Content uploaded by Carole J Hickman
Author content
All content in this area was uploaded by Carole J Hickman
Content may be subject to copyright.
MAJOR ARTICLE
Mumps Antibody Levels Among Students Before
a Mumps Outbreak: In Search of a Correlate of
Immunity
Margaret M. Cortese,
1
Albert E. Barskey,
1
Gary E. Tegtmeier,
3
Cheryl Zhang,
4
Laurie Ngo,
4
Moe H. Kyaw,
1,
a
Andrew L. Baughman,
2
Jay E. Menitove,
3
Carole J. Hickman,
1
William J. Bellini,
1
Gustavo H. Dayan,
1,
a
Gail R. Hansen,
5,
a
and Steven Rubin
4
1
Division of Viral Diseases, National Center for Immunization and Respiratory Diseases, and
2
Center for Global Health, Centers for Disease Control and
Prevention, Atlanta, Georgia;
3
Community Blood Center of Greater Kansas City, Missouri;
4
Center for Biologics Evaluation and Research, Food and Drug
Administration, Bethesda, Maryland; and
5
Kansas Department of Health and Environment, Topeka
Background. In 2006, a mumps outbreak occurred on a university campus despite $ 95% coverage of students
with 2 doses of measles-mumps-rubella (MMR) vaccine. Using plasma samples from a blood drive held on campus
before identification of mumps cases, we compared vaccine-induced preoutbreak mumps antibody levels between
individuals who developed mumps (case patients) and those who did not develop mumps (nonpatients).
Methods. Preoutbreak samples were available from 11 case patients, 22 nonpatients who reported mumps
exposure but no mumps symptoms, and 103 nonpatients who reported no known exposure and no symptoms.
Antibody titers were measured by plaque reduction neutralization assay using Jeryl Lynn vaccine virus and the
outbreak virus Iowa-G/USA-06 and by enzyme immunoassay (EIA).
Results. Preoutbreak Jeryl Lynn virus neutralization titers were significantly lower among case patients than
unexposed nonpatients (P 5 .023), and EIA results were significantly lower among case patients than exposed
nonpatients (P 5 .007) and unexposed nonpatients (P 5 .009). Proportionately more case patients than exposed
nonpatients had a preoutbreak anti–Jeryl Lynn titer , 31 (64% vs 27%, respectively; P 5 .065), an anti–Iowa-G/
USA-06 titer , 8 (55% vs 14%; P 5 .033), and EIA index standard ratio , 1.40 (64% vs 9%; P 5 .002) and , 1.71
(73% vs 14%, P 5 .001).
Discussion. Case patients generally had lower preoutbreak mumps antibody levels than nonpatients. However,
titers overlapped and no cutoff points separated all mumps case patients from all nonpatients.
In 2006, a multistate outbreak of mumps occurred in
theUnitedStateswithover6500reportedcases.This
was the largest outbreak in the United States in 19 years
[1, 2]. The greatest number of cases occurred among
college-aged persons, and most persons with mumps
had received 2 doses of measles-mumps-rubella (MMR)
vaccine as currently recommended by the Advisory
Committee on Immunization Practices [3]. Waning
vaccine-induced immunity may have played a role in
the resurgence [1, 2]. The occurrence of a blood drive
held shortly before identifica tion of mumps cases
at a Kansas university provided a key opportunity to
assess the relationship between levels of vaccine-induced
neutralizing antibody and protection from mumps,
which currently is not known. At the university there
were a total of 174 persons reported to have mumps
(attack rate, 0.9% among 19 155 undergraduate
students) despite $ 95% 2-dose MMR vaccine cover-
age among students [4, 5]. Mumps cases were also re-
ported among young adults in other Kansas locations.
Using plasma samples remaining from blood dona-
tions, we sought to compare vaccine-induced preout-
break mumps antibody levels between individuals with
Received 12 March 2011; accepted 7 June 2011.
a
Present affiliations: Novartis (M. H. K.); Sanofi Pasteur (G. H. D.); Pew
Charitable Trusts (G. R. H.).
Correspondence: Margaret M. Cortese, Centers for Disease Control and
Prevention, 1600 Clifton Rd, MSA34, Atlanta, GA 30333 (mcortese@cdc.gov).
The Journal of Infectious Diseases 2011;204:1413–22
Published by Oxford University Press on behalf of the Infectious Diseases Society of
America 2011.
0022-1899 (print)/1537-6613 (online)/2011/2049-0015$14.00
DOI: 10.1093/infdis/jir526
Mumps Antibody Levels Before an Outbreak
d
JID 2011:204 (1 November)
d
1413
at CDC Public Health Library & Information Center on October 29, 2013http://jid.oxfordjournals.org/Downloaded from at CDC Public Health Library & Information Center on October 29, 2013http://jid.oxfordjournals.org/Downloaded from at CDC Public Health Library & Information Center on October 29, 2013http://jid.oxfordjournals.org/Downloaded from at CDC Public Health Library & Information Center on October 29, 2013http://jid.oxfordjournals.org/Downloaded from at CDC Public Health Library & Information Center on October 29, 2013http://jid.oxfordjournals.org/Downloaded from at CDC Public Health Library & Information Center on October 29, 2013http://jid.oxfordjournals.org/Downloaded from at CDC Public Health Library & Information Center on October 29, 2013http://jid.oxfordjournals.org/Downloaded from at CDC Public Health Library & Information Center on October 29, 2013http://jid.oxfordjournals.org/Downloaded from at CDC Public Health Library & Information Center on October 29, 2013http://jid.oxfordjournals.org/Downloaded from at CDC Public Health Library & Information Center on October 29, 2013http://jid.oxfordjournals.org/Downloaded from
mumps (case patients) and those without mumps (non-
patients) to determine a correlate of immunity. If such a cor-
relate could be identified, it may suggest there would be benefit
in providing additional mumps vaccine doses to raise an
antibody level above the threshold, in persons at risk of mumps
(eg, young adults in crowded settings). We also sought to
estimate the mumps infection rate at the university by the
proportion of donors with an increase in mumps antibody
titer during the outbreak.
