizing antibodies was lower in infantswhose mothers were born after 1983 than in infants whose motherswere born before 1983 (128.6 mIU/ml versus 212.1 mIU/ml, respectively)(Table 2), but this difference was not statistically significant(P = 0.43).

View this table:
[in this window]
[in a new window]

TABLE 2. Measles-neutralizing serum antibody concentrations in infants in 2006 according to their mothers’ year of birth in France

The proportion of infants with protective measles antibody concentrationswas also lower in infants whose mothers were born after 1983than in infants whose mothers were born before 1983 (9.1% versus22.2%, respectively). This difference, however, was not statisticallysignificant (P = 0.15).

Measles-neutralizing antibody concentrations in infants according to the mother's reported history of measles (natural infection versus vaccination). Infants 0 to 3 months old whose mothers had a reported historyof measles vaccination had lower antibody concentrations (GMC,155.0 mIU/ml; 95% CI, 95.6 to 251.5) than infants whose mothershad a reported history of measles infection (GMC, 263.8 mIU/ml;95% CI, 172.8 to 402.8), but the difference was not statisticallysignificant (P = 0.11) and disappeared rapidly (Fig. 3). Forinfants 6 to 9 months of age, the level of maternal measlesantibodies was the same for all infants, whether their mothershad had natural measles infections or had been vaccinated.

View larger version (14K):
[in this window]
[in a new window]

FIG. 3. Measles-neutralizing serum antibody concentration (GMC) according to infant age (3-month age groups) and maternal measles status (natural infection or vaccination) in France, 2006.

DISCUSSION

Top

DISCUSSION

Early measles infection may lead to severe complications suchas bacterial superinfection and subacute sclerosing panencephalitis;the highest hospitalization rates for measles are observed forinfants under 1 year of age (7, 18, 33). Newborns are protectedagainst measles by maternal antibodies actively transferredvia the placenta during the last trimester of pregnancy (41,46). These maternal antibodies disappear during the first monthsof life as the newborn's own immune system develops.

The optimal age for measles vaccination is an important publichealth issue. Vaccination at an early age may fail because ofimmaturity of the immune system and also because of the presenceof maternal antibodies, which neutralize the vaccine antigensbefore a specific immune response develops (19, 25, 36). However,any delay in vaccination may increase the risk of complicateddisease in infants. It is consequently important for each countryto adapt the timing of vaccination to its measles seroepidemiologicalsituation (27). This study is, to our knowledge, the first todetermine the kinetics of maternal measles antibody declinein infants in France. It involved seven hospitals located insix regions throughout France: Normandie, Bretagne, Pays deLoire, greater Paris, Rhônes-Alpes, and Aquitaine.

A rapid decline in levels of maternal measles-neutralizing antibodiesafter birth was observed, as measured by the PRN assay. Thistest is currently considered the gold standard due to its highsensitivity and ability to measure functional neutralizing antibodies(6, 9, 51). Although the percentage of infants with detectablemeasles-neutralizing antibody levels remained around 40% forinfants 7 to 8 months old, 90% of the infants included in thestudy had antibody concentrations under 120 mIU/ml after 6 monthsof age. Studies carried out in other countries have also shownthat infants over 6 months of age in a well-immunized populationmay be poorly protected against measles (5, 12, 24, 32, 40,44).

The 120-mIU/ml antibody level correlates with protection inactively immunized (vaccinated) children; it may not correlatewith protection in passively immunized children because cell-mediatedimmunity, which also plays an important role in protection,is lacking in passive immunity. Thus, while some children vaccinatedagainst measles could be protected even in the absence of detectablecirculating antibodies through cell-mediated immunity (47),infants passively immunized through maternal antibodies maynot be protected against disease, even with antibody levelsof >120 mIU/ml.

The rate of decline of maternal antibodies in infants is dependentupon initial levels of maternal antibody, which reflect bothmaternal antibody levels at the time of pregnancy and the extentof placental transfer. Since antibody transfer occurs mainlyduring the third trimester of pregnancy, preterm infants havelow levels of maternal antibodies (5, 21, 26, 28, 31, 41, 46).Placental transfer has also been found to be less efficientin infants in developing countries than in industrialized countries(22). Coexisting infections such as malaria and human immunodeficiencyvirus infections are also associated with decreased levels ofmaternal antibodies in infants (5, 17). These factors shouldnot have complicated the results of the present study, sincepremature infants and infants whose mothers were not livingin metropolitan France at the time of pregnancy were excludedfrom the study.

