Mumps Vaccine Performance among University Students during a Mumps Outbreak

Abstract

Background. The largest reported mumps outbreak at a US college in 19 years occurred in 2006 at a Kansas university with a 2-dose measles-mumps-rubella (MMR) vaccination policy. We assessed vaccine performance and mumps risk factors, including the possibility of waning vaccine protection.

Methods. Case students were compared with a cohort of the university's ∼19,000 undergraduates. The secondary attack rate for clinical mumps was determined among roommates exposed to case students. Time from receipt of the second dose of MMR vaccine was compared between case students and roommates without mumps.

Results. Coverage with ⩾2 dose of MMR vaccine was ⩾95% among 140 undergraduate case students and 444 cohort students. The secondary attack rate for clinical mumps among roommates who had received 2 doses of vaccine ranged from 2.2% to 7.7%, depending on the case definition. Compared with roommates without mumps, case students were more likely (odds ratio, 2.46; 95% confidence interval, 1.25–4.82) to have received their second dose of MMR vaccine ⩾10 years earlier. The odds of being a case student increased with each 1-year increase in time from receipt of the second dose of MMR vaccine (odds ratio, 1.36; 95% confidence interval, 1.10–1.68) among case students and roommates aged 18–19 years but not among those aged ⩾20 years. Students aged 18–19 years had a higher risk of mumps (risk ratio, 3.14; 95% confidence interval, 1.60–6.16), compared with students aged ⩾22 years; women living in dormitories had increased risk of mumps (risk ratio, 1.95; 95% confidence interval, 1.01–3.76), compared with men not living in dormitories.

Conclusion. High 2-dose MMR coverage protected many students from developing mumps but was not sufficient to prevent the mumps outbreak. Vaccine-induced protection may wane. Similar US settings where large numbers of young adults from wild-type naive cohorts live closely together may be at particular risk for mumps outbreaks.

From 1999 through 2005, <400 cases of mumps were reported annually in the United States [1]; this low number was believed to be largely a result of a high rate of mumps vaccine coverage among persons without prior exposure to wild-type virus. In 1977, the Advisory Committee on Immunization Practices recommended 1 dose of mumps vaccine for all children [2, 3]. In 1989, the Advisory Committee on Immunization Practices recommended that 2 doses of measles vaccine, preferably given as measles-mumps-rubella (MMR) vaccine, be administered to all school-aged children and persons entering college to improve measles control [3, 4]. In 2006, however, the United States experienced a mumps resurgence, with 6584 reported cases—the largest annual number in 19 years (Centers for Disease Control and Prevention, unpublished data) [5–7]. Most cases occurred in the Midwest, and almost 200 cases were reported from a Kansas university. This university had instituted a 2-dose MMR vaccination requirement in 1993 and estimated high 2-dose MMR vaccine coverage among its students. Therefore, as was true for the 2006 US resurgence in general, the large size of the outbreak at this institution was unexpected and raised questions about the effectiveness of mumps protection from 2 doses of MMR vaccine. The objectives of this study were to measure mumps vaccine coverage and vaccine performance among the students and to determine whether there was evidence of waning vaccine-induced protection.

Methods

Mumps Vaccine Coverage

Mumps vaccine coverage was assessed for case students and a cohort of undergraduate students. Suspected cases of mumps in students (according to the Centers for Disease Control and Prevention and Council of State and Territorial Epidemiologists mumps definition) were reported to the Lawrence-Douglas County Health Department, Kansas State Department of Health and Environment, or the university's student health service [8]. Study investigators attempted to interview all students reported to have mumps in person or via telephone using a standard questionnaire and reviewed physician records.

A case student was defined as a student enrolled at the university's main campus for the 2006 spring term who was reported to have suspected mumps and who met at least 1 of the following criteria: (1) the physician record described enlargement or tenderness of parotid gland(s) or, for males, testicular swelling or tenderness, without a likely alternative diagnosis; (2) the student reported enlargement of parotid gland(s) for ⩾2 days (i.e., the student was asked, “Did you have swelling of the parotid glands, the glands near your cheeks and ears?”), or a male student reported testicular swelling or pain without likely alternative diagnosis; (3) mumps virus was isolated from buccal swab or urine specimens, or mumps RNA was detected by RT-PCR of a buccal swab specimen from a symptomatic student. A possible case student was defined as a student reported to have suspected mumps who did not meet the case student definition and did not have an alternate diagnosis in the medical record.

