Persistence of Vaccine‐Induced Antibody to Measles 26–33 Years after Vaccination

Article (PDF Available)inThe Journal of Infectious Diseases 189 Suppl 1(s1):S123-30 · May 2004with16 Reads
DOI: 10.1086/380308 · Source: PubMed
Abstract
Because measles-specific antibody titer after vaccination is lower than after natural infection, there is concern that vaccinated persons may gradually lose protection from measles. To examine the persistence of vaccine-induced antibody, participants of a vaccine study in 1971, with documentation of antibody 1–7 years after vaccination, were followed up in 1997–1999 to determine the presence and titer of measles antibody. Of the 56 participants (77% were 2-dose recipients), all had antibodies detected by the plaque reduction neutralization (PRN) antibody assay an average of 26–33 years after the first or second dose of measles vaccine; 92% had a PRN titer considered protective (>1:120). Baseline hemagglutination inhibition antibody titer in 1971 strongly predicted follow-up PRN antibody titer (P<.001). Persistence of antibody in these primarily 2-dose recipients supports the current elimination strategy to achieve and sustain high population immunity with a 2-dose schedule.
Duration of Measles Vaccine–Induced Antibody JID 2004:189 (Suppl 1) S123
SUPPLEMENT ARTICLE
Persistence of Vaccine-Induced Antibody to Measles
26–33 Years after Vaccination
Mark S. Dine,
1
Sonja S. Hutchins,
2
Ann Thomas,
2,a
Irene Williams,
3
William J. Bellini,
3
and Stephen C. Redd
4
1
Private practice, Cincinnati, Ohio;
2
National Immunization Program,
3
National Center for Infectious Diseases, and
4
National Center
for Environmental Health, Centers for Disease Control and Prevention, Atlanta, Georgia
Because measles-specific antibody titer after vaccination is lower than after natural infection, there is concern
that vaccinated persons may gradually lose protection from measles. To examine the persistence of vaccine-
induced antibody, participants of a vaccine study in 1971, with documentation of antibody 1–7 years after
vaccination, were followed up in 1997–1999 to determine the presence and titer of measles antibody. Of the 56
participants (77% were 2-dose recipients), all had antibodies detected by the plaque reduction neutralization
(PRN) antibody assay an average of 26–33 years after the first or second dose of measles vaccine; 92% had a
PRN titer considered protective (
11:120). Baseline hemagglutination inhibition antibody titer in 1971 strongly
predicted follow-up PRN antibody titer ( ). Persistence of antibody in these primarily 2-dose recipientsP
! .001
supports the current elimination strategy to achieve and sustain high population immunity with a 2-dose schedule.
To prevent reestablishment of indigenous measles trans-
mission in the United States, a high level of protection
against measles in the population must be sustained
[1–3]. A potential obstacle to sustaining this high level
is waning of vaccine-induced protection over time, or
secondary vaccine failure [4–7]. Because measles-spe-
cific antibody titer after vaccination is lower than after
natural infection, there is concern that vaccinated per-
sons may gradually lose protection from measles. Sec-
ondary vaccine failure, in contrast to primary vaccine
failure (no protection after vaccination), is a concern
because of the potential insidious challenge to measles
elimination. For instance, if vaccine-induced immu-
nity waned to nonprotective levels in a high propor-
tion of vaccinated adults, the level of population pro-
tection might decline to allow recurrence of endemic
disease. By means of statistical modeling, Mossong et
Presented in part: 34th National Immunization Conference, 5–8 July 2000,
Washington, DC.
a
Present affiliations: Office of Disease Prevention and Epidemiology, Oregon
Department of Human Services, Portland (A.T.); National Center for Environmental
Health, Centers for Disease Control and Prevention, Atlanta (S.R.).
Reprints or correspondence: Dr. Sonja S. Hutchins, National Immunization
Program, Mailstop E-61, Centers for Disease Control and Prevention, 1600 Clifton
Road, Atlanta, GA 30333 (ssh1@cdc.gov).
The Journal of Infectious Diseases 2004;189:S123–30
2004 by the Infectious Diseases Society of America. All rights reserved.
0022-1899/2004/18909S1-0020$15.00
al. [8] predicted waning of vaccine-induced immunity
25 years after immunization.
Isolated cases of measles in persons with documented
seroconversion after vaccination have been reported [4,
6, 7]. These secondary vaccine failures among adults
in the United States, Canada, and China are of concern
and provide evidence that vaccine-induced protection
may wane. However, the magnitude of the problem
seems small. The number of reported measles cases in
the United States has been at record low levels since
1996, and the number of reported cases in adults has
been very small (32 cases in persons 20 years of age
in 1999). Though a small number, measles cases in
vaccinated adults (22% of cases in adults) suggest that
waning vaccine-induced protection may explain some
vaccine failures [9, 10].