It is recognized that measles antibody levels induced by vaccinationare lower than antibody levels induced by the disease (23, 24).Several studies have shown that newborns from vaccinated mothershave both lower levels of transplacentally transferred measlesantibodies than infants born from women who had measles andan earlier decline in maternal antibodies (1, 4, 12, 15, 30,37, 38, 40). In this study, the measles antibody levels determinedfor infants at 0 to 3 months of age were higher in infants bornfrom mothers who had had natural measles infections than ininfants of vaccinated mothers. This difference was not significantand disappeared rapidly (Fig. 3). However, one limitation tothe study is that the mothers’ histories were obtainedonly through interview, without any confirmation using individualhealth records or serology. The same trend was observed whenantibody levels in infants were analyzed according to the mother'syear of birth; for infants whose mothers were born before 1983(measles vaccination was recommended in 1983), there was a nonsignificanttrend toward higher antibody levels. This result should be interpretedwith caution, taking into account the small sample size of infantswith mothers born after 1983.

A seroepidemiological study carried out in France in 2005 to2006 among women of childbearing age showed that the measles-neutralizingserum antibody GMCs were significantly lower in women born after1983 and decreased significantly with increasing birth year,corresponding to increasing vaccine coverage (43). This decreasein antibody levels in women of childbearing age may be amplifiedwith the increasing age at which the majority of women havechildren in the Western population, due to the increased intervalbetween childhood vaccination and childbirth (27). In addition,boosting by wild-type viruses occurs less often as vaccinationcoverage increases, and this may contribute further to lowerantibody levels in the adult population (either after vaccinationor after natural disease) (39).

Conclusion. This study is the first to show that in France, passively transferredmaternal antibodies against measles decline rapidly during theinfant's first months of life, to reach mean concentrationsunder the seroprotection level at 6 months of age. Infants over6 months are poorly protected by maternal antibodies againstmeasles (90% of infants are no longer protected against measlesat 6 months of age), which is consistent with observations inother countries, where populations are vaccinated routinelyagainst measles. The progressive increase in measles vaccinationcoverage in infants born in France after 1983 led to a significantdecrease in measles incidence and consequently a loss of naturalboosters from wild-type measles virus. On the other hand, thecurrent vaccination coverage (87% for one dose at 24 monthsof age in 2004) is not sufficient to eliminate the disease inFrance, and measles outbreaks are still occurring (10, 14).This places infants older than 6 months who have not yet beenvaccinated in a critical window at risk for measles infectionand potentially serious complications such as subacute sclerosingpanencephalitis. While lowering the age for the first dose ofmeasles vaccine to 6 months may not be optimal even in the absenceof maternal antibody, because of the immaturity of the immunesystem (19, 20, 25, 36), lowering the age of routine vaccinationfrom 12 to 9 months of age would reduce the immunity gap betweenpassive and active protection.

High vaccine coverage is the most important factor needed tointerrupt and control measles transmission (14). Efforts arealso required to reach a vaccine coverage rate of at least 95%with two doses, estimated as necessary for measles elimination(3, 49).

Infant protection against measles could be optimized both byincreasing herd immunity through an increased vaccine coverageand by lowering the age of routine vaccination from 12 to 9months.

ACKNOWLEDGMENTS

This work was supported by Sanofi Pasteur MSD.

We thank the coinvestigators of the study. Coinvestigators includeI. Abadie, L. Abalea, F. Audic-Gérard, C. Bailly-Botuha,F. Babre, S. Berberian, L. Berthomieu, C. Boscher, V. Brossard,I. Ceruti-Hazart, N. Delaperrière, N. Fargier, H. Feghaki,E. Fleurence, C. Fourmaux-Poulain, S. Gaubicher, C. Gay, N.Godon, M. Grall-Lerosey, L. Kohen-Couderc, E. Lachassine, D.Mamireau, C. Lardennois, P. Lemoine, F. Madhi, C. Metz, B. Peyret,N. Remus, O. Richer, C. Roumegoux, E. Sabbagh-Helali, J. Sarlangue,L. Sarthou, J.-F. Segura, A. Seiz, and L. Tripodi. We also thankthe personnel of the seven French hospital laboratories whoparticipated in the study: S. Capdepont, M. Gueudin, P. Ledudal,V. Narbonne, S. Pillet, I. Poilane, C. Robin, and P. Volle.In addition, we thank Desiree Doblas of the HPA Centre for Infectionsfor performing measles PRN assays; Christel Saussier and RemiGauchoux (Mapi-Naxis) for data analysis; and Betty Dodet (DBS,Lyon, France) for help in writing the manuscript.

FOOTNOTES

* Corresponding author. Present address: Département de Pédiatrie, Unité de Néonatalogie, CHUS Fleurimont 3001, 12ème Avenue Nord, Sherbrooke, QC J1H 5N4, Canada. Phone: (819) 346-1110. Fax: (819) 564-5398. E-mail: arnaud.gagneur{at}usherbrooke.ca

Published ahead of print on 24 September 2008.

REFERENCES

Top