For the cohort of students, 450 students were selected from a randomly ordered list of all undergraduate students (∼19,000) for the spring term 2006 (A. Huang, M. Cortese, A. Curns, R. Bitsko, H. Jordan, F. Soud, J. Villalon-Gomez, P. Dennning, K. Ens, G. Dayan, unpublished data). Six students were subsequently excluded, because they primarily attended a satellite campus. The final cohort consisted of 444 students.

Mumps vaccination status of students was determined through review of hard-copy vaccination records required at admission to the university, vaccination documents in student health service medical records, and the student health service electronic vaccination database. If a vaccination record was incomplete, attempts were made to contact the student to obtain the record or to locate information for their physician so that the record could be obtained. Students born before 1957 were considered to be immune because of natural exposure [3].

Vaccine Performance: Secondary Attack Rate among Roommates

The clinical mumps secondary attack rate was determined among roommates of case students. Roommates were identified during case student interviews or through health department records. A standard questionnaire was administered to consenting roommates via telephone or e-mail.

An exposed roommate was defined as a roommate who lived in a room, suite, or apartment with a case student for at least 1 night during the case student's infectious period (from 3 days before to 2 days after illness onset) [9]. Two days after onset was chosen as the end of the infectious period, because case students were advised to use masks and isolate themselves, potentially shortening the transmission period.

A secondary case student was defined as a roommate who had onset of mumps (meeting the study case definition) 12–25 days after the primary case student's illness onset. A probable secondary case student was a roommate who developed only parotid gland or jaw pain during that period, without physician confirmation. A coprimary case student was defined as a roommate with mumps onset within 11 days after the primary case student's onset. Vaccination records of roommates were assessed as previously described.

Secondary attack rates were calculated by dividing the number of exposed symptomatic roommates by the total number of exposed roommates. Secondary attack rates were calculated by both excluding and including possible case students (as primary cases) and their roommates.

Assessment of Possible Waning Protection

Case-cohort study. To evaluate time from the second dose of MMR vaccine and other potential risk factors for disease among vaccinated students, undergraduate case students who received 2 MMR vaccine doses before the outbreak were compared with cohort students who received 2 MMR vaccine doses before the outbreak. Additional analyses were performed (1) excluding students born before 1984, because they may have been protected from developing mumps in 2006 by exposure to wild-type virus during the US mumps resurgence in the late 1980s [10, 11] (students born after 1983 were younger than school-age during that resurgence); (2) excluding students who received MMR vaccine at ⩾18 years of age who may have received 2 MMR vaccine doses before matriculation but were revaccinated because the vaccination record was unavailable; (3) excluding students designated as international students, because they may have been exposed to wild-type virus in their native country; (4) including students who had received ⩾3 MMR vaccine doses and using the time from the last received dose of MMR vaccine.

Age, in completed years, was calculated by subtracting date of birth from date of illness onset (case students) or the outbreak midpoint, April 2006 (cohort students). Time from receipt of the second dose of MMR vaccine, rounded to the nearest one-tenth of a year, was calculated by subtracting date of receipt of the second dose of MMR vaccine from date of illness onset or the outbreak midpoint.

Analyses were performed using SAS, version 9.1.3 (SAS Institute). Case and cohort students were compared in univariate analyses by χ² tests. Risk ratios with 95% CIs were estimated. Variables with a P value <.05 were considered to be statistically significant and were included in multivariate logistic regression analyses. Interaction between variables was assessed. Forward and backward selection algorithms were used to identify factors independently associated with case status.

Case student and roommate study. Time from receipt of the second dose of MMR vaccine and other potential risk factors were also compared between case students and roommates who did not develop clinical mumps. Undergraduate and graduate students who had received 2 doses of MMR vaccine were included. For roommates, age at exposure (primary case student's illness onset date) and time from receipt of the second dose of MMR vaccine to exposure were used. Analyses were performed as described for the case student and cohort student study.

Results

Mumps Vaccine Coverage

An interview was completed for 135 (78%) of 174 students who were reported as having suspected mumps, and medical records were reviewed for 168 (97%) of these 174 students. A total of 149 (86%) of 174 case students met the case definition; 140 were undergraduate students, and 9 were graduate students. Mumps virus was isolated from 5 students, and mumps RNA was detected in 8 others. The cohort included 1 case student.

