Response to third rubella vaccine dose

ABSTRACT

Limited data exist on the immunogenicity of a third dose of the measles, mumps, and rubella vaccine (MMR). In this study, our aim was to evaluate the long-term rubella immunogenicity afforded by two childhood MMR doses of the Norwegian vaccination program in a cohort of conscripts and to determine the effect of an additional dose of MMR vaccine, in order to inform vaccination policy. Blood samples from Norwegian conscripts (n = 495) taken both before and eight months after administration of a dose of MMR vaccine were tested using an enzyme immunoassay to measure anti-rubella IgG. Concentrations <5 IU/mL were regarded as negative, 5.0-9.9 IU/mL as equivocal, and ≥10 IU/mL as positive. Overall, the seropositivity before vaccination was 84.6%, and 99.0% of the conscripts had anti-rubella IgG concentrations ≥5 IU/mL. The seropositivity after vaccination was 94.5%, and 99.8% of the conscripts had antibody concentrations ≥5 IU/mL. The geometrical mean IgG concentrations increased from 21.4 IU/mL before vaccination to 28.9 IU/mL after. Four out of five conscripts, with seronegative concentrations before administrations of an additional MMR dose, had equivocal or seropositive results following vaccination. The cohort of young adults in Norway, which was eligible for two childhood MMR doses, was protected against rubella, and efforts should be made to maintain high vaccine coverage to ensure immunity in the future. A third dose of MMR administered in early adulthood led to an increase in the antibody concentration in our cohort and seroconversion for the majority of seronegative persons.

KEYWORDS:

• MMR
• rubella
• seroprevalence
• vaccination

Introduction

Rubella is an acute viral disease transmitted through airborne droplets when infected people sneeze or cough. A rubella infection in early pregnancy or just prior to conception may result in miscarriage, fetal death, or congenital rubella syndrome (CRS). Children should therefore receive measles mumps rubella (MMR) vaccine and non-immune pregnant women should be offered the MMR vaccine after giving birth to prevent complications in subsequent pregnancies. In the World Health Organization (WHO) Global Vaccine Action Plan, rubella is targeted for elimination in five WHO Regions by 2020.¹ Globally, large-scale rubella vaccination has dramatically reduced rubella occurrence in the last decade. The most commonly used vaccines are rubella-containing vaccines that also confer protection against other viral diseases such as measles, mumps, and varicella.²

In Norway, two doses of MMR vaccine have been included in the Childhood Immunization Program since 1983. Prior to that, one dose of the rubella vaccine was offered to 15-year-old girls since 1978. For more than a decade, fewer than five cases of rubella have been reported annually to the Norwegian Surveillance System for Communicable Diseases and the latest report of congenital rubella syndrome was in 2002.³ According to the 2016 evaluation from the European Regional Verification Committee for Measles and Rubella elimination, rubella has been eliminated in Norway.⁴

Seroprevalence studies are important to monitor the effect of vaccination programs and assess possible immunity gaps. The vaccine response mounted by the immune system to MMR is usually good, but the long-term immunity afforded by two doses of MMR is uncertain, because vaccine-induced immunity may wane more rapidly than immunity conferred through natural infection that is boosted by rubella circulation.⁵ The recommended age for the second dose of MMR varies between countries from under 2 years to 15 years depending on the national vaccination program.⁶ This may also be reflected in observed antibody levels in adulthood. Previous studies have indicated that adolescent and young adults aged 15–30 years have lower levels of anti-rubella antibodies than other age groups and may require booster vaccinations.⁷ To our knowledge, few data exist on the immunogenicity of a third MMR vaccine dose on immunity to rubella, following a two dose childhood vaccination program.