METHODS
Blood Donors
A blood drive had been held by the Community Blood
Center (CBC) and the American Red Cross (ARC) at the
university 6 210 March 2006. A mumps outbr eak was iden-
tifiedon30March2006[5] and peaked in mid-April 2006.
In June 2006, investigators co ntacted t he blood centers to
obtain excess samples from the drives. Samples were frozen
in storage at CBC only. Blood drives were subsequently held
at the univers ity in September 2006 and March 2007,
approximately 5.5 and 11 months after the outbreak
peak. Investigators attended the March 2007 drive a nd en-
roll ed interested donors in-person into the study. Donors
were asked for permission via written informed consent to
test leftover plasma samples for mumps antibodies, to
obtain their MMR vaccination records, and to obtain their
medical records if they had sought medical care for
symptoms possibly consistent with mumps. Enrollees were
also asked to complete a writt en questionnaire that queried
demographic factors, student status, living arrangements,
history of mumps symptoms (including parotid swelling,
parotid pain, jaw pain, swelling below the jaw, and
testicular pain or swelling) during the period 1 January 2006
to date of questionnaire completion, date of symptom onset,
mumps exposure during 200622007, and history of ever
having physician- diagnosed mumps. Pictures showing
location of parotid glands and parotid enlargement and
a calendar of campus events were included. For assessi ng
mumps exposure, donors were asked if they had been face-
to-face (‘‘as close as if you were talking with someone right in
front of you’’) with someone ill with mumps who was not
wearing a mask.
Some persons who had not participated in the March 2007
drive but who had donated blood in 2006 before the outbreak
were contacted by email, letter, or telephone to request study
participation. These included reported mumps patients, room-
mates/housemates of mumps patients [4], and those with pre-
and postoutbreak samples available (‘‘repeat donors’’). Samples
were tested for mumps virus–specific antibody by plaque-
reduction neutralization (PRN) assay and a commercially avail-
able enzyme immunoassay (EIA).
Enrollees were asked if they desired their mumps EIA result
and those who did were notified by letter. Those who tested
negative were advised to discuss the results with their physician
and consider revaccination with MMR. The study was reviewed
by the institutional review boards of the Centers for Disease
Control and Prevention, CBC, ARC, and the Kansas Department
of Health and Environment (KDHE).
Donor Categories
Case Patients
Case patients were donors who had been reported by a clinician
to KDHE as having a case of mumps and who, upon review of
available records, had an illness suggestive of mumps and who
had a plasma sample obtained at least 25 days (1 incubation
period) before illness onset.
Nonpatients
Nonpatients were students at the university during the outbreak
who were not reported to have mumps and who completed
a questionnaire and did not report having symptoms suggestive
of mumps between 1 January 2006 and the date of questionnaire
completion. Although jaw pain may be a symptom of mumps,
persons meeting the nonpatient criteria who reported having
only jaw pain sometime during 2006–2007 but not specifically
during the outbreak period (March 2006–August 2006) were
considered a nonpatient.
Nonpatients were classified into 2 groups. Group 1 included
those likely exposed to mumps, defined either as being
a roommate/housemate of a person with mumps, reporting on
the questionnaire that they had face-to-face exposure to a person
with mumps during the outbreak, or repeat donors who re-
ported no mumps symptoms but who demonstrated a $ 4-fold
rise in neutralizing antibody titer against either of the mumps
virus strains tested (Jeryl Lynn vaccine or Iowa-G/USA06).
Group 2 nonpatients were those who reported no known
mumps exposure during the outbreak (and, if a repeat donor,
did not have a $ 4-fold rise in antibody titer).
Some donors were excluded from analysis before results
were obtained because they may have been exposed to mumps
before their earliest sample. During the outbreak investigation,
5 mumps cases were retrospectively identified to have occurred
at the university before the blood drive [5]. We therefore
excluded donors if their reported mumps exposure occurred
before the donation date, if they had lived in the 1 dormitory
with a case in February 2006, or if they were a member of
the sorority or fraternity with a cas e in February–early March
2006.
Repeat donors were university students who donated a first
sample before 16 March 2006 and a second sample during May
20062May 2007, and who were not reported as having mumps.
Some were also in nonpatient groups 1 or 2. Those who may
have had mumps exposure before donating their first sample
were excluded.