A mumps vaccination record that was considered to be complete was obtained for almost all undergraduate case and cohort students (table 1). Counting students with incomplete or missing vaccination records as not having received 2 MMR vaccine doses, 135 (96%) of 140 undergraduate case students and 421 (95%) of 443 cohort students born after 1957 had received ⩾2 MMR vaccine doses before the outbreak. No student received MMR vaccine after January 2006.

Table 1

Case-cohort study: mumps vaccination status among undergraduate case and cohort students.

Vaccine Performance: Mumps Transmission among Roommates

Of the 149 students who met the case definition, the presence or absence of roommates was assessed for 121 (81%). One hundred (83%) of 121 of the assessed case students had at least 1 roommate, for a total of 177 potentially exposed roommates. Of these roommates, 36 could not be contacted, 14 declined participation, 13 were not exposed, and 8 were not university students. None of the students in the first 3 categories were reported as having mumps. Of the remaining 106 roommates, 16 were excluded from analysis because they were coprimary case students (n=12), were possible primary case students (n=3), or had received only 1 dose of MMR vaccine (n=1); therefore, 90 roommates remained in the main analysis. Of the 64 case students who had roommates included in the analysis, 62 had received ⩾2 doses of MMR vaccine, 1 had received 1 dose, and the record was incomplete for 1 case student. There were an additional 23 possible case students; roommates were assessed for 16 (70%), and 11 had roommates who were “exposed.”

Two secondary cases of mumps were identified among the 90 exposed roommates (attack rate, 2.2%; 95% CI, 0.3%–7.8%); 2 probable secondary cases were also identified (overall attack rate, 4.4%; 95% CI, 1.2%–11.0%) (table 2). Lengthening the incubation window by 12 days (to 6–31 days) to account for possible errors in onset dates did not capture additional transmissions. Attack rates increased when the outcome and case criteria were relaxed (table 2). Attack rates increased by ≤1% when roommates without mumps symptoms who were born before 1984 were removed from the denominator (data not shown).

Table 2

Mumps secondary attack rate among roommates of students with mumps.

Assessment of Possible Waning Protection

Case-cohort study. Of the 132 undergraduate case students who had received 2 doses of MMR vaccine, 42% were aged 18–19 years; 88% had received their second dose of MMR vaccine ⩾10 years earlier, and 80% received their second dose of MMR vaccine during 1990–1994. Time from receipt of the second dose of MMR vaccine was not statistically significantly different between the case and cohort students (P=.50, by Wilcoxon rank sum test). By univariate analyses, case students were more likely than cohort students to be younger and female and to have received their second dose of MMR vaccine at 1–6 years of age (table 3).

Table 3

Case-cohort study: characteristics and univariate analysis comparing undergraduate students with mumps and cohort students.

By multivariate analyses, younger students had increased risk of mumps (table 4). The effect of sex on risk of mumps was modified by dormitory residency; women living in dormitories had increased risk of mumps, compared with both men living in dormitories and men not living in dormitories, although the risk of mumps was not different between women and men who did not live in dormitories. Neither time from receipt of the second dose of MMR vaccine nor age at receipt of the second dose of MMR vaccine was significantly associated with case status. Results from additional analyses were similar to those reported above (data not shown).

Table 4

Case-cohort study: multivariate analysis comparing undergraduate students with mumps and cohort students.

Case student and roommate study. The 139 case students were compared with 88 roommates who did not develop mumps. The majority of case students and roommates were aged 18–19 years, were female, were not dormitory residents, and had received their second dose of MMR vaccine ⩾10 years earlier (table 5). By univariate analysis, case students were more likely to have received their second dose of MMR vaccine ⩾10 years earlier and to have received the 2 MMR vaccine doses <6 years apart, compared with roommates. Time from receipt of the second dose of MMR vaccine was not different between case students and roommates by Wilcoxon rank sum test (P=.12).

Table 5

Case student and roommate study: characteristics and univariate analysis comparing students with mumps and roommates without mumps.