Although waning vaccine-induced immunity does
occur, population-based prevalence is not known in the
United States. Little is known about the levels of measles
protection among adults. The third National Health
and Nutrition Examination Survey (NHANES III), con-
ducted during 1988–1994, found that 93% of partici-
pants aged 6 years had measles antibody [11]. This
finding suggests that measles immunity in the United
States is high. However, measles vaccination history was
not available for NHANES participants and could not
be linked with seroprevalence results. Even if vaccina-
by guest on November 11, 2015http://jid.oxfordjournals.org/Downloaded from
S124 JID 2004:189 (Suppl 1) Dine et al.
tion status were known, we still could not distinguish between
primary and secondary vaccine failure without knowing the
initial response to vaccination. The present study was under-
taken to estimate the prevalence of measles antibody among a
population of adults with documentation of vaccination an
average of 26–33 years before and whose initial documentation
of measles antibody titers occurred in 1971 after initial vacci-
nation [12]. By studying blood drawn several decades after
vaccination and documentation of initial measles antibody, it
should be possible to estimate whether measles antibody per-
sists in this population of vaccinated adults and to examine
whether loss of antibody over time results in susceptibility to
measles in the United States when continued antigenic stim-
ulation is unlikely because the risk of measles exposure is low.
METHODS
1971 Serological Study
In the initial 1971 study in Cincinnati, Ohio, measles serological
status was examined for 182 children before and after vacci-
nation with a live measles virus vaccine [12]. Of the 182 chil-
dren, 131 were initially vaccinated at
!1 year of age and 51
were vaccinated at 1 year of age. Of the 182 children, 97
received a live attenuated measles vaccine (74 with immune
globulin and 23 with no record of receipt of immune globulin)
and 85 received a live, further attenuated measles vaccine.
About one-half were revaccinated between 19 months and 13
years of age with a live, further attenuated vaccine (Moraten
vaccine strain, the only strain now currently in use in the United
States). Revaccination was 1–7 years after the first dose of mea-
sles vaccine. All of the study participants had a documented
measles antibody titer after the first or second dose of vacci-
nation [12]. Antibody was measured by the standard hemag-
glutination inhibition (HI) assay, an assay no longer commonly
used. Participants who had received 2 doses were revaccinated
during the 1971 study because their first dose was usually ad-
ministered at
!1 year of age. In 1965, the recommended age
at vaccination was increased from 9 months to 12 months [3].
At that time, the Advisory Committee on Immunization Prac-
tices recommended revaccination of children vaccinated at
!1
year of age [3]. In addition, children in the 1971 study were
also revaccinated if their HI antibody titer was 1:40.
1997–1999 Follow-Up Study
Study design and population. In the follow-up study, we
recruited as many of the participants from the 1971 study pop-
ulation as possible. Many of the patients still resided in the
Cincinnati area. Participants were found through pediatric
medical records, local phone books, and Web-based directories
for the United States. The study protocol was reviewed and
approved by the Institutional Review Boards of the Children’s
Hospital, Cincinnati, and of the Centers for Disease Control
and Prevention (CDC).
Collection of information and blood. Participants in the
1971 study who agreed to participate in the follow-up study
were interviewed and had blood drawn in the pediatrician’s
office or a medical laboratory. Arrangements were made for a
representative from a local chapter of the Visiting Nurses’ As-
sociation to complete the interview and draw blood from par-
ticipants who lived outside the Cincinnati area. Participants
were interviewed to obtain demographic information and his-
tory of measles exposure, disease, and vaccination in the in-
tervening years between 1971 and 1999. Initial vaccination his-
tory and antibody titers were abstracted from the original office
medical records from the 1971 study. In the follow-up study,
1 blood specimen was collected for each participant between
1997 and 1999. Sera were frozen and shipped to the CDC for
analysis.
Participants in the 1997–1999 follow-up study were classified
into 3 groups on the basis of type of vaccine received, number
of doses received, and timing of blood drawn at the initial
study in 1971. In group A, participants received a single dose
of either a live attenuated vaccine strain (Edmonston B) with
immune globulin or a live, further attenuated vaccine (Schwartz
or Moraten) without immune globulin before the 1971 study.
Blood for HI antibody testing was collected in 1971, on average,
4 years after initial vaccination (range, 2–8 years). In group B,
participants received a first dose of a live, attenuated vaccine
strain with immune globulin, a live, further attenuated vaccine
without immune globulin, or an inactivated vaccine (Edmon-
ston B) before the 1971 study and a second dose of a live,
further attenuated vaccine (Moraten only) during the 1971
study. For this group, blood for HI antibody testing was col-
lected in 1971, an average of 6 years (range, 2–8 years) after
the first dose of vaccine and immediately before the second
dose. Some children returned to have a second blood sample
drawn 2–11 weeks after revaccination. In group C, participants
received a first dose of a live, attenuated vaccine or an inac-
tivated vaccine as in group B, the second dose was either the
Moraten vaccine or an inactivated vaccine, and the third dose
was a Moraten vaccine. Blood for HI testing was collected in
1971, an average of 7 years (range, 6–7 years) after the first
dose of vaccine and immediately before the second or third
dose if the second dose was an inactivated vaccine. For the
purposes of the follow-up study, for all 3 groups, only the HI
titer from blood collected after the first dose of measles vac-
cination and before revaccination during the 1971 study for
groups B and C was defined as the baseline HI titer.
Laboratory Methods
Two serological assays were performed at the CDC for each blood
specimen from a participant to detect measles antibody. First,
by guest on November 11, 2015http://jid.oxfordjournals.org/Downloaded from
Duration of Measles Vaccine–Induced Antibody JID 2004:189 (Suppl 1) S125
the plaque reduction neutralization (PRN) assay was done, which
is considered the “gold standard” [13]. This test is 12%–17%
more sensitive than the EIA and is considered to be 60-fold
more sensitive than the HI assay used in the 1971 study [14–
17]. Because the EIA is practical and widely used, serum sam-
ples were also tested by EIA. We were not able also to examine
follow-up serum samples for HI antibody because the HI assay
is no longer widely available.