1414
d
JID 2011:204 (1 November)
d
Cortese et al
Laboratory Testing
Plasma samples were kept at –70°C at the blood centers until
shipment to the US Food and Drug Administration and CDC.
Laboratory scientists were blinded to donor information, in-
cluding case status, until after testing was completed.
Neutralization Titers
PRN assay was used to determine virus neutralizing antibody
titer in plasma as previously described [6]. Because levels of
mumps virus neutralizing antibody titers detected in the PRN
are a function of the challenge virus strain used [6–9], 2
different PRN a ssays were used: 1 using the Jeryl Lynn strain
(the mumps virus component of the MMR vaccine) to
measure the concentration of antibody that had been induced
by immunization, and 1 using a contemporary wild-type
virus isolated from the 2006 US mumps outbreak (Iowa-G/
USA-06) [6] to approximate the potency o f vaccine-induced
antibody to the circulating wild-type strain. The neutralizing
antibody titer is reported as the sample dilution capable of
reducing the mean number of virus plaques by 50% or greater
compared with the mean number of plaques in virus control
wells using the Karber formula [10].
Enzyme Immunoassay
An indirect enzyme-linked immunosorbent immunoglobulin
G (IgG) assay (Mumps IgG ELISA II, Wampole Laboratories)
was used to determine mumps IgG antibody levels in plasma
specimens, as described by the manufacturer. In this assay,
index standard ratio (ISR) values , 0.90 are seronegative,
values 0.91–1.09 are indeterminate, and values .1.10 a re
seropositive. All samples with seronegative or indeterminate
results were retested along with an equal number of randomly
selected positive specimens. The original ISR result was used
in analysis if the result categories from the original and repeat
runwerethesame,andthemeanISRresultfromthe2runs
was used if the original and repeat categories were different.
Statistical Analyses
The dist ributions of preoutbreak neutralization titers and
EIA ISRs were compared between case patients and non-
patient groups using the Kolmogorov-Smirnov test. The
preoutbreak geometric mean (with 95% confidence interval
[CI]) for case patients and nonpatients was calculated for
each of the 3 assays. F or each assay, cutoff points for pre-
outbreak samples were assessed for their abili ty to discrim-
inate case pa tients from nonpatients. Cuto ff points were
defined by the midpoint between a case patient’s value and
the next highest value among the group 1 (‘‘exposed’’)
nonpatients. The proportions of persons in each group with
preoutbreak values below the cutoff points were compared
using Fisher exact test, and odds ratios (ORs) with 95% CI
were calculated. For all a nalyses, a P value , .05 was con-
sidered statistically significant. Analyses were performed
using St ata software version 11.0.
RESULTS
Case Patients and Nonpatients
Eleven individuals met study criteria defining a case patient
(Table 1). Three were laboratory-confirmed: patient 1 by virus
isolation, patient 6 by reverse-transcription polymerase chain
reaction [11], and patient 3 by a $ 4-fold rise in neutralizing
antibody titer between pre- and postillness plasma samples.
Oral swabs from the other 8 individuals were not obtained for
mumps virus testing, likely because the KDHE laboratory had
recommended against continued testing at confirmed outbreak
locations [5]. Two of these 8 individuals (patients 9 and 11) had
postillness plasma samples available. All 11 case patients
had received 2 MMR doses before their preoutbreak sample
(Table 1).
Twenty-two persons met criteria for a group 1 (exposed)
nonpatient. Three were roommates of persons with mumps and 2
others were selected into the group because they reported no
mumps symptoms but had a $ 4-fold rise in antibody titers.
Twelve (55%) group 1 nonpatients were repeat donors
(Figure 1). A total of 103 persons met criteria for a group 2
(nonexposed) nonpatient, and 58 (56%) of these were repeat
donors (Figure 1). Immunization records were available on 119
(95%) of the 125 nonpatients; 118 (99%) of these had received
$ 2 MMR doses before their preoutbreak sample and 1 (1%)
had received 1 dose. No case patients or nonpatients reported
a history of ever having physician-diagnosed mumps before
enrollment at the university.
Anti–Jeryl Lynn Vaccine Virus Neutralizing Antibody Titers
Among Case Patients Versus Nonpatients
Among case patients, the median preillness neutralization anti-
body titer against the Jeryl Lynn vaccine virus was 23.2 (Table 2).
The 3 case patients with laboratory-confirmed mumps had
preillness titers in the lower range (8.2, 9.5, and 23.2; Table 3). Of
the 3 case patients with a postillness sample available, 1 had
a 506-fold increase 2 months after illness, 1 had a nominal in-
crease 5 months later, and 1 had a 3.9-fold drop in titer 1.1
months after illness (Table 1).
The median preoutbreak anti–Jeryl Lynn neutralization an-
tibody titer value among group 1 nonpatients was 73.4 and the
median among group 2 nonpatients was 57.6 (Table 2). The
2 persons who reported no symptoms but had a $ 4-fold rise in
titer had preoutbreak titers of 70.8 and 76.0 (with 13.5- and 4.5-
fold rises, respectively) (Table 3).
Preoutbreak Jeryl Lynn titers among case patients were sig-
nificantly different from those of group 2 nonpatients (P 5 .023)
but not group 1 nonpatients (P 5 .12; Table 2; Figure 2).