Our study addresses this issue in a cohort of Norwegian conscripts entering their military service in 2004.⁸ This cohort of conscripts represent one of the first birth cohorts in Norway to be offered two doses of MMR through the national childhood immunization program at 15 months and 12–13 year of age. At the time, military service was mandatory for males and voluntary for women. Between January 1999 and May 2011, military conscripts were given a dose of MMR at the start of their military service as part of the health program for new recruits, since an earlier seroprevalence study showed that approximately 8% of the cohort was susceptible to measles.⁹ Previous studies have been conducted on this 2004 cohort to determine measles seroprevalence,⁸ but the seroprevalence of rubella following the additional MMR dose from this rare set of samples has not been previously investigated. The aim of this study was therefore to determine the levels of antibodies against rubella in a cohort of Norwegian conscripts, in order to evaluate the long-term rubella immunogenicity afforded by two childhood doses of MMR of the Norwegian vaccination program. Furthermore, our aim was to determine the effect of a booster dose of MMR vaccine given in adulthood, in order to monitor the immunogenicity of the vaccination.

Results

In total, specimens were available at both S1 and S2 from 495 conscripts in the 2004 conscription cohort. The majority of these conscripts were male (98%, n = 484) and they represented 18/20 counties in Norway. The mean age was 19 years, range 18–26 years, with 98% (n = 487) of the cohort being below 22 years old.

The seropositivity at S1 was 84.6% (95% CI, 81.2-87.6%) overall in the cohort (Table 1) and 99.0% (95% CI, 97.6-99.6%) of the conscripts had antibody concentrations ≥5 IU/mL at S1. The seropositivity eight months after vaccination, at S2, was 94.5% (95% CI, 92.2-96.2%) overall and 99.8% (95% CI, 98.6-100.0%) of the conscripts had antibody concentrations ≥5 IU/mL at S2. The proportion of seropositive samples by year of age of conscript ranged between 80–100% at S1 and 94–100% at S2 (Table 1) and was lowest among the youngest age groups, although the differences were non-significant. At S1, the geometrical mean of anti-rubella IgG concentrations was 22 IU/mL (95% CI, 20–23) and at S2, it was 29 IU/mL (95% CI, 27–31). The increase from S1 to S2 was statistically significant (p < 0.01).

Table 1. Study samples from Norwegian conscripts collected in 2004–2005, before (S1) and eight months after (S2) the administration of a third dose of MMR vaccine, described by sex and age.

			S1				S2
Characteristic		n	Sero-negative <5 IU/mL, %	Equivocal 5.0–9.9 IU/mL, %	Sero-positive ≥10 IU/mL, %	Geometric mean antibody concentration, IU/mL (95% CI)	Sero-negative <5 IU/mL, %	Equivocal 5.0–9.9 IU/mL, %	Sero-positive ≥10 IU/mL, %	Geometric mean antibody concentration, IU/mL (95% CI)
Sex	Male	484	1.0	14.7	84.3	21 (20–23)	0.2	5.4	94.4	29 (27–30)
	Female	11	0.0	0.0	100.0	33 (24–46)	0.0	0.0	100.0	38 (27–55)
Age	18	136	0.7	19.1	80.1	20 (18–22)	0.0	5.9	94.1	27 (24–30)
	19	257	1.6	13.2	85.2	21 (19–23)	0.4	5.8	93.8	28 (26–30)
	20	76	0.0	13.2	86.8	26 (22–32)	0.0	3.9	96.1	34 (29–40)
	21	18	0.0	5.6	94.4	22 (16–29)	0.0	0.0	100.0	30 (21–45)
	22–26	8	0.0	0.0	100.0	32 (18–55)	0.0	0.0	100.0	37 (22–63)
Total		495	1.0	14.3	84.6	22 (20–23)	0.2	5.3	94.5	29 (27–31)

Among the conscripts aged <22 years, who had previously been offered two doses of MMR vaccine as part of the national childhood immunization program in Norway, the seropositivity at S1 was 84.4% (95% CI, 80.9–87.4%). The seropositivity eight months after vaccination at S2 was 94.5% (95% CI, 92.0–96.2%) in this age-group. The anti-rubella IgG geometric mean concentrations increased from 21 IU/mL (95% CI, 20–23) at S1 to 29 IU/mL (95% CI, 27–31) at S2. As for the cohort as a whole, the increase was statistically significant (p < 0.01).