To detect PRN antibody, serum was serially diluted beginning
at 1:8 [13]. The PRN antibody titer was reported as the dilution
of serum that reduced plaques by 50%. Adults with an antibody
titer of 1:120 were considered protected from measles because
a titer at that level was found in several outbreaks to be as-
sociated with protection against classical measles [16, 18, 19].
Analysis of risk factors for measles susceptibility was limited
only to PRN antibody results, because a serological correlate
of protection had been identified for this assay.
The EIA detected measles-specific IgG antibody to recombi-
nant measles virus nucleoprotein [14]. Serum was defined as
positive for antibody by an optical density reading of P N
and , where P was the average optical density of0.09 P/N 3.00
duplicate positive wells and N was the average value of duplicate
negative wells.
Statistical Analysis
Because the sample size was relatively small for estimating prev-
alence and risk factors for measles susceptibility, exact statis-
tics were used. StatXact-3 and LogXact software were used for
univariate, bivariate, and multivariate analyses [20, 21]. We
estimated the 95% confidence interval of the prevalence of
antibody using the binomial distribution. The Fisher-Freeman-
Halton test was performed to detect differences between pro-
portions, and the Kruskal-Wallis test was used to examine dif-
ferences in the overall distribution of a continuous variable in
a population [20]. Although this is a prospective study, the
distribution of a continuous variable (exposure variable) in a
bivariate analysis could be examined relative to the PRN titer
(outcome variable) with this sample size only by use of exact
statistics. Thus, instead of examining the difference in the PRN
antibody titer between exposure populations, we examined the
distribution of the exposure variable by the outcome variable.
To conduct unconditional logistic regression, the unstratified
version of logistic regression in LogXact was attempted but was
unsuccessful because of very sparse data [21]. All antibody titers
were reported by computing the reciprocal of the titer.
RESULTS
Study population. Of the 182 patients in the initial 1971 co-
hort, we were able to locate the charts of 174. Of these, 64
consented to participate in the follow-up study, were interviewed,
and had blood drawn. Of the 64 participants, 4 were excluded
because they did not have a vaccination history or were found
later to have had a history of measles exposure or disease. An
additional 4 persons were excluded because they had received a
second or third dose of measles vaccine within 10 years of the
1997–1999 follow-up study. None of the remaining 56 partici-
pants reported having had measles or a known exposure to mea-
sles during the follow-up period after the 1971 study.
The 56 study participants in the 1997–1999 follow-up study
were slightly older and were more likely to be female than were
nonparticipants ( ) (table 1). In addition, participants
P
! .01
were slightly older than nonparticipants at the first and second
dose of measles vaccination ( and , respectively).
P p .04 P
! .01
There were no other statistically significant differences between
study participants and nonparticipants by vaccine group (type
and number of vaccine doses), years since vaccination, baseline
HI antibody titer, the age at HI testing, and years since HI titer
was determined.
Study participants were born between 1957 and 1969, the
early part of the vaccine era. Their mean age at the 1997–1999
follow-up study was 34 years (range, 29–40 years; table 1).
Seventy-one percent of participants were female. Most partic-
ipants (77%) had received 2 doses of measles vaccine and were
classified as group B. Of the 43 participants in group B, 42 had
received a first dose of the live, attenuated vaccine with immune
globulin or a live, further attenuated vaccine without immune
globulin (table 2). The remaining participant in group B re-
ceived a first dose of an inactivated vaccine before revaccination
with the live vaccine. Ten participants in group A received a
single dose of a live vaccine, whereas 2 participants in group
C received 3 doses of a live vaccine. The remaining participant
in group C received 2 doses of an inactivated vaccine before a
live vaccine. Overall, the mean age at the first dose was 19
months, at the second dose was 8 years, and at the third dose
was 17 years (table 1). Participants in group C with 3 doses
had received the third dose in 1971, 1983, or 1987.
Depending on the number of doses of measles vaccine re-
ceived, the mean number of years between the last dose of
measles vaccine and the 1997–1999 follow-up study was 18–
33 years (table 1). Most participants (77%) had received 2 doses
of measles vaccine. The second dose was received in 1971, a
mean of 26 years before collection of the follow-up blood sam-
ple; most 2-dose recipients (75%) had received the second dose
of vaccine 26–33 years previously. Eighteen percent of partic-
ipants received only a single dose and were vaccinated a mean
of 32 years before the follow-up blood sample was obtained;
the single dose of vaccine was received between 28 and 36 years
previously. The third dose was received a mean of 18 years (11,
16, and 27 years) before collection of follow-up blood.
All participants had documentation of a baseline HI antibody
titer in 1971 after initial vaccination. Mean age at collection of
by guest on November 11, 2015http://jid.oxfordjournals.org/Downloaded from
S126 JID 2004:189 (Suppl 1) Dine et al.
Table 1. Characteristics of study participants (1997–1999) and nonpartici-
pants (1971).
