Seropositivity of measles antibodies in the Israeli population prior to the nationwide 2018 – 2019 outbreak

ABSTRACT

Measles vaccine is administered in Israel as part of the routine childhood immunization program, at ages 1 and 6 years. In this study, we assessed seropositivity of the Israeli population against measles before the onset and propagation of the 2018–2019 measles outbreak. From the Israel Center for Disease Control National Serum Bank, 3,164 samples collected during 2015 were tested for measles antibodies. All the tests were performed using Enzyme-Linked Immunosorbent Assay (ELISA) commercial kit (Enzygnost, Anti-Measles Virus/IgG: Behring, Marburg, Germany). The overall seropositivity rate for measles was 90.7%. The seropositivity rate at 6 months and younger was 48.9%, and decreased to 3.8% among infants aged 6–11 months. Seropositivity increased to 90.7% in the 1-4-year age group, and reached 96.1% for 5–9 year-old children. Our results suggest high immunity in the Israeli population against measles virus, but not high enough to prevent outbreaks because of pockets of specific population groups with low immunization coverage. Infants between ages 6 and 11 months and children younger than 2 years had the lowest seropositivity rates being the age groups with the highest attack rates of measles during the epidemic of 2018. Efforts should be aimed at avoiding any delay in vaccination once a child reaches the age of 1 year and improving immunity levels in children aged 1–4 years.

Introduction

Measles is an exceptionally contagious viral infection that is capable of causing epidemics.¹ Measles is highly communicable and infectivity is higher than 90% among susceptible individuals. In the pre-vaccine era, measles affected nearly every individual during childhood.^1,2

The measles vaccine was introduced in Israel in 1967, for infants aged 9 months, and has dramatically changed the epidemiology of the disease.³ In 1971, the vaccine dose administration was postponed to 12 months, and in 1980 to 15 months. Since the introduction of the vaccine in Israel, the incidence of measles declined dramatically, but outbreaks have still been documented. The largest of these, in 1982, presented 7,864 new cases and an incidence rate of 195.3 per 100,000.³ In 1988, the measles vaccine was introduced in the MMR combination, with mumps and rubella, at age 15 months.³ In 1990, a booster was introduced at age 6 years. Since 1994, the first MMR dose is administered at 12 months.³ Overall, Israel reached measles elimination in 2014,⁴ and high immunization coverage continues to be reported (97%).⁵ However, the measles virus has caused frequent outbreaks over the years, for example in 2003–2004, 2007–2008, 2012, 2015 and 2017;^3,6,7 these have originated from imported cases. Since March 2018, Israel has experienced a large, ongoing measles outbreak, with more than 4,000 confirmed cases, which constitutes part of a global outbreak.^8–11

The aim of the current study was to assess the seropositivity rate of measles antibodies in the Israeli population before a large measles outbreak.

Materials and methods

Study design and sampling

A cross-sectional sero-survey was conducted using 3,164 residual serum samples, retrieved from the Israel Center for Disease Control National Serum Bank. Established in 1997, this database has served as an ongoing collection since. The samples were collected during 2015, from five laboratories, located geographically throughout Israel and representing the distribution of the Israeli population by age, gender and population group (“Jews and others” included Jews, non-Arabic Christians and populations not affiliated with a religion; “Arabs” included Muslims, Arab Christians and Druze). The samples were collected from both healthy blood donors and residual sera from diagnostic laboratories. Samples from patients with suspicious or confirmed immune deficiency were disregarded. The number of serum samples tested, stratified by age groups in equal numbers for males and females, was determined using the European Sero-Epidemiology Network (ESEN) guidelines.¹² We calculated the sample size required to detect an anticipated difference between age groups of 15% in seroprevalence, by using a 5%-level of significance, two-sided with 80% power.

Sera sample collection was approved by the legal department of the Israeli Ministry of Health.

Laboratory methods

Levels of measles virus specific IgG antibodies were measured using Enzyme-Linked Immunosorbent Assay (ELISA) commercial kit (Enzygnost® Anti-Measles Virus/IgG, Marburg, Germany). The samples were classified as follows: optical density (OD)>0.2 was considered positive, 0.1–0.2 equivocal, and <0.1 negative. According to the manufacturer, kit sensitivity was 99.6% and specificity was 100.0%.