Only one conscript, whose first sample was seronegative, had not seroconverted at S2 (Table 2). All other four conscripts with seronegative results at S1 had equivocal or seropositive antibody concentrations at S2. 71% (54/76) of conscripts with seronegative or equivocal IgG concentrations at S1 were seropositive at S2.

Table 2. The number of seronegative, equivocal, and seropositive samples from Norwegian conscripts collected in 2004–2005 before (S1) and eight months after (S2) the administration of a third dose of MMR vaccine.

		S2 (n)
		Seronegative <5 IU/mL	Equivocal 5.0–9.9 IU/mL	Seropositive ≥10 IU/mL	Total
S1 (n)	Seronegative <5 IU/mL	1	3	1	5
	Equivocal 5.0–9.9 IU/mL	0	18	53	71
	Seropositive ≥10 IU/mL	0	5	414	419
	Total	1	26	468	495

For the conscripts, who were seronegative at S1, the geometric mean anti-body concentration at S1 was 4.1 IU/mL (95% CI, 3.9–4.4) and 8.3 IU/mL (95% CI, 4.4–16) at S2. For the group with equivocal results at S1, the geometric mean concentration was 8.1 IU/mL (95% CI, 7.8–8.4) at S1 and 13 IU/mL (95% CI, 12–14) at S2. For the group of conscripts, who were seropositive at S1, the geometric mean anti-body concentration was 25.7–26 IU/mL (95% CI, 24–27) at S1 and 33.6–34 IU/mL (95% CI, 32–36) at S2. All these increases were significant, p < 0.01.

The qualitative results based on optical density as described by the assay manufacturer and the antibody levels calculated by the α-method were not fully compatible. Two samples at S1 and one sample at S2, which were equivocal by qualitative evaluation based on OD, were seronegative based on the calculated antibody levels. Similarly, 69 samples at S1 and 26 samples at S2, which were positive by qualitative evaluation based on OD, were equivocal based on the calculated antibody levels.

Discussion

Our study showed that the seroprevalence of rubella antibodies, in a cohort that had been offered two childhood doses of MMR vaccine, reached 99%, when including both the equivocal and seropositive results. This therefore exceeds the 95% herd protection threshold for vaccination coverage of one dose, as defined in the WHO rubella elimination goal.^1,10 Since the introduction of the MMR vaccine in 1983, national vaccination coverage has been 90–95% for both doses.¹¹ The annual rubella vaccination coverage in Norway among 16-year-olds has been 94–95% in 2009–2014, when the collected coverage data did not specifically target age-appropriate vaccination coverage.¹² Since 2015, the vaccination coverage has been 97% for one dose of MMR vaccine and 91% for two doses.¹³

Our study evaluated the effect on immunity of a dose of MMR vaccine among a cohort of Norwegian conscripts, the majority of whom most likely had received two earlier MMR vaccine doses. Our study provides evidence on the effect of an additional MMR dose in early adulthood, something that has previously been lacking in the literature. Only five conscripts in our cohort were seronegative at the start of the study and all but one of them obtained antibody levels of ≥5 IU/mL following vaccination. As their previous personal vaccination histories were unavailable, it is possible that the five seronegative conscripts were unvaccinated at the start of the study and that some of the conscripts had not completed their childhood vaccination schedule. In addition, the majority of conscripts with equivocal concentrations became seropositive after booster vaccination, which can be considered a good response. Overall, the dose given to the conscripts led to an increase in antibody concentration, which indicates increased protection against rubella in the cohort as a whole. We saw similar significant increase when stratifying the cohort based on their seroreactivity at S1, which shows that the seronegative, equivocal, and seropositive groups all benefited from the third dose of MMR vaccine.