By multivariate analysis, only time from receipt of the second dose of MMR vaccine was significantly associated with case status (table 6). Compared with roommates without mumps, case students were more likely (OR, 2.46; 95% CI, 1.25–4.82) to have received their second dose of MMR vaccine ⩾10 years earlier. No statistically significant effect modification was detected between time from receipt of the second dose of MMR vaccine as a categorical variable and other variables (for interaction between time from receipt of the second dose of MMR vaccine and age, P=.09). When examined as a continuous variable, the effect of time from receipt of the second dose of MMR vaccine on mumps risk was modified by age. Among students aged 18–19 years, the odds of being a case student increased by a 36% (95% CI, 10%–68%) with each 1-year increase in time from receipt of the second dose of MMR vaccine. Among students aged ⩾20 years, the odds did not significantly increase with 1-year increases in time from receipt of the second dose of MMR vaccine (table 6 and figure 1). Adding a squared term in the model for time from receipt of the second dose of MMR vaccine did not significantly improve model fit. Results from additional analyses were similar to those reported above (data not shown).

Table 6

Case student and roommate study: multivariate analysis comparing students with mumps and roommates without mumps.

Figure 1

Proportion of case students with mumps and roommates without mumps, by time from receipt of the second dose of measles-mumps-rubella vaccine (MMR2). A, Students aged 18–19 years. B, Students aged ⩾20 years.

Discussion

A large mumps outbreak occurred at this university despite high 2-dose MMR vaccine coverage among undergraduate students. Because no vaccine is 100% effective, cases of mumps in vaccinated persons can be expected, even in those who have received 2 doses of MMR vaccine [12–14]. However, the large size of the outbreak raised general questions about mumps protection in young adults.

The low attack rate (2%–8%) in the roommate study suggests that, overall, 2 MMR vaccine doses protected most exposed students from developing clinical mumps. Reported attack rates among unvaccinated household members without a history of mumps have ranged from 31% to 48% in observational studies [15–17]. The low attack rate (measured among roommates who had received 2 doses of MMR vaccine and were exposed to case students, virtually all of whom had also received 2 doses of MMR vaccine) likely occurred because of adequate immunity in roommates. Whether patients with “breakthrough” mumps shed relatively low levels of virus, thereby contributing to the low attack rate, is unknown. Although vaccine effectiveness could not be estimated because of the lack of an unvaccinated comparison group, the data suggest that the majority of roommates were protected against clinical mumps by their 2 doses of MMR vaccine. There are limitations to this assertion. Protection would be overestimated if some roommates were not actually exposed to infectious virus. This might have occurred if some case students did not truly have mumps. Our case definition relied largely on clinical criteria and allowed for self report of symptoms among students with suspected mumps reported by health care professionals. However, our definition was similar to the Centers for Disease Control and Prevention and Council of State and Territorial Epidemiologists definition, and by reviewing medical records, we were able to document mumps-specific features for most cases (A. Huang, M. Cortese, A. Curns, R. Bitsko, H. Jordan, F. Soud, J. Villalon-Gomez, P. Dennning, K. Ens, G. Dayan, unpublished data).

Exposure would also have been overestimated if the type of contact between roommates and case students was not always sufficient to transmit virus. Mumps virus is transmitted through saliva or large droplets. Data from the prevaccine era indicate that mumps is less infectious than measles and varicella, evidenced by the older mean age at the time of mumps acquisition [18] and the lower estimated attack rate among household members [15]. Vaccine-induced protection would also be overestimated if some roommates were actually protected by pre-2006 exposure to wild-type virus. This seems unlikely because of the epidemiology of reported mumps in the United States during the lifetime of these young adults. Notably, only a small proportion of roommates were school-aged during the previous mumps resurgence in the late 1980s [10, 11].

Although the roommate study indicates that 2 doses of MMR vaccine were protective against clinical mumps, none of the studies presented here address how well MMR vaccine protected against mumps infection and, most importantly, against infectiousness. In unvaccinated persons, mumps exposure can result in asymptomatic or minimally symptomatic infections with viral shedding [19, 20]; the incidence and infectiousness of such outcomes among persons who have received 2 doses of vaccine are unknown. If vaccine effectiveness against infectiousness is much lower than effectiveness against clinical disease, minimally symptomatic persons may play an important role in an outbreak such as this one, in which most exposed persons were vaccinated.

In the case-cohort student study, students who were exposed, developed mumps, and were reported by health care professionals were compared with a sample of undergraduate students; therefore, factors associated with mumps in this analysis may be those that increased risk of exposure, increased risk of clinical disease once exposed, or increased likelihood of being reported. Younger students and women who lived in dormitories were found to be at increased risk of clinical mumps; these factors are likely to be associated with increased exposure risk and were also associated with mumps illness in the case-control study performed on this campus (A. Huang, M. Cortese, A. Curns, et al., unpublished data). In our comparison of case students and the cohort of undergraduate students, time from receipt of the second dose of MMR vaccine was not associated with case status.