Characteristic Participants Nonparticipants
Total population 56 104
Age, years
Mean 34
a
32
Range 29–40 27–39
Sex, %
Female 71
a
47
Male 29 53
Vaccine group, %
A1827
B7771
C52
Vaccination age, mean (range)
First dose, months 19 (5–70)
a
14 (4–12)
1-dose recipient 28 (5–70) 18 (7–73)
2-dose recipient 17 (5–63) 12 (4–56)
Second dose, years 8 (2–18)
a
6 (0–20)
Third dose, years 17 (7–25) 6 (1–8)
Years since vaccination, mean (range)
First dose, months 33 (28–36) 31 (26–34)
1-dose recipient 32 (28–36) 31 (27–34)
2-dose recipient 33 (28–36) 31 (26–34)
Second dose, years 26 (12–33) 25 (10–32)
Third dose, years 18 (11–27) 28 (26–32)
Baseline HI titer,
b
%
5–20 59 55
40 21 16
140 20 29
Age at baseline titer determination, years
Mean 7 5
Range 2–13 0–12
Time since baseline HI titer determined, years
Mean 27 27
Range 26–28 26–30
NOTE. HI, hemagglutination inhibition.
a
Statistically significant difference between participants and nonparticipants; .P ! .05
b
Reciprocal hemagglutination inhibition antibody titer.
the baseline blood sample for HI assay was 7 years (table 1).
Thirty-three participants (59%) had HI antibody reciprocal ti-
ters between 5 and 20; the remaining 23 (41%) had titers of
40 (12 participants) or
140 (11 participants). Only 15 (27%)
of the 56 participants also had a second titer measured 2 to 8
weeks after the second dose of vaccine. Of the 15, 11 (73%)
had a titer of 40 after the second dose, whereas 4 (27%) had
a titer of between 5 and 20. The baseline HI titer was obtained
a mean of 27 years before the 1997–1999 follow-up blood sam-
ple was obtained.
Prevalence of vaccine-induced antibody. The prevalence
of measles antibody was very high a mean of 26–33 years after
1 or more doses of measles vaccine. All 56 participants had
PRN antibody detected (PRN antibody titer of 1:8), and
nearly all (92%) had a positive EIA antibody test result. Fifty-
one of the participants had a reciprocal PRN titer of
1120; only
5 (prevalence, 9%; 95% confidence interval, 3%–20%) had a
PRN titer not considered to be protective (120) (table 2).
To identify risk factors for measles susceptibility, we exam-
ined the study participants’ characteristics relative to a low
reciprocal PRN titer of 120 and a higher PRN titer of
1120.
We found slightly more women than men with a PRN titer of
by guest on November 11, 2015http://jid.oxfordjournals.org/Downloaded from
Duration of Measles Vaccine–Induced Antibody JID 2004:189 (Suppl 1) S127
Table 2. Characteristics of study population by plaque reduction neutralizing (PRN)
antibody titer, 1997–1999.
Characteristic
Total
population
PRN antibody titer
a
120 1120
Total population 56 5 51
Age, years
Mean 54 33 34
Range 29–36 29–40
Sex, %
Female 39 10 90
Male 16 0 100
Vaccine group, %
A 10 0 100
B431288
C 3 0 100
Vaccination age, mean (range)
First dose, months 56 12 (10–14) 20 (5–70)
1-dose recipient 10 28 (5–70)
2-dose recipient 46 12 (10–14) 18 (5–63)
Second dose, years 43 6 (3–8) 8 (2–18)
Third dose, years 3 17 (7–25)
Years since vaccination, mean (range)
First dose, months 54 32 (28–34) 33 (28–36)
1-dose recipient 9 32 (28–36)
2-dose recipient 45 32 (28–34) 33 (28–36)
Second dose, years 42 27 (26–28) 26 (12–33)
Third dose, years 3 18 (11–27)
Baseline HI titer,
a
%
5–20 33 15 85
40 12 0 100
140 11 0 100
Age at baseline titer determination, years 56
Mean 6 7
Range 3–8 2–13
Time since baseline HI titer determination, years 56
Mean 27 27
Range 26–28 26–28
NOTE. HI, hemagglutination inhibition; PRN, plaque reduction neutralization. No statistical differences
( ) were seen by characteristics relative to PRN antibody titer. For categorical variables, differencesP
! .05
were tested between levels of characteristic, and for continuous variables, differences were tested be-
tween levels of PRN antibody titer.
a
Reciprocal antibody titer.
120, but the difference was not statistically significant (table
2). Although more group B participants and participants with
baseline HI titers of 5–20 had PRN antibody titers of 120
than did other groups or those with higher baseline HI titers,
the differences were not statistically significant. In addition, the
distribution of continuous variables (i.e., age at and years since
vaccination and age at and years since baseline HI titer deter-
mination) by PRN antibody was not statistically different, al-
though the mean age at first and second dose was younger
among participants with a PRN antibody titer of 120 than
among those with a higher antibody titer.
We further examined the relationship between the baseline
HI antibody titer and the follow-up PRN antibody titer and
found a strong, positive relationship of grouped HI antibody
titers (i.e., 5–20, 40, and
140) and geometric mean PRN an-
tibody titer ( ) (figure 1). The higher the baseline HIP
! .001
by guest on November 11, 2015http://jid.oxfordjournals.org/Downloaded from
S128 JID 2004:189 (Suppl 1) Dine et al.