A Plaque Reduction Neutralization Test (PRNT) was performed on 12 samples, which were classified according to the ELISA test as equivocal. PRNT was performed by the laboratory protocol established for the WHO measles aerosol project.¹³ In brief, the assay was performed on Vero cells infected with the measles virus (Schwartz vaccine strain, NIBSC code: 92/648). The antibody concentration was standardized using the WHO 3^rd International Standard for measles antibody, containing 3000 mIU/mL (NIBSC code 97/648). This enabled the transformation of a 50% neutralizing antibody end-point dose (titer, ND₅₀) of test samples to antibody concentrations in terms of mIU/mL. The analytical cutoff value in this assay is ND₅₀ < 1/8, which is the lowest dilution at which sera are tested. The interpretive cutoff value was ≥120 mIU/mL, the concentration of antibody considered to be protective.¹⁴

Data analysis

Descriptive statistics were conducted for age, gender and population-groups. Results were expressed as frequencies, and 95% confidence intervals (CI) of measles antibody seropositivity rates were calculated. Simple and multiple logistic regressions were applied to examine associations of seronegativity with age groups, gender and population group. Unadjusted and adjusted odds ratio (OR) and 95% CI were calculated. A sensitivity analysis was performed, using two models. In the first model, we analyzed the data while considering all equivocal samples as negative. In the second model, we considered all equivocal samples as positive. Statistical significance was defined at a level of p-value <0.05. Statistical analyses were performed using SAS Enterprise Guide software package (version 7.12, SAS Institute Inc., Cary, NC, USA).

Results

Measles IgG antibody levels were measured in 3,164 samples. Of them, 2,559 (80.9%; 95% CI: 79.5–82.2%) were positive, 310 (9.8%; 95% CI: 8.8–10.9%) equivocal and 295 (9.3%; 95% CI: 8.3–10.4%) negative. The demographic characteristics of the persons from whom the samples were tested are shown in Table 1. Examination of twelve equivocal samples by PRNT demonstrated protective levels of measles IgG antibody (≥120 mIU/mL). Based on these data, all equivocal samples were considered positive and the overall seropositivity rate was 90.7%; 95% CI: 89.6–91.7%. In the model in which all equivocal samples were classified as negative, only 80.9% (95% CI: 79.5–82.2%) were seropositive.

Table 1.

Demographic characteristics of the persons from whom the samples were tested for measles, N = 3,164, 2015

		N	%
Age Group (Years)	≤0.50	47	1.5
	0.51–0.99	53	1.7
	1.00–4.99	399	12.6
	5.00–9.99	480	15.2
	10.00–14.99	499	15.8
	15.00–19.99	488	15.4
	20.00–24.99	200	6.3
	25.00–29.99	200	6.3
	30.00–34.99	200	6.3
	35.00–39.99	200	6.3
	40.00–49.99	200	6.3
	50.00–59.99	112	3.5
	60.00+	86	2.7
Gender	Female	1,565	49.5
	Male	1,599	50.5
Population Group	Jews and Others	1,780	56.3
	Arabs	1,384	43.7

Among children aged up to 6 months, 48.9% of the samples tested positive. Seropositivity decreased dramatically to 3.8% among infants aged 6–11 months (Table 2, Figure 1). In the 1–4 year age group, seropositivity increased to 90.7%, though it was only 82.0% and 88.0% in the second and third years of life, respectively. In the age 5–9 year group, seropositivity increased to 96.0%. The seropositivity rate declined to 91.5% in the 40–49 year age group. In the 50–59 and 60 years and above age groups, seropositivity rates were 95.5% and 98.8%, respectively. The total seropositivity rate in the study population aged ≥1.0 years was 92.8%; 95% CI: 91.8–93.7%. The proportions of those categorized as equivocal in the 15, 17, 18, 20–24, 25–29 and 30–34 year age groups were high (19.0%, 16.2%, 16.5%, 19.0%, 15.5% and 20.0%, respectively) (Figure 1).

Table 2.

Seroprevalence of measles antibodies by demographic characteristics and multiple analysis of seronegativity to measles in Israel, 2015 (n = 3,164)