Our results, showing that eligibility for two childhood doses of the MMR vaccine gives rubella immunity to 85% of the cohort, are also in line with previous studies on measles and mumps seroprevalence studies from the conscripts entering military service at the same time as our cohort. Overall seroprevalence of measles antibodies was 89% among a larger cohort including 1,405 conscripts.⁸ In a partial cohort, the seroprevalence of mumps antibodies was 76% or 85% depending on the laboratory method used.¹⁴ While the protection rate among young adults eligible for two childhood doses of MMR vaccine is high for both measles and rubella, and both diseases have been eliminated in Norway,⁴ efforts to maintain the MMR vaccination coverage are important. For example, since 2015, school nurses have received reminders to ensure that all 16 year olds have been offered two doses of MMR vaccine.¹³

Previous studies have indicated that adolescents and young adults aged 15–30 years have lower levels of anti-rubella antibodies than other age-groups and may require booster vaccinations.⁷ While we measured a lower geometrical mean antibody concentration compared to some other studies,⁷ the cohort of conscripts was protected against rubella by the two vaccine doses offered through the Norwegian childhood vaccination program and our results do not indicate a need to change current guidelines.

A study from Hong Kong has shown that a higher proportion of women, who were covered by two-dose rubella vaccinations, were seronegative compared to those born before the vaccinations were implemented.¹⁵ Similarly, a previous study conducted in Norway focusing on seroprevalence among pregnant women, found that women, born before the national childhood vaccination program included two doses of MMR vaccine, had higher antibody concentrations compared to those born after.¹⁶ This is may be due to the effect afforded by exposure to the circulating virus, while the immune response to the vaccine may be less persistent.⁵ Although the differences were nonsignificant, it is interesting to note that the youngest age group (18 years) in our study had a lower proportion of seropositive conscripts compared to the other age groups, even though they had most recently been offered the second MMR vaccination through the childhood vaccination program. In our study, before the booster MMR vaccine dose, at S1, the overall geometric mean of the antibody concentration is lower than in the aforementioned Norwegian study of pregnant women when the study population is taken as a whole, but it is similar to the youngest included age groups, which are comparable in age to the conscripts in our current study.¹⁶ The results from our study are also comparable with rubella seroprevalence studies of pregnant women in Canada.^17,18 Since 2016, it is recommended that only pregnant women, who have not received two doses of MMR, are screened for rubella antibodies in Norway. If antenatal antibody concentrations are 5–9.9 IU/mL, one dose of MMR vaccine is offered, and if they are <5 IU/mL, 2 doses of MMR vaccine are offered postpartum to prevent CRS in future pregnancies.¹⁹ The results from our study show that for persons with equivocal or negative antibody concentrations, a dose of MMR vaccine will for the majority lead to seroconversion and protection of any future fetus from CRS, thus supporting the current guidelines. If increased protection would be needed, our study shows that a third dose of MMR increases anti-rubella IgG levels and therefore protection against rubella. Additional vaccinations have been mentioned elsewhere, as a possible approach for extending rubella protection throughout the child-bearing period of women's lives.²⁰

Even while large-scale vaccination has enabled elimination of rubella in several countries, questions have been raised about the possibilities to eliminate rubella through implementation of universal two-dose vaccination schedules alone. This is mainly because reports of failure to sustain long-term high-level rubella antibody levels in a vaccinated population.^15,20-22 Further similar results may be expected following known decreases in childhood vaccination coverage.²² The correlates of protection in a vaccinated population may differ from those in an unvaccinated population, and a lower anti-rubella IgG concentration may be sufficient if the virus is no longer circulating.^5,17 However, in light of the lower proportions of seropositive persons covered by a two-dose vaccination schedule and research results showing that vaccine induced antibodies against rubella wane after the second vaccine dose⁵, it is increasingly important to identify susceptible populations as elimination is achieved in several countries. Future seroprevalence studies can play an important role in guiding considerations of additional vaccinations for vulnerable groups.