Figure 1. Relationship of baseline hemagglutination inhibition (HI) titer and follow-up geometric mean (GMT) plaque reduction neutralization (PRN)
antibody titer on average 26–33 years after vaccination. Titers are reciprocal antibody titers. CI, confidence interval.
titer, the higher the mean PRN titer of the follow-up serum
samples. When we further classified participants into 3 groups
on the basis of their PRN antibody titer to distinguish very high
antibody titer (
1900) from low (120) and moderate (121–900)
titers, we found baseline HI titers to predict PRN titers and also
to be statistically significant ( ) (table 3).P
! .01
Comparing participant characteristics relative to the 3-level
classification of PRN titers showed no other statistically sig-
nificant relationships by age at follow-up study, sex, age and
time since vaccination, and age and time since baseline HI titer
determination. Categorical analysis of age at vaccination and
years since first and second of measles vaccine did not show
statistically significant results (table 3). Although relatively more
recipients with a first dose at
!12 months of age and those
vaccinated
126 or 130 years after the second dose had a PRN
titer of 120 than those vaccinated older or sooner, these pro-
portions were not statistically different.
DISCUSSION
This is the first study in the United States to examine persistence
of vaccine-induced antibody an average of 25 years after vac-
cination [22–25]. All 56 participants had detectable PRN an-
tibody and nearly all (92%) had PRN titers considered to be
protective (PRN antibody titer of
11:120), despite no known
exposure to the wild measles virus or measles disease after
vaccination. Measles was not reported in any of the 56 vacci-
nated participants, although measles was circulating at low lev-
els in the United States (0.04 cases/100,000 population in 1999
to 23 cases/100,000 population in 1970) and very low levels in
Cincinnati (0–4 cases annually since 1985, the first year of
national electronic reporting).
If the PRN assay is 60-fold more sensitive than the HI assay
used in 1971 [17], we estimate that a decline in antibody titer
may have occurred over time in some study participants. The
lowest detectable antibody titer by HI is 1:5. This titer translates
to a PRN titer of 1:300. Of the 9 participants with a PRN titer
of
!1:300 at follow-up, all had baseline HI titers between 1:5
and 1:40 in 1971; most (7) had titers between 1:5 and 1:10.
Thus, in participants with the lowest titers in 1971, antibody
may have waned to levels not considered protective. However,
waning vaccine-induced antibody resulting in no likely pro-
tection is small. We would expect vaccine-induced antibody to
wane some, as we would expect antibody induced by disease
to wane [22–24]. However, waning antibody does not neces-
sarily mean that protection from measles also has waned, be-
cause cellular immunity may persist, resulting in a vigorous
secondary antibody response upon exposure to measles virus
[26, 27]. Moreover, a recent study suggests that one-half of
persons with a single dose of vaccine resulting in a PRN titer
of
!125 are protected from measles, although this study ex-
amined measles up to 11 years after vaccination [18]. The in-
teraction of humoral and cellular immunity needs to be eval-
uated to understand why low antibody titers persist even when
a person has cellular immunity and is found to be protected
from measles.
Presence of low vaccine-induced antibody (PRN antibody
titer of 120) after documentation of adequate HI antibody
following initial vaccination suggests that measles antibody for
a small number of participants may have decreased to titers
considered nonprotective. The 9% prevalence of low PRN an-
tibody titer found in the present study is consistent with the
secondary vaccine failure rate found in Canada (5%) [6] and
with a meta-analysis of 9 studies conducted by Anders et al.
[28] that found a secondary vaccine failure rate of 0% to 6%,
which includes the Canadian study. Anders et al. [28] pooled
data from studies conducted in North America, Taiwan, and
Japan. These studies include cohorts of healthy persons who
received a live measles-containing vaccine, had a blood test
demonstrating serological immunity after measles vaccination,
were exposed to measles, and developed disease at least 21 days
after vaccination, as judged on the basis of clinical or serological
by guest on November 11, 2015http://jid.oxfordjournals.org/Downloaded from
Duration of Measles Vaccine–Induced Antibody JID 2004:189 (Suppl 1) S129
Table 3. Risk factors for low measles antibody titer (PRN
antibody titer 1:120), 1997–1999.
Characteristic
PRN antibody titer
a
P120 121–900 1900
Total population 5 21 30
HI titer
a
5–20 5 18 10 !.01
40 0 2 10
140 0 1 10
Age at vaccination
Age at first dose
!12 months 4 11 17 .62
12 months 1 10 13
Age at second dose
6 years 3 6 4 .27
16 years 2 12 19
Years since vaccination
Years since first dose
30 1 3 5 .87
130 3 18 23
Years since second dose
26 years 0 1 5 .28
126 years 5 17 17
NOTE. HI, hemagglutination inhibition; PRN, plaque reduction
neutralization.
a
Reciprocal antibody titer.
data. These investigators found a 6% (95% confidence interval,
4%–8%) secondary vaccine failure rate for 3 studies (2 of which
were conducted outside the United States) with a total of 572
participants who had received a live measles vaccine. The age
at vaccination was unknown in 2 of the 3 studies; 12 months
was the reported age in the other study. Pooled results for the
remaining 6 studies (5 which were conducted in the United
States) found an even rarer secondary vaccine failure rate (0;
95% confidence interval, 0–0.15%) among
12000 participants.
For half of these studies, age at vaccination was available and
ranged from 48 to 56 months, which is older than ages in the
present study. The number of years since vaccination was not
available for any of the studies and was not reported in the
meta-analysis, in contrast to the present study, which estimates
prevalence of antibody an average of 26–33 years after
vaccination.