In the case student and roommate study, however, an increase in the odds of having mumps was detected among students who had received their second dose of MMR vaccine ⩾10 years earlier, suggesting waning of protection. An increase in odds of having mumps with each 1-year increase in time from receipt of the second dose of MMR vaccine was found among students aged 18–19 years but not among those aged ⩾20 years. A consistent finding across age groups would be stronger evidence of waning protection. One possible explanation for the disparate results could be that only the cohort of students aged 18–19 years was completely naive to wild-type virus before the outbreak and, therefore, was the group in which waning vaccine-induced protection could be detected. As discussed earlier, this was not very likely. If protection does indeed wane, a linear decrease may poorly approximate the clinical expression of the biological process in a population. Finally, primary mumps vaccine failure occurs, and this reduces the ability to detect secondary failure with our study design.

Waning protection against mumps after 1 dose of MMR vaccine was suggested by authors investigating 2 US outbreaks [11, 21] but was not detected in other investigations [13, 22, 23]. A Belgian study involving predominantly school-aged children who had received 1 dose of vaccine found that the odds of having mumps increased 27% per 1-year increase in time from vaccination, beginning 1 year after receipt of the last dose of MMR vaccine [24]. A recent UK evaluation suggested that protection waned after 1 and 2 doses of MMR vaccine, with vaccine effectiveness from 2 doses of MMR vaccine of 86% among children aged 11–12 years (who had received the second dose of vaccine 6–7 years earlier), compared with 99% effectiveness among children aged 5–6 years (who had recently received the second 2 dose of vaccine) [13]. Although our data also suggest that protection may have waned, the low attack rate among roommates still indicates that the majority of young adults who had received 2 doses of MMR vaccine were protected against clinical mumps.

An outbreak among students who had received 2 doses of MMR vaccine at a US college was not reported before the mumps resurgence in 2006. More than 1 factor likely contributed to our mumps outbreak. A population-level explanation is the accumulation of students from wild-type naive cohorts who were susceptible to mumps because of primary or secondary vaccine failure to some critical level needed to sustain an outbreak. This was also postulated as a cause of the 1988 outbreak in Douglas County, Kansas, which occurred among secondary school children with high 1-dose MMR vaccine coverage [11]. A possible vaccine-related explanation is reduced protection of MMR vaccine against transmission of this particular mumps strain [25]. Alternatively, perhaps a relatively less effective vaccine was given to this cohort, compared with earlier cohorts of students who also did not have prematriculation natural immunity. Because a maximum of 35% of case students received the first or second dose of MMR vaccine in any calendar year, however, administration of a less effective vaccine for ⩾2 years would need to be invoked to account for most cases in our outbreak. Considering exposure-related factors, it seems to be unlikely that a mumps outbreak at a college campus was not reported from 1989 through 2005 because mumps virus had not been introduced onto a campus during that period. In our outbreak, multiple introductions of virus onto campus may have occurred, given the mobility of students and the number of mumps cases among young adults in surrounding states [5].

The data indicate that 2 doses of MMR vaccine protected the majority of students from developing clinical mumps but were not effective enough to prevent an outbreak of mumps despite a high coverage rate. We found evidence that vaccine-induced protection may wane. Additional studies are needed to better elucidate the issues of waning immunity and vaccine effectiveness against mumps transmission, particularly among cohorts of young adults who are naive to wild-type virus and live in congregate settings.

Acknowledgments

We thank the following individuals for their assistance with the investigations: Joe Gillespie, Diane Hendy, Sherry Rhine, Andrea Gore, Ronald Mier, and the physicians and medical records staff at the Watkins Memorial Health Center; Barbara Schnitker, Kay Kent, and the public health nurses at the Lawrence-Douglas County Health Department; Gail Hansen of the Kansas Department of Health and Environment; Angela Huang, Rebecca Bitsko, Fatma Soud, José Villalon-Gomez, Lumbe Davis, Mona Marin, Charles LeBaron, Umesh Parashar, and Paul Gargiullo from the Centers for Disease Control and Prevention; and the students at the University of Kansas. We thank Susan Goldstein for reviewing the manuscript.

Potential conflicts of interest. All authors: no conflicts.

The findings and conclusions of this article are those of the authors and do not necessarily represent the views of the funding agency.

Present affiliation: Sanofi Pasteur, Swiftwater, Pennsylvania.