Compared to group 1 nonpatients, case patients had 4.7 times
greater odds of having a preoutbreak Jeryl Lynn titer , 31 and
5.7 times greater odds of having a titer , 41 (Table 4). Com-
pared to group 2 nonpatients, case patients had a statistically
Mumps Antibody Levels Before an Outbreak
d
JID 2011:204 (1 November)
d
1415
Table 1. Clinical Characteristics and Mumps Antibody Levels in 11 Mumps Patients
Patient
no.
Age at
illness
onset, y Sex
Student at
the
university
History of
possible
exposure
Physician
impression Physician exam
Patient reported
to CDC staff Sample
Years
from
MMR2 to
sample
Months
from
presample
to
illness
onset
Months
from
illness
onset
to post
sample
Anti-Jeryl
Lynn
neutralization
titer
Anti-Iowa
G/USA-06
neutralization
titer
EIA
result
EIA
ISR
1 18.1 M No Per student, friend’s
sibling had mumps
Parotitis,
lymphadenitis
Parotid swelling
mild moderate
a
Parotitis, jaw pain
b
pre 12.6 4.1 8.2 5.3 pos 1.14
2 20.3 M Yes Mumps Parotid swelling Parotitis, jaw pain,
testicular swelling,
testicular pain
pre 14.5 1.3 8.8 6.0 neg 0.82
3 19.3 M Yes Wrestling opponent
developed mumps
shortly after match
Swollen glands,
possible mumps
Submandibular
swelling
Parotitis and
parotid pain,
jaw pain
b
pre 13.9 1.3 9.5 7.5 neg 0.76
post 2.0 4812.3 1868.2 pos 5.32
4 18.4 M No Visited the university
several weeks
before illness
NA (reported
to health
dept as probable
mumps)
NA Parotitis
c
pre NA
d
1.0 14.9 3.9 pos 1.16
5 21.9 M Yes Roommate reported
to health dept as
mumps case 2
weeks earlier
Probable mumps Submandibular
gland swelling
Parotitis and
parotid pain
b
pre 9.7 6.0 17.7 7.2 pos 1.35
6 21.0 M Yes Mumps Parotid swelling
e
Parotitis, jaw pain pre 13.3 2.0 23.2 5.8 neg 0.83
7 18.7 M No Lived in the
city where
university is
located
Mumps Parotid swelling Parotitis, jaw pain
b
pre 12.6 4.8 27.7 30.1 neg 0.66
8 18.3 F No NA (reported to
health dept as
probable mumps)
Jawline swelling
per nurse’s notes
Parotitis and
parotid pain,
jaw pain,
facial swelling
b
pre 12.9 2.1 36.8 19.8 pos 2.51
9 18.9 F Yes Mumps Parotid tenderness
only
Swelling below
jaw only
pre 13.6 1.4 76.0 19.4 pos 1.67
post 5.3 92.0 34.4 pos 1.93
10 17.7 F No Outbreak at
high school
Mumps Parotid swelling Parotitis, jaw pain
b
pre 12.6 1.2 472.5 152.3 pos 3.74
11 20.0 M Yes Probable mumps Parotid swelling,
mild
Parotitis, jaw pain,
testicular pain
pre 14.9 1.8 495.7 84.3 pos 2.96
post 1.1 127.1 70.9 pos 2.76
post 10.3 147.9 86.9 pos 2.80
Patients are ordered in table by anti-Jeryl Lynn neutralization titer in preillness sample.
Abbreviations: CDC, Centers for Disease Control and Prevention; EIA, enzyme immunoassay; ISR, index standard ratio; MMR2, second dose of measles-mumps-rubella vaccine; NA, not available; neg, negative; pos,
positive; post, post-illness; pre, preillness.
a
Mumps virus isolated from throat washing.
b
Per information provided by patient in 2007. All cases reported in 2006 to health department as having parotitis.
c
Per health department interview.
d
Exact date of MMR2 not available.
e
Mumps virus RNA detected from buccal swab by reverse-transcription polymerase chain reaction.
1416
d
JID 2011:204 (1 November)
d
Cortese et al
significant 6.4 times and 4.8 times greater odds, respectively, of
having a titer below those cutpoints.
Anti–Iowa-G/USA-06 Virus Neutralizing Antibody Titers Among
Case Patients Versus Nonpatients
Among case patients, the median preillness neutralization an-
tibody titer against the Iowa-G/USA-06 virus was 7.5 (Table 2).
The 3 case patients with laboratory-confirmed mumps had
preillness titers in the lower range (5.3, 5.8, and 7.5). Of the 3
case patients with postillness samples, only 1 had a $ 4-fold rise
in titer (Table 1).
The median preoutbreak Iowa-G/USA-06 titer was 20.3
among the group 1 nonpatients and 19.5 among the group 2
nonpatients. The 2 nonpatients demonstrating $ 4-fold rise had
preoutbreak titers of 28.5 and 37.0 (with a 7.9- and 8.2-fold rise,
respectively).