Before the present study, the longest follow-up study of per-
sistence of measles antibody after vaccination and disease was
conducted by Krugman et al. [22–24]. In follow-up studies that
examined antibody to measles as much as 16 years after wild
virus infection or vaccination, no evidence was found that pro-
tection against measles had waned to nonprotective levels from
serological results after revaccination, although antibody after
both natural infection (47 participants) or vaccination (70 par-
ticipants) decreased over time. Krugman found typical booster
responses after revaccination among persons with negative re-
sults of HI assay, suggesting persistence of immunologic mem-
ory. In addition, among 16 participants with low HI antibody
titers (1:2–1:4) 6–15 years after vaccination, PRN antibody
titers were higher (1:4–1:46), suggesting that actual titers may
be higher than measured by the HI assay. These results suggest
that protection does not seem to wane and also document a
higher sensitivity of the PRN assay than the HI assay in de-
tecting measles antibody.
A potential limitation of our study is the low participation
rate, which may be expected in a follow-up study nearly 3 dec-
ades after the initial study. Although participants were slightly
older, more likely to be female, and received the first or second
dose of measles vaccine at a slightly older age than nonpartici-
pants, the distribution of baseline HI titer was similar among
participants and nonparticipants. This is an important finding,
because the baseline HI titer was the most important predictor
of the later PRN titer, the basic measurement of measles an-
tibody in this study. All 5 participants with low PRN antibody
considered susceptible had low (10 or 20) baseline HI titers
nearly 3 decades before; all but 1 had 2 doses of measles vaccine.
Another potential limitation is that blood for most participants
(78%) was obtained before revaccination and not after. These
HI titers do not reflect the last titer after vaccination in 1971,
which would include boosting of antibody after revaccination.
However, several studies suggest that while boosting does occur,
it is only transient [25, 29, 30]. The longest prospective cohort
study of revaccinated children aged 4–6 years found boosting
to last just 6 months after revaccination [30]. Antibody titers
increased after vaccination for all participants but declined to
prevaccination levels for nearly all participants at 6 months.
Finally, because only 56 participants were examined, we did
not have sufficient power to detect small differences in age at
vaccination, years since vaccination, age at baseline HI titer
determination, or years since the baseline HI titer was deter-
mined. Our statistical power was only 2% to detect differences
of
145% between the group with low PRN antibody titer
(120) and the group with higher PRN antibody titer. Both
young age (
!12 months) at vaccination and many years since
vaccination have been found to predict low seroprevalence and
seroconversion rates [31]. In addition, higher attack rates dur-
ing outbreaks have been found among children vaccinated at
!12 months of age and in the more distant past [31].
This provider-based study does demonstrate that data in the
offices of physicians in private practice can be a valuable re-
source in answering medical questions of national importance,
such as the degree of persistence of vaccine-induced measles
antibody. Although the study population included adults who
were vaccinated with 1, 2, or 3 doses of measles vaccine many
years before, our results primarily reflect persistence of antibody
by guest on November 11, 2015http://jid.oxfordjournals.org/Downloaded from
S130 JID 2004:189 (Suppl 1) Dine et al.
among 2-dose recipients, among whom one-half had received
the first dose at age
!1 year. Our findings suggest that persons
who have received either a single dose or 2 doses of vaccine a
mean of 26–33 years before may still have detectable neutral-
izing antibody, and nearly all may still have titers considered
to be protective (
11:120). The persistence of antibody in this
diverse group of vaccine recipients supports the current elim-
ination strategy to achieve high population immunity with 2
doses of measles vaccine.
Acknowledgments
Special thanks to all of the study participants for their will-
ingness to participant and take time from their busy schedules
many decades after the initial study in 1971, and to the Visiting
Nurses Association for interviewing and collecting blood. In
addition, special thanks to Mary McCauley for editorial advice
and to Mark Papania and Jane Seward for scientific and edi-
torial assistance.
References
1. Hinman AR, Orenstein WA, Papania MJ. Evolution of measles elim-
ination strategies in the United States. J Infect Dis 2004; 189(Suppl 1):
17–22.
2. Katz SL, Hinman AR. Summary and conclusions: measles elimination
meeting, 16–17 March 2000. J Infect Dis 2004; 189(Suppl 1):43–7.
3. Centers for Disease Control and Prevention. Measles prevention.
MMWR Morb Mortal Wkly Rep 1989; 38(SS-9):1–18.
4. Markowitz LE, Preblud SR, Fine PE, Orenstein WA. Duration of live
measles vaccine-induced immunity. Pediatr Infect Dis J 1990; 9:101–10.
5. Hutchins SS, Bellini W, Coronado V, Jiles R, Wooten K, Deladisma A.
Population immunity to measles in the United States, 1999. J Infect
Dis 2004; 189(Suppl 1):91–7.
6. Mathias RG, Merrison WG, Arcand TA, et al. The role of secondary
vaccine failures in measles outbreaks. Am J Public Health 1989; 79:475–8.
7. Reyes MA, Franky DeBorrero M, Roa J, et al. Measles vaccine failure
after documented seroconversion. Pediatr Infect Dis J 1987; 6:848–51.
8. Mossong J, Nokes DJ, Edmunds WJ, et al. Modeling the impact of
subclinical measles transmission in vaccinated populations with waning
immunity. Am J Epidemiol 1999; 150:1238–49.