Populations where the wild-type rubella virus is no longer circulating are also challenged by imported cases that could transmit rubella and potentially reestablish endemic transmission if the population is not immune. This may pose a threat to rubella elimination. For example, in the United States, 42 rubella cases were import associated in 2004–2011, a period when endemic rubella and CRS had been eliminated. Of these, 14% resulted in transmission to one further established case.²³ However, there are no reports of larger outbreaks caused by spread from imported rubella cases in countries where the disease is eliminated. This differs from the situation for measles and mumps, and indicates that elimination of rubella might be easier than for measles.

For the purposes of this study, we have defined those conscripts with anti-rubella IgG levels ≥10 IU/mL as seropositive. However, the challenge of determining an antibody concentration cut-off that affords clear indication of immunity is well known and ties in with the discussion above about antibody concentrations following vaccinations and the correlates for protection in a vaccinated population. In 1996, when the Rubella Subcommittee of the National Committee for Clinical Laboratory Standards in the United States lowered its recommended breakpoint for seropositivity from 15 to 10 IU/mL, it noted that reinfection has been documented also in persons with antibody concentrations ≥15 IU/mL.²⁴ Varying threshold values from >7 to >15 IU/mL to define seropositivity have been employed in the literature, and some studies argue that low positive (equivocal) concentrations (4–7 IU/mL) offer protection against rubella.^{5,10,17,18,25} It has been suggested that in presence of herd immunity, in populations where the rubella virus is not circulating, detectable anti-rubella IgG levels (≥4 IU/mL) are sufficient for immunity.^5,17 In addition to humoral immunity, cell mediated immunity is also likely to provide protection against rubella, although the longevity of response is uncertain.^26,27 We believe that in Norway, where vaccination coverage is high and rubella has been eliminated, the antibody level of ≥5 IU/mL offers protection against rubella.

The discordant results, between quantitative OD based assessment and assessment based on IU concentration in our study, derive from the cut-offs we use for seroreactivity, and the discrepancy disappears if cut-offs for seronegative and seropositive results are lowered to <4 IU/mL and >6 IU/mL, respectively. Different cut-offs employed in laboratory assays may contribute to confusion in determining clear cut-off values for protection. Discordant results obtained by different immunoassays for rubella seroreactivity when compared to reference tests have been previously described, as a study of eight immunoassays in Europe found that over half of the screened women, who were assessed as being susceptible to rubella, were in fact seropositive when studied by the reference tests.²⁸

It has been argued that long-term storage of sera may lead to decrease in antibody levels. However, freezing for long-term storage is routine practice in serological investigations and a study comparing serology, including anti-rubella antibodies, from fresh and frozen specimens found no difference between results.²⁹ Therefore, we do not find reason to believe that storage has affected our results.

There are several limitations in this study. Firstly, we do not have access to personal vaccination records of the conscripts included in our cohort, and therefore do not know for certain whether they have received the two doses of MMR vaccine that they have been offered through the Norwegian childhood vaccination program. However, we know that vaccination coverage in Norway is high.^11-13 The cohort in our study is geographically diverse within Norway and the results can be considered applicable to the young adult male population in the country. However, our cohort lacks foreign residents and persons who do not serve in the military e.g. because of health or ideological considerations, and it is principally male. Previous studies have shown no differences between rubella seropositivity status and sex.^30,31 Some studies indicate that adult females generally may develop higher antibody responses to vaccines than males.³² Based on this, we believe that our study can be used to inform public health policy regarding rubella vaccinations of the general healthy young adult population and pregnant women. The S2 samples in our study were taken eight months after the third MMR vaccine dose was administered. By then, antibody levels are likely to have declined from peak values,³³ and our study may therefore offer a better understanding of the long-term effect of the vaccine.