The preoutbreak anti-Iowa-G/USA-06 virus neutralizing an-
tibody titers among case patients were not significantly different
from those of group 1 (P 5 .12) or group 2 nonpatients
(P 5 .13) (Table 2; Figure 2). Case patients had a statistically
significant 7.6 times greater odds of having a preoutbreak anti–
Iowa-G/USA-06 virus titer , 8 compared with group 1
Figure 1. Donor categories among students at the university. Number in parentheses at bottom indicate number of persons with a . 4-fold rise in
neutralizing antibody titer.
a
The sixth mumps patient from the university donated a preillness sample after 16 March 2006.
b
See page 2 of the text.
Number in parentheses at bottom indicate number of persons with a $ 4-fold rise in neutralizing antibody titer.
Mumps Antibody Levels Before an Outbreak
d
JID 2011:204 (1 November)
d
1417
nonpatients and 4.7 times greater odds compared with group 2
nonpatients (Table 4).
EIA ISR Results Among Case Patients Versus Nonpatients
Preoutbreak EIA ISR values among case patients differed sig-
nificantly from those of group 1 (P 5 .007) and group 2
(P 5 .009) nonpatients (Table 2; Figure 2). Thirty-six percent of
case patients (including 2 of 3 with laboratory-confirmed
mumps) were EIA-negative or indeterminant preoutbreak,
compared with 5% of group 1 nonpatients and 9% of group 2
nonpatients. Three EIA cutoffs (1.24, 1.40, and 1.71) discrimi-
nated case patients from nonpatients (eg, case patients had
a statistically significant 17.5 times greater odds of having
a preoutbreak EIA ISR , 1.40 compared with group 1 non-
patients, and 8.3 times greater odds compared with group 2
nonpatients (Table 4).
Table 2. Preoutbreak Mumps Neutralization Antibody Titers and EIA ISR Results of Case Patients and Nonpatients
Case patients (n 5 11) Group 1 nonpatients (n 5 22) P value
a
Group 2 nonpatients (n 5 103) P value
a
Jeryl Lynn neutralization antibody titer
Median titer (IQR) 23.2 (9.5–76.0) 73.4 (22.5–124.2)
b
.12 57.6 (32.5–118.1) .023
GMT (95% CI) 34.3 (12.8–91.8) 59.6 (38.8–91.8) 59.6 (48.5–73.1)
Iowa-G/USA-06 neutralization antibody titer
Median titer (IQR) 7.5 (5.8–30.1) 20.3 (12.2–58.2)
b
.12 19.5 (8.7–41.8) .13
GMT (95% CI) 14.5 (6.4–32.6) 22.1 (14.7–33.1) 21.0 (17.3–25.5)
EIA ISR
Median (IQR) 1.16 (0.82–2.51) 2.40 (1.81–3.54)
b
.007 2.58 (1.70–3.54) .009
Geometric mean (95% CI) 1.36 (.91–2.01) 2.52 (2.06–3.08) 2.41 (2.18–2.65)
Abbreviations: CI, confidence interval; EIA, enzyme immunoassay; GMT, geometric mean titer; IQR, interquartile range; ISR, index standard ratio.
a
Exact P value by Kolmogorov–Smirnov test for nonpatient group compared with case patients.
b
The 3 roommates/housemates of persons with mumps had Jeryl Lynn neutralization antibody titers, Iowa-G/USA-06 neutralization antibody titers, and EIA ISR
results, respectively, as follows: roommate A: 11.8 (13.9 6.2 months after reported exposure), 10.7 (13.1 6.2 months after reported exposure) and 2.61; roommate
B (to a polymerase chain reaction–positive case patient): 18.3, 6.6, and 2.09; roommate C: 261.8, 78.2, and 4.04.
Table 3. Comparison of Preoutbreak Mumps Neutralization Antibody Titers and EIA ISR Results Between case Patients and Nonpatients
Case patients
(n 5 11)
Group 1
nonpatients
(n 5 22)
% (n)
Case patients vs group 1
nonpatients
Group 2
nonpatients
(n 5 103)
% (n)
Case patients vs group 2
nonpatients
% (n) OR (95% CI) P value
a
OR (95% CI) P value
a
Jeryl Lynn neutralization antibody titer
Proportion (n) below this level
,16 36.4 (4) 9.1 (2) 5.7 (.6–71.8) .15 8.7 (9) 6.0 (1.05–28.8) .022
,19 45.4 (5) 22.7 (5) 2.8 (.5–17.3) .24 12.6 (13) 5.8 (1.2–26.0) .014
,31 63.6 (7) 27.3 (6) 4.7 (.8–29.3) .065 21.4 (22) 6.4 (1.5–32.2) .006
,41 72.7 (8) 31.8 (7) 5.7 (.9–41.6) .061 35.9 (37) 4.8 (1.05–29.1) .024
,83 81.8 (9) 54.5 (12) 3.8 (.6–42.0) .25 62.1 (64) 2.7 (.5–27.2) .32
Iowa-G/USA-06 neutralization antibody titer
Proportion (n) below this level
,7 36.4 (4) 9.1 (2) 5.7 (.6–71.8) .15 12.6 (13) 4.0 (.7–18.0) .058
,8 54.5 (6) 13.6 (3) 7.6 (1.1–60.5) .033 20.4 (21) 4.7 (1.06–21.1) .020
,21 72.7 (8) 50.0 (11) 2.7 (.5–19.3) .28 51.5 (53) 2.5 (.6–15.4) .22
,34 81.8 (9) 63.6 (14) 2.6 (.4–29.5) .43 68.0 (70) 2.1 (.4–21.1) .5
EIA ISR
Proportion (n) below this level
,1.10 (nonpositive) 36.4 (4) 4.6 (1) 12.0 (.9–618.1) .032 8.7 (9) 6.0 (1.05–28.8) .022
,1.24 54.5 (6) 4.6 (1) 25.2 (2.0–1212) .002 13.6 (14) 7.6 (1.6–35.4) .004
,1.40 63.6 (7) 9.1 (2) 17.5 (2.0–207.1) .002 17.5 (18) 8.3 (1.8–41.6) .002
,1.71 72.7 (8) 13.6 (3) 16.9 (2.2–148.9) .001 25.2 (26) 7.9 (1.7–48.6) .003
,2.56 81.8 (9) 54.5 (12) 3.8 (.6–42.0) .25 49.5 (51) 4.6 (.9–45.1) .057
Abbreviations: CI, confidence interval; EIA, enzyme immunoassay; ISR, index standard ratio; OR, odds ratio.