		Tested	Seropositive		Seronegativity
		N	%; 95% CI¥	p-value¦	OR§	95% CI¥	p-value¤
Age Group (Years)	≤0.50	47	48.9; 34.1–63.9	<0.01	93.1	11.9–726.2	<0.01
	0.51–0.99	53	3.8; 0.5–13.0		>999.9	189.4->999.9	<0.01
	1.00–4.99	399	90.7; 87.4–93.4		8.6	1.2–63.6	0.03
	5.00–9.99	480	96.0; 93.9–97.6		3.5	0.5–26.3	0.23
	10.00–14.99	499	92.6; 89.9–94.7		6.8	0.9–50.0	0.06
	15.00–19.99	488	92.6; 89.9–94.8		6.7	0.9–49.7	0.06
	20.00–24.99	200	93.5; 89.1–96.5		5.8	0.7–45.2	0.09
	25.00–29.99	200	93.0; 88.5–96.1		6.3	0.8–49.0	0.08
	30.00–34.99	200	88.5; 83.2–92.6		11.0	1.5–82.6	0.02
	35.00–39.99	200	91.0; 86.2–94.6		8.3	1.1–63.1	0.04
	40.00–49.99	200	91.5; 86.7–95.0		7.8	1.0–59.3	0.04
	50.00–59.99	112	95.5; 89.9–98.5		3.9	0.4–34.0	0.22
	60.00+	86	98.8; 93.7–100.0		Ref.	-	-
Gender	Female	1,565	91.6; 90.1–92.9	0.09	Ref.	-	-
	Male	1,599	89.8; 88.2–91.2		1.3	0.9–1.7	0.09
Population Group	Jews and Others	1,780	91.0; 89.5–92.2	0.54	Ref.	-	-
	Arabs	1,384	90.3; 88.6–91.8		1.2	0.9–1.5	0.28

^¥CI-Confidence Interval

^¦ P-value representing the difference between the categories within the variable

^§ OR-Odds Ratio

^¤ P-value representing the difference of the category to the reference category

Figure 1.

Age-specific seroprevalence of measles IgG antibodies in Israel, 2015 (n = 3,164)

The seropositivity rate was 91.6% among females and 89.8% among males; and 91.0% among Jews and others, and 90.3% among Arabs (Table 2). No significant differences were observed between the genders (p-value = 0.09) and between the population groups (p-value = 0.54) (Table 2).

In the simple and multiple analyses, the odds ratios of being seronegative to measles were significantly higher in the following age groups: first 6 months of life, 6 to 11 months of age; and 1–4, 30–34, 35–39 and 40–49 years (Table 2)

Logistic regression was used to examine interactions between age group, gender and population group, but no significant interactions were detected.

Discussion

We demonstrated a 90.7% seropositivity rate of measles antibodies among the general Israeli population in 2015, which preceded the large-scale outbreak that has been observed since March 2018. A seroprevalence of 90.7% is lower than the WHO recommended 93–95% threshold.¹⁵ The relatively low seroprevalence of 92.8% among those vaccinated (≥1.0 years) is mainly due to the 82.0–88.0% seropositivity in the 1 and 2 year-old age groups. This, in addition to the possibly waning immunity among individuals aged 15–35 years, may enable transmission of the virus from imported cases by means of susceptible groups that are not immune. As in Israel, also in the Netherlands, an outbreak occurred within a highly vaccinated population with high seropositivity rates.¹⁶ Trentini suggested that current vaccination policies are not sufficient to achieve and maintain measles elimination in most countries, and recommended focusing on increasing childhood immunization rates.^17,18

The seropositivity rate documented in our study (90.7%) is higher than seropositivity rates reported in Turkey (82.2%),¹⁹ Shaanxi province-China (85.9%),²⁰ and Australia (89.7%);²¹ but lower than rates in the United States (92.0%),²² the Czech Republic (93.0%),²³ Dongguan-China (93.4%)²⁴ and the Netherlands (95.7%).²⁵ Differences between vaccination programs and implementation years can explain the wide variation between countries. The high seropositivity rate observed in the present study is expected, in light of the high vaccination coverage reported by the Israel Ministry of Health and the high effectiveness (97%) of the vaccine in preventing measles,^1,26 especially for those who receive both doses. Targeting low age groups to adhere to vaccination schedule, especially at age 12 months, may bridge the gap between the observed seropositivity rate and the WHO 93–95% recommended threshold.¹⁵

We showed significant differences in seropositivity rates between various age groups. In the first year of life, before the administration of the first vaccine dose, seropositivity was the lowest; in the first six months of life, almost half were seropositive to measles, probably due to maternal antibodies. This was followed by a dramatic decrease in seropositivity in infants aged 6–11 months, due to loss of maternal antibodies and very low exposure to the natural virus. Our findings, corroborate the Czech Republic,²³ the Netherlands²⁵ and Korea²⁷and suggest that infants younger than age 1 year are at the highest risk for contracting measles especially those aged 6–11 months. Indeed, children of this age were highly affected in the current measles outbreak in Israel (incidence rate of 196.0 per 100,000),²⁸ similar to previous measles outbreaks in Israel and abroad.^29,30 Infants younger than 12 months can have atypical measles presentation and are at higher risk for complications.³¹ Therefore, during a large scale outbreak, administration of an additional vaccine dose in the 9–12 months age group, in addition to the routine vaccine schedule, should be considered. This is despite relatively lower seroconversion rates and neutralizing titers that are achieved in these age groups.^32,33