9. Centers for Disease Control and Prevention. Measles—United States,
1999. MMWR Morb Mortal Wkly Rep 2000; 49:557–60.
10. Atkinson W. Measles. In: Atkinson W, Murphy L, Humiston SG, et
al, eds. Epidemiology and prevention of vaccine-preventable diseases.
7th ed. Atlanta: Centers for Disease Control and Prevention, 2002.
11. Hutchins SS, Redd S, Schrag S. National serologic survey of measles
immunity among persons 6 years of age or older, 1988–1994. Med-
scape General Medical Journal 2001; 3:E5. Available at: http://www
.medscape.com.
12. Linneman CC Jr, Dine MS, Bloom JE, Schiff GM. Measles antibody
in previously immunized children. Am J Dis Child 1972; 124:53–7.
13. Albrecht P, Herrmann K, Burns GR. Role of virus strain in conventional
and enhanced measles plaque neutralization test. J Virol Methods 1981;
3:251–60.
14. Hummel KB, Erdman DD, Heath J, Bellini WJ. Baculovirus expression
of the nucleoprotein gene of measles virus and utility of the recom-
binant protein in diagnostic enzyme immunoassays. J Clin Microbiol
1992; 30:2874–80.
15. Ratnam S, Gadag V, West R, et al. Comparison of commercial enzyme
immunoassay kits with plaque reduction neutralization test for detec-
tion of measles virus antibody. J Clin Microbiol 1995; 33:811–5.
16. Chen RT, Markowitz LE, Albrecht P, et al. Measles antibody: reeval-
uation of protective titers. J Infect Dis 1990; 162:1036–42.
17. Brunell PA. Measles control in the 1990s: measles serology. Geneva:
World Health Organization, 1990; document no. WHO/EPI/Gen/90.4.
18. Samb B, Aaby P, Whittle HC, et al. Serologic status and measles attack
rates among vaccinated and unvaccinated children in rural Senegal.
Pediatr Infect Dis J 1995; 14:203–9.
19. De Serres G, Boulianne N, Ratnam S, et al. Effectiveness of vaccination
at 6 to 11 months of age during an outbreak of measles. Pediatrics 1996;
97:232–5.
20. Mehta C, Patel N. StatXact 3 for windows. Cambridge, MA: Cytel
Software Corporation, 1996.
21. Mehta C, Patel N. LogXact for windows. Cambridge, MA: Cytel Soft-
ware Corporation, 1996.
22. Krugman S, Giles JP, Friedman H, et al. Studies on immunity to mea-
sles. J Pediatr 1965; 66:471–88.
23. Krugman S. Present status of measles and rubella immunization in the
United States: a medical progress report. J Pediatr 1977; 90:1–12.
24. Krugman S. Further-attenuated measles vaccine: characteristics and
use. Rev Infect Dis 1983; 5:477–81.
25. Markowitz LE, Albrecht P, Orenstein WA, et al. Persistence of measles
antibody after revaccination. J Infect Dis 1992; 166:205–8.
26. Ward BJ, Boulianne N, Ratnam S, et al. Cellular immunity in measles
vaccine failure: demonstration of measles antigen—specific lympho-
proliferative responses despite limited serum antibody production after
revaccination. J Infect Dis 1995; 172:1591–5.
27. Bautista-Lopez N, Ward BJ, Mills E, et al. Development and durability
of measles antigen—specific lymphoproliferative response after MMR
vaccination. Vaccine 2000; 18:1393–401.
28. Anders JF, Jacobson RM, Poland GA. Secondary failure rates of measles
vaccines: a metaanalysis of published studies. Pediatr Infect Dis J
1996; 15:62–6.
29. Stetler HC, Orenstein WA, Bernier RH. Impact of revaccinating chil-
dren who initially received measles vaccine before 10 months. Pediatrics
1986; 77:471–6.
30. LeBaron C, Beeler J, Redd S, et al. Immunogenicity and adverse events
associated with the 2nd dose of MMR vaccine (abstract no. H78b). In:
Program and abstracts of the 36th Interscience Conference on Anti-
microbial Agents and Chemotherapy. Washington, DC: American So-
ciety for Microbiology, 1996.
31. Redd SC, Markowitz LE, Katz SL. Measles vaccine. In: Plotkin SA,
Orenstein WA, eds. Vaccines. Philadelphia: WB Saunders, 1999:222–66.
by guest on November 11, 2015http://jid.oxfordjournals.org/Downloaded from
    • "However, despite the vaccination programs adopted by many developed countries and advancements towards the goal of measles elimination, outbreaks continue to occur in Europe and more recently in the USA [1][2][3][4][5][6][7][8]. Lifelong immunity is generally reported after a wild-type measles infection, while in vaccinated people a waning immunity has been reported in correlation with lower levels or more rapid decrease of measles specific antibodies [9][10][11]. Asymptomatic reinfections or disease were observed in subjects where the immune system was challenged by 2-vaccine doses, probably because of an inadequate response. "
    [Show abstract] [Hide abstract] ABSTRACT: The objective of this study is primarily to compare the performance of the VIDAS(®) Measles immunoglobulin (Ig)G assay to that of two other serological assays using an immunoassay technique, Enzygnost(®) Anti-measles Virus/IgG (Siemens) and Measles IgG CAPTURE EIA(®) (Microimmune). The sensitivity and the agreement of the VIDAS(®) Measles IgG assay compared to the Enzygnost(®) Anti-measles Virus/IgG assay and the Measles IgG CAPTURE EIA(®) assay are 100%, 97.2% and 99.0%, 98.4%, respectively. The very low number of negative sera for IgG antibodies does not allow calculation of specificity. As a secondary objective, we have evaluated the ability of the VIDAS(®) Measles IgG assay to measure anti-measles virus IgG antibody avidity with the help of the VIDAS(®) CMV IgG Avidity reagent, using 76 sera from subjects with measles and 238 other sera. Different groups of populations were analyzed. In the primary infection measles group, the mean IgG avidity index was 0.16 (range of 0.07 to 0.93) compared to 0.79 (range of 0.25 to 1) in the serum group positive for IgG antibodies and negative for IgM. These data allow to define a weak anti-measles virus IgG antibody avidity as an avidity index (AI) < 0.3 and a strong avidity as an AI > 0.6. The VIDAS(®) Measles IgG assay has a performance equivalent to that of other available products. Its use, individual and quick, is well adapted to testing for anti-measles immunity in exposed subjects.