a
Comparison of the proportions below the cutoff point by Fisher exact test.
1418
d
JID 2011:204 (1 November)
d
Cortese et al
Figure 2. Preoutbreak antibody levels by patient category. Line indicates a cutpoint value presented in Table 3. Top panel: Anti–Jeryl Lynn virus neutralizing
antibody titers. Middle panel: Anti–Iowa-G/USA-06 virus neutralizing antibody titers. Bottom panel: Enyzme immunoassay index standard ratio (EIA ISR) results.
Mumps Antibody Levels Before an Outbreak
d
JID 2011:204 (1 November)
d
1419
Increases in Neutralizing Antibody Titers Among Repeat Donors
A total of 113 persons were repeat donors (Figure 1). Of the 31
who had a first postoutbreak sample obtained during May
20062July 2006, 2 (6.5%) had a $ 4-fold rise in Jeryl Lynn or
Iowa-G/USA-06 titers. One was a group 1 nonpatient described
earlier and the other did not complete the symptom question-
naire so symptom information was unavailable (preoutbreak
Jeryl Lynn titer, 39.9; Iowa-G/USA-06 titer, 8.0). Of the 45
students with a first postoutbreak sample obtained in September
2006, 3 (6.7%) had a $ 4-fold rise in titer. One was a group 1
nonpatient described earlier, 1 did not complete the question-
naire so symptom information was unavailable (preoutbreak
Jeryl Lynn titer, 134.1; Iowa-G/USA-06 titer, 34.1), and 1 re-
ported mumps exposure during the outbreak and developed jaw
pain (preoutbreak Jeryl Lynn titer, 30.1; Iowa-G/USA-06 titer,
12.2; values approximating the cutoffs used in previous analysis)
(Table 3). Overall, of the 47 persons who donated their post
sample during May 20062September 2006 and completed the
symptom questionnaire, 3 (6.4%) had the specified titer rise.
None of the 37 students whose first postoutbreak sample was
obtained February 20072May 2007 had a $ 4-fold rise. One
student with a March 2008 postoutbreak sample had a 5.3-fold
rise in Jeryl Lynn titer (preoutbreak Jeryl Lynn titer, 57.1; Iowa-
G/USA-06, 18.1; EIA ISR, 2.21). He had been a resident of
a dormitory where mumps cases were reported and in mid-
2008, he reported no mumps symptoms during the outbreak or
subsequently.
DISCUSSION
In our very highly vaccinated population of young adults, pre-
outbreak Jeryl Lynn vaccine virus neutralization titers were
lower among case patients than among grou p 2 nonpatients,
and EIA ISR results were lower among case patients than
among both nonpatient groups. We identified cutoff points in
each assay that statistically discriminated case patients from
nonpatients; however, titers overlapped and there were no
cutoff points that separated all case patients from all non-
patients. Of note, the data presented in this report were based
on testing of plasma samples. While it is likely that similar
results would be obtained had sera been available for testing,
differences cannot be ruled out.
Interestingly, the 1:8 cutoff identified in this study using
a contemporary wild-type isolate (Iowa-G/USA06) is similar to
that reported by others in studies conducted in the prevaccine
era [12–15]. No such studies have been conducted subsequently.
Notably, levels of virus neutralizing antibody titers measured in
vitro are dependent on the challenge virus strain used in the
assay [6–9], highlighting the importance of defining a serological
correlate of protection based on appropriate selection of the
challenge virus strain, among other considerations. In our study,
the minimal level of antibody correlating with protection is best
investigated using the vaccine virus strain, given that antibodies
detected in pre-exposure plasma are vaccine derived. Un-
fortunately, a robust correlate of protection using the Jeryl Lynn
virus could not be identified in our study. Despite the difficulty
in establishing a correlate of protection, anti-Iowa-G/USA06
neutralizing antibodies were present in preexposure plasma
obtained from all case patients, suggesting that susceptibility to
disease was not due to the inability of vaccine-induced antibody
to neutralize the outbreak strain (ie, immune escape). In addi-
tion, the detection of neutralizing antibodies in preexposure
plasma from all case patients suggests that their susceptibility to
mumps was unlikely to have been a result of primary vaccine
failure. It is important to acknowledge, however, that meas-
urements of virus neutralizing antibody in vitro may not be fully
predictive of immunological activity in vivo given that Fc-me-
diated phagocytosis, antibody-dependent cell-mediated cyto-
toxicity, and other processes that occur in the host are not
reflected in the assays used to measure virus viability in vitro.