Among children aged 1–4 years, who should have received the first vaccine dose, seropositivity increased significantly and reached 90.7%. For children in their second and third years of life, the seropositivity rates of 82.0% and 88.0%, respectively, are lower than the threshold needed to confer herd immunity. The odds of being seronegative to measles in this age group were higher than among older persons who had been exposed to the natural virus. Most probably, among this group, there is a high rate of persons with delayed first dose immunization, and who were thus unvaccinated at the time of sample collection. The high odds for seronegativity in this age group compared with older ages were reflected in the 2018–2019 measles outbreak that occurred in Israel. Accordingly, this is the age group with the second highest incidence rate of disease (118.9 per 100,000),²⁸ after the infant age group. To better protect the <1 year-old population, with no or waning levels of maternal measles antibodies, all efforts should be aimed at avoiding any delay in vaccination once a child reaches age 1 year. Strengthening immunity levels in children aged 1–4 years, together with the high observed seroprevalence of >95% in the 5–9 years age group, may promote herd immunity in the <1 year age group.

We showed a decline over time in measles seropositivity following vaccination, from 99.0% in those aged 4 years who were vaccinated with one dose, to 92.9% at age 6 years, prior to the second dose. Among persons who were vaccinated with two doses, a decline in seropositivity was observed, from 97.0% at age 10 years to 88.5% at age 30–34 years. Similar declines were reported by others.^{18,23–25,27,34}

The highest seropositivity rates were found among persons older than age 50 years. This is most probably associated with naturally-acquired immunity in childhood, as has also been reported by others.^23,24,34

In our study, we regarded the equivocal test results as positive. This decision was supported by the findings on the neutralizing capabilities, which are evident from the subsample of sera with equivocal ELISA antibody levels. This approach, also adopted by Ratnam et al. in 1995,³⁵ Hogg et al. in 2006,³⁶ Cohen et al. in 2007¹³ and Dorigo-Zetsma et al. in 2015,³⁷ demonstrated that equivocal samples have >120mlU/ml in PRNT, the cutoff of protection. In a scenario in which all equivocal samples would be classified as negative, only 80.9% (95% CI: 79.5–82.2%) would be seropositive. This contrasts with the 90.7% seropositivity (95% CI: 89.6–91.7%) in the selected scenario in which all equivocal samples were classified as positive. If the measles seropositivity rate in Israel had been 80.9%, which is significantly lower than the WHO recommendations, the occurrence of measles would have been higher than that observed, and most probably more extensively spread in the general Israeli population. Classifying equivocal samples as positive, and thus attaining the 90.7% seroprevalence, corroborates the incidence rates and patterns of measles transmission observed in Israel. We report here a high prevalence of equivocal samples among persons aged 15–34 years, as was reported previously.²⁴ The explanation for the high proportion of equivocal samples in this age group may be the decrease in antibody levels that occurs after vaccination (waning immunity). Yet, the incidence rates of measles in the 10–19, 20–29 and 30–39 year age groups were low in the 2018–2019 outbreak (21.7, 13.3 and 12.1 per 100,000, respectively).²⁸ Thus, an antibody level that is not high enough to be classified as positive may confer protection against infection. The neutralizing capabilities shown by the subsample of sera with equivocal ELISA antibody levels support this assumption. Further examination is important, of the implication of the high proportion of equivocal samples, and of whether individuals with equivocal results are capable of transmitting the virus.

The strength of our study is the relatively large number of samples tested overall, in all age groups, and prior to a large outbreak of measles in 2018–2019.

The study has limitations. We did not have data available on individual history of vaccination against measles. The numbers of samples tested in the ≤0.50, 0.51–0.99 and 60.00+ year age groups were lower than for other age groups. Further, our classification of equivocal results as seropositive is supported by the testing of only a relatively small number of equivocal samples by a neutralization assay.

In conclusion, despite the overall 90.7% measles seropositivity in Israel, pockets of unvaccinated age groups enable propagation of the virus, and explain the occurrence of outbreaks. Establishing solid herd immunity by reaching the desired high level of vaccination coverage and eliminating pockets of under coverage is necessary to interrupt measles transmission.³⁸

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.