    Full-text · Article · Aug 2016
    • "There was some discussion around these recommendations in the reviewed publications, although the necessity of vaccinating infants was accepted without question. Several authors demonstrated that vaccinated individuals have lower GMT compared with those acquiring immunity through natural infection [9,10,[15][16][17]19,[24][25][26]30,34,69,96,99]. There was insufficient information provided to determine if certain vaccination programmes resulted in a lower GMT. "
    [Show abstract] [Hide abstract] ABSTRACT: The determination of the seroprevalence of vaccine-preventable diseases is critical in monitoring the efficacy of vaccination programmes and to assess the gaps in population immunity but requires extensive organisation and is time and resource intensive. The results of the studies are frequently reported in peer-reviewed scientific, government and non-government publications. A review of scientific literature was undertaken to advise the development of WHO guidelines for the assessment of measles and rubella seroprevalence. A search of the National Library of Medicine’s PubMed online publications using key words of ‘measles’, ‘rubella’, combined with ‘serosurvey’, ‘seroprevalence’, ‘immunity’ and ‘population immunity’ was conducted. A total of 97 articles published between January 1998 and June 2014 were retrieved, 68 describing serosurveys for measles and 58 serosurveys for rubella, conducted in 37 and 36 different countries respectively. Only 13 (19%) and 8 (14%) respectively were UN classified “least developed countries”. The study sample varied markedly and included combinations of male and female infants, children, adolescents and adults. The study sizes also varied with 28% and 33% of measles and rubella studies respectively, having greater than 2000 participants. Microtitre plate enzyme immunoassays were used in 52 (76%) measles studies and 40 (69%) rubella studies. A total of 39 (57%) measles and 44 (76%) rubella studies reported quantitative test results. Seroprevalence ranged from 60.8% to 95.9% for measles and 53.0% to 99.3% for rubella studies. The review highlighted that infants lost maternally-acquired immunity within 9 months of birth and were unprotected until vaccination. Two groups at higher risk of infection were identified: young adults between the ages of 15 and 30 years and immigrants.
    Article · Jun 2016
    • "Despite the availability of a safe and effective vaccine, measles is one of the leading causes of child death worldwide, particularly in poor nutrition and inefficient health care systems settings [1, 2] . All measles virus isolates characterised to date belong to a single serotype and both natural infection and vaccination produce lifelong immunity against the virus [3, 4]. The increase in vaccination rates has led to a dramatic reduction in measles deaths: in the period between 2000 and 2013 a 75% reduction in measles deaths was recorded (from 544,200 to 145,700 deaths/year), with an estimated 15.6 million deaths prevented in this period [5]. "
    [Show abstract] [Hide abstract] ABSTRACT: Measles is a highly infectious disease caused by measles virus (MeV). Despite the availability of a safe and cost-effective vaccine, measles is one of the world-leading causes of death in young children. Within Europe, there is a target for eliminating endemic measles in 2015, with molecular epidemiology required on 80% of cases for inclusion/exclusion of outbreak transmission chains. Currently, MeV is genotyped on the basis of a 450 nucleotide region of the nucleoprotein gene (N-450) and the hemagglutinin gene (H). However, this is not sufficiently informative for distinguishing endemic from imported MeV. We have developed an amplicon-based method for obtaining whole genome sequences (WGS) using NGS or Sanger methodologies from cell culture isolates or oral fluid specimens, and have sequenced over 60 samples, including 42 from the 2012 outbreak in the UK.Overall, NGS coverage was over 90% for approximately 71% of the samples tested. Analysis of 32 WGS excluding 3' and 5' termini (WGS-t) obtained from the outbreak indicates that the single nucleotide difference found between the two major groups of N-450 sequences detected during the outbreak is most likely a result of stochastic viral mutation during endemic transmission rather than of multiple importation events: earlier strains appear to have evolved into two distinct strain clusters in 2013, one containing strains with both outbreak-associated N-450 sequences. Additionally, phylogenetic analysis of each genomic region of MeV for the strains in this study suggests that the most information is acquired from the non-coding region located between the matrix and fusion protein genes (M/F NCR) and the N-450 genotyping sequence, an observation supported by entropy analysis across genotypes.We suggest that both M/F NCR and WGS-t could be used to complement the information from classical epidemiology and N-450 sequencing to address specific questions in the context of measles elimination.
    Full-text · Article · Nov 2015
Show more