Several aspects of our study limited our ability to identify an
antibody-based correlate of immunity, if one indeed exists. The
Table 4. Preoutbreak Mumps Antibody Levels in Persons With Laboratory-Confirmed Mumps Exposure
Jeryl Lynn
neutralization
antibody titer
Iowa-G/USA-06
neutralization
antibody titer EIA ISR
Cases of laboratory-confirmed mumps
Patient 1 8.2 5.3 1.14 pos
Patient 3 9.5 7.5 0.76 neg
Patient 6 23.2 5.8 0.83 neg
Nonpatient with $ 4-fold rise in titer and minimal
symptoms (jaw pain) during outbreak
30.1 12.2 2.38 pos
Nonpatients with $ 4-fold rise in titer and no symptoms 70.8 37.0 2.39 pos
76.0 28.5 1.33 pos
Abbreviations: pos, positive; neg, negative.
1420
d
JID 2011:204 (1 November)
d
Cortese et al
most significant limitation was that preexposure samples were
available from only 11 case patients. Given the small number of
case patients, misclassification of even a few nonpatients with
high preoutbreak titers into this group would impact our results.
We wanted to compare case patients to a group of nonpatients
who had truly been exposed to mumps virus. Short of demon-
strating an increase in antibody titer in ideally timed samples
(our convenience samples were not ideally timed), identifying
such persons is inherently difficult. The greater the enrichment
of our nonpatient groups with truly unexposed persons, the less
likely we would have been to identify an immune correlate.
Additionally, the epidemic curve indicates there was limited
mumps virus circulation before the first blood drive. It is
therefore possible that some samples from nonpatients that we
considered preexposure samples were actually postexposure
samples, inflating the differences between case patients and
nonpatients. To address this, we excluded donors if the
information available raised a possibility of mumps exposure
before the first donation.
It is possible that the level of immunity required to protect
against classic clinical mumps illness depends on the inoculum
of virus one is exposed to, so that protection at a particular
antibody titer is not absolute. This would also make identifying
a protective level more difficult. As in our university outbreak
overall [5] and other similar outbreaks [16, 17], most of the 11
case patients in our study could not identify a specific exposure.
This may indicate that transmission occurred from persons
asymptomatic at the time of exposure, and therefore quantifying
exposure intensity (as a marker for viral inoculum) would be
very difficult for most case patients in these outbreaks.
The infectiousness of asymptomatic or minimally symptom-
atic 2-dose vaccinees is not known, but it is possible such per-
sons play an important role in outbreaks in highly vaccinated
populations. Viral shedding from unvaccinated, minimally
symptomatic persons has been documented [15, 18]. Among
our repeat donors who did not develop clinical mumps, the
minimum mumps infection rate was 6.4% (3 of 47 repeat do-
nors who did not develop clinical mumps had a $ 4-fold rise in
titer). This is a minimum rate because most post samples were
from routine blood donations several months after the outbreak
peak and hence transient titer rises would have been missed.
Furthermore, it is not known if all mumps infections in 2-dose
vaccinees are associated with a $ 4-fold rise in antibody titer.
While we do not know how representative our blood donors are
of the entire student body with regard to mumps exposure, by
dividing this percentage (6.4%) of repeat donors who had a $ 4-
fold rise in titer but no or minimal mumps symptoms by the
percentage of March 2006 donors who developed clinical
mumps (1.6%; 5 of 313, Figure 1), we estimate that there may
have been at least 4 individuals with asymptomatic or minimally
symptomatic infections during the university outbreak for every
reported case. This proportion (4 of 5; 80%) of asymptomatic or
minimally symptomatic persons appears higher than the
20%–40% generally reported in the literature from the pre-
vaccine era [15, 19, 20], suggesting, perhaps not surprisingly,
that subclinical or mild infections may be more common in
vaccinated persons.
Identifying a vaccine-induced correlate of immunity for pro-
tection against mumps disease and infectiousness would be of
great benefit to immunization programs. We assessed only 1
component of the immune response to the Jeryl Lynn vaccine, and
while there are limitations to using samples obtained from blood
donations around a mumps outbreak, our results should be useful
for additional investigations to help identify the best correlate.
Notes
Acknowledgments. We are grateful to the students for participating in
the study. We thank the following individuals for their efforts in the in-
vestigation: Cherie Erikson (Community Blood Center of Greater Kansas
City), Dr Mei-chien Fucci (American Red Cross), Dr Charley LeBaron, Nobia
Williams, Robin Jones, Daisy Valentine, Susanna Haynes (CDC), Andrea
Gore, Sherie Rhine, and Joe Gillespie (Watkins Memorial Health Center).
Financial support. This work was supported by the Centers for Disease
Control and Prevention (contract 200-2008-M-27695).
Potential conflicts of interest. All authors: No reported conflicts.
All authors have submitted the ICMJE Form for Disclosure of Potential
Conflicts of Interest. Conflicts that the editors consider relevant to the
content of the manuscript have been disclosed.