In our study, the cohort of healthy young adults in Norway was protected against rubella, after being eligible for two childhood MMR vaccine doses, and efforts should be made to maintain high vaccine coverage to ensure immunity in the future. A third dose of MMR vaccine administered in early adulthood provided increased protection against rubella, as it led to an increase in the antibody levels in our cohort and seroconversion for the majority of seronegative persons, and can therefore be used to increase protection against rubella, if needed.

Materials and methods

Study population

All conscripts entering their military service in August 2004 were invited to participate in the study.⁸ We collected information on month and year of birth, sex, and municipality of residence. Individual childhood vaccination status of the conscripts was not available.

Specimen collection

Blood samples were collected from Norwegian conscripts (n = 495) in 2004, before (S1) and in 2005, eight months after (S2) an additional dose of MMR vaccine was given. The samples were stored at -20°C kept at monitored temperatures until testing at the Norwegian Institute for Public Health (NIPH) in 2015.

Laboratory methods

An indirect enzyme-linked immunosorbent assay (Enzygnost Anti-Rubella-Virus IgG, RUB/IgG, Siemens) was used to measure the anti-rubella immunoglobulin G (IgG) in the samples according to the manufacturer's instructions. Samples with equivocal results were retested, if enough specimen material was available. The optical density (OD) cut-offs were OD <0.100, negative; OD 0.100–0.200, equivocal; OD >0.200, positive. The conversion of the OD test result into concentration in international units (IU) per mL was performed by the α-method, as recommended by the assay manufacturer.

Definition of seroreactivity

Anti-rubella IgG concentrations <5 IU/mL were regarded as negative, concentrations from 5.0–9.9 IU/mL as equivocal, and ≥10 IU/mL as positive, based on previous Norwegian studies and national guidelines.^16,19 Samples with IgG antibody activity lower than the OD cut-off (<0.100) or with uncorrected OD readings higher than 2.5 were recorded as the lowest and highest values of detection, respectively.

Data analysis

We calculated the overall seroprevalence before and after the administered MMR dose for the cohort of conscripts (seropositive tests/number of performed test) with 95% confidence intervals (95% CI). Data was stratified for analysis by sex, age in years, and the following age groups: <22 year-old (eligible for two childhood doses of MMR vaccine through the national vaccination program) and ≥22 year-old (ineligible for two childhood doses of MMR vaccine through the national vaccination program). Differences in the proportion of seropositive samples were examined for statistical significance using the Chi-square test. Geometrical means of anti-rubella IgG were calculated, statistical significance of differences between mean antibody concentrations at S1 and S2 were examined with the Kruskal-Wallis test. A p value <0.05 was considered as significant and STATA 14 (StataCorp LLC) was used for statistical testing.

Ethics approval and consent to participate

The Norwegian Regional Committee of Medical Research Ethics reviewed and approved the study. Written informed consent was obtained from the conscripts prior to enrolment in the study.

Abbreviations

CRS	=	congenital rubella syndrome
IU	=	international units
MMR	=	measles mumps rubella
NIPH	=	Norwegian Institute for Public Health
OD	=	optical density
WHO	=	World Health Organization

Disclosure of potential conflicts of interest

L.S. is a co-investigator in an unrelated study for which the National Institute for Health and Welfare, Finland, has received research funding from GlaxoSmithKline Biologicals SA. The other authors report no potential conflicts of interest.

Acknowledgments

We gratefully acknowledge the conscripts participating in our study and the late Kirsti Vainio for her substantial role in collecting the material and designing the study. We would like to warmly thank Hang Thi Ngoc Le and other staff at the Department of Virology, NIPH, for their skillful technical work. We gratefully acknowledge local and international EUPHEM coordinators and supervisors for guidance during this project and Emily MacDonald and Loredana Ingrosso for reviewing the manuscript.

Additional information

Funding

No external funding received.