The findings and conclusions of this article are those of the authors and
do not necessarily represent the views of the Centers for Disease Control
and Prevention.
At the time of his involvement in this investigation, Dr Kyaw was an
employee of the National Center for Immunization and Respiratory Dis-
eases, CDC. He is currently an employee of Novartis. At the time of his
involvement in this investigation, Dr Dayan was an employee of the Na-
tional Center for Immunization and Respiratory Diseases, CDC, and is
currently an employee of Sanofi Pasteur, which manufactures measles-
mumps-rubella vaccine in France. At the time of her involvement in this
investigation, Dr Hansen was an employee of the Kansas Department of
Health and Environment. She is currently an employee of The Pew
Charitable Trusts.
References
1. Dayan GH, Quinlisk MP, Parker AA, et al. Recent resurgence of
mumps in the United States. N Engl J Med 2008; 358:1580–9.
2. Barskey AE, Glasser JW, LeBaron CW. Mumps resurgences in the
United States: a historical perspective on unexpected elements. Vaccine
2009; 27:6186–95.
3. Watson JC, Hadler SC, Dykewicz CA, Reef S, Phillips L. Measles,
mumps, and rubella–vaccine use and strategies for elimination of
measles, rubella, and congenital rubella syndrome and control of
mumps: recommendations of the Advisory Committee on Immuni-
zation Practices (ACIP). MMWR Recomm Rep 1998; 47:1–57.
4. Cortese MM, Jordan HT, Curns AT, et al. Mumps vaccine performance
among university students during a mumps outbreak. Clin Infect Dis
2008; 46:1172–80.
5. Huang AS, Cortese MM, Curns AT, et al. Risk factors for mumps at
a university with a large mumps outbreak. Public Health Rep 2009;
124:419–26.
6. Rubin SA, Qi L, Audet SA, et al. Antibody induced by immunization
with the Jeryl Lynn mumps vaccine strain effectively neutralizes
a heterologous wild-type mumps virus associated with a large outbreak.
J Infect Dis 2008; 198:508–15.
Mumps Antibody Levels Before an Outbreak
d
JID 2011:204 (1 November)
d
1421
7. Orvell C, Tecle T, Johansson B, Saito H, Samuelson A. Antigenic re-
lationships between six genotypes of the small hydrophobic protein
gene of mumps virus. J Gen Virol 2002; 83:2489–96.
8. Yates PJ, Afzal MA, Minor PD. Antigenic and genetic variation of the
HN protein of mumps virus strains. J Gen Virol 1996; 77:2491–7.
9. Pipkin PA, Afzal MA, Heath AB, Minor PD. Assay of humoral im-
munity to mumps virus. J Virol Method 1999; 79:219–25.
10. Cohen BJ, Audet S, Andrews N, Beeler J. WHO working group on
measles plaque reduction neutralization test. Vaccine 2007; 26:59–66.
11. Bitsko RH, Cortese MM, Dayan GH, et al. Detection of R NA of
mumps virus during an outbreak in a population with a high level of
measles, mumps, and rubella vaccine coverage. J Clin Microbiol 2008;
46:1101–3 .
12. Ennis FA. Immunity to mumps in an institutional epidemic. Correlation
of insusceptibility to mumps with serum plaque neutralizing and
hemagglutination-inhibiting antibodies. J Infect Dis 1969; 119:654–7.
13. Meyer MB, Stifler WC, Joseph JM. Evaluation of mumps vaccine given
after exposure to mumps, with special reference to the exposed adult.
Pediatrics 1966; 37:304–15.
14. Bashe WJ, Gotlieb T, Henle G, Henle W. Studies on the prevention of
mumps. VI. The relationship of neutralizing antibodies to the de-
termination of susceptibility and to the evaluation of immunization
procedures. J Immunol 1953; 71:76–85.
15. Brunell PA, Brickman A, O’Hare D, Steinberg S. Ineffectiveness of
isolation of patients as a method of preventing the spread of mumps.
Failure of the mumps skin-test antigen to predict immune status. N
Engl J Med 1968; 279:1357–61.
16. Marin M, Quinlisk P, Shimabukuro T, Sawhney C, Brown C, Lebaron
CW. Mumps vaccination coverage and vaccine effectiveness in a large
outbreak among college studentsdIowa, 2006. Vaccine 2008; 26:3601–7.
17. Sosin DM, Cochi SL, Gunn RA, Jennings CE, Preblud SR. Changing
epidemiology of mumps and its impact on university campuses. Pe-
diatrics 1989; 84:779–84.
18. Henle G, Henle W, Wendell KK, Rosenberg P. Isolation of mumps
virus from human beings with induced apparent or inapparent in-
fections. J Exp Med 1948; 88:223–32.
19. Falk WA, Buchan A, Dow M, et al. The epidemiology of mumps in
southern Alberta 1980–1982. Am J Epidemiol 1989; 130:736–49.
20. Cooney MK, Fox JP, Hall CE. The Seattle Virus Watch. VI. Ob-
servations of infections with and illness due to parainfluenza, mumps
and respiratory syncytial viruses and Mycoplasma pneumoniae.AmJ
Epidemiol 1975; 101:532–51.
1422
d
JID 2011:204 (1 November)
d
Cortese et al
