Mumps Vaccines

Clinical Description and Complications

Table 41.1 summarizes the clinical manifestations of mumps. Symptoms and duration vary widely among studies and are dependent on age group, sex, vaccination status, and the basis for defining a case. Complications are more frequent in adults than children; by factoring out sex-specific manifestations (e.g., orchitis, oophoritis, mastitis), rates of complications are similar in males and females, except for neurological manifestations, which have a 3:1 or greater male-to-female ratio. The most common manifestation is salivary gland swelling, usually the parotid, which serves as the basis for the clinical case definition. Parotitis may be unilateral ( Fig. 41.1 ) or, more commonly, bilateral. Parotitis develops an average of 16–18 days after exposure ^, and may be preceded by several days of non-specific symptoms, including fever, headache, malaise, myalgias, and anorexia. Fever usually lasts 1–6 days, but enlargement of the parotid gland may persist 10 days or longer.

If viruria is used as an indication of kidney infection, then kidney involvement is common in mumps, although clinical nephritis is rarely diagnosed. Utz and colleagues found asymptomatic excretion of mumps virus in the urine of 72% of mumps cases evaluated. A similar frequency of viruria was identified among military personnel with mumps admitted to a medical center. Abnormal renal function can be measured in more than a third of cases; however, kidney-specific manifestations are rare. ^,

Orchitis, typically unilateral, occurs in up to a third of cases in postpubertal men. ^{, ,} In cases of bilateral orchitis, hypofertility and testicular atrophy have been reported; however, these are uncommon and sterility is exceedingly rare. An increased risk of testicular cancer has been reported in association with mumps orchitis but remains unproved.

In females, oophoritis and mastitis can occur, with rates ranging from 5% to 30% of cases. ^{, ,} Infertility subsequent to oophoritis has been suggested but not confirmed.

Involvement of the CNS occurs in fewer than 1% to up to 62% of clinical cases, depending on the stringency of assessment. If lumbar punctures are performed routinely, greater than half of persons with mumps may show evidence of meningitis based on pleocytosis of the cerebrospinal fluid (CSF), ^, whereas clinical evidence of meningitis is seen in approximately 5% of cases. Encephalitis is much more serious, occurs in approximately 0.5% of clinically apparent mumps infections, and is responsible for the majority of fatal cases. ^{, ,} Neurological manifestations are generally preceded by parotitis by 4–5 days but can occur before or in the absence of detectable salivary gland swelling.

Pancreatitis, usually mild, is seen in approximately 4% of cases. ^, Significantly higher rates have been reported in some outbreaks, although these are likely overestimates given that the diagnosis is often based on elevated serum amylase levels irrespective of clinical symptoms.

Other infrequent complications include endocardial fibroelastosis, ^, cerebellar ataxia, ^, transverse myelitis, ^, hydrocephalus, ^, arthropathy, ^, autoimmune hemolytic anemia, ^, thyroiditis, ^, thrombocytopenia, ^, and hearing loss. Hearing loss is typically unilateral and transient, but sudden onset permanent deafness occurs in approximately 1 in 20,000 cases. ^,

Non-specific respiratory symptoms have been ascribed to mumps, but it is not clear if such symptoms are caused by the virus or are coincidental. ^,

Based on assessments of seroconversion in longitudinal studies or outbreak studies, asymptomatic infections occur in approximately 20–40% of infected persons. ^{, ,} The high rate of subclinical mumps infections was experimentally demonstrated in a small study by Henle and colleagues in 1947. Of 15 initially seronegative children orally inoculated with high doses of live mumps virus, all seroconverted, yet 7 (47%) showed no evidence of mumps-specific symptoms.

Virology

Granata was the first to suggest a viral etiology for the disease in 1908 based on experiments in rabbits, work that was replicated in monkeys by Nicolle and Conseil in 1913 and by Gordon in 1914. A viral etiology was finally proven in 1935 by Johnson and Goodpasture in a remarkable study, although perhaps not for its time, wherein Koch’s postulates were fulfilled using children residing in the neighborhood of one of the investigators.

The virus is an enveloped, negative-strand RNA virus in the Paramyxoviridae family, genus Rubulavirus . The genome consists of 15,384 nucleotides encoding seven genes: the nucleoprotein (N), V, matrix (M), fusion (F), small hydrophobic (SH), hemagglutinin-neuraminidase (HN), and the large (L) protein genes. Each gene encodes a single protein, except the V gene, which additionally encodes the phosphoprotein (P) and I protein via cotranscriptional insertion of nontemplated guanine residues. ^,

The role of the N protein is to protectively and functionally coat the viral RNA to form the ribonucleoprotein, which is the template for RNA synthesis, performed by the RNA-dependent RNA polymerase, a complex consisting of the P and L proteins. The viral envelope is a host-cell–derived lipid bilayer membrane containing the M protein on its internal surface and the HN and F proteins on its outer surface. The M protein appears to be involved in the assembly of viral proteins by forming a bridge between the N proteins of the ribonucleoprotein and the cytoplasmic tails of the HN and F proteins followed by budding from the cell. ^, The HN protein is responsible for adsorption of the virion to its cellular receptor, sialic acid, and release of nascent virions from the cell surface. ^, The HN and F proteins work cooperatively to induce fusion of the virion membrane with the host cell membrane and to induce cell-to-cell fusion. ^, The nonstructural V protein is involved in circumventing the interferon (IFN)-mediated antiviral responses by blocking its signaling and limiting its production. ^, The SH protein, like the V protein, also appears to aid in evasion of the host antiviral response, but by interfering with TNF-α-mediated NF-kB activation (Franz et al, 2017). Neither of these proteins are essential to virus replication, as demonstrated by the rescue of recombinant viruses engineered to not express these proteins. ^, The role of the putative I protein in virus replication and infection is not known.

Only antibodies directed to the cell surface expressed F and HN proteins definitively have been shown to neutralize the infectivity of the virus in vitro and confer protection in vivo.

Mumps virus is considered serologically monotypic; however, there is some evidence that antigenic differences between strains may allow for breakthrough infection by heterologous strains in persons with low levels of neutralizing antibodies (see “Risk Factors for Vaccine Failure and Recent Outbreaks”).

Pathogenesis as It Relates to Prevention

Based on knowledge of the pathogenesis of related respiratory viruses, it is presumed that mumps virus initially infects and multiplies in the upper respiratory tract before spreading to draining lymph nodes and disseminating via plasma viremia. Given the wide array of acute inflammatory reactions, there is little doubt that viremia ensues; however, primary sites of virus replication following exposure have not been demonstrated and virus has been detected in blood only rarely.

Reported symptoms and laboratory findings indicate glandular tissues (parotid glands, testes, breasts, and pancreas) and the central nervous system (CNS) as the most common sites of mumps virus dissemination. If viruria is used as an indication of kidney infection, then kidney involvement is also common.

Virus is shed in saliva as early as 11 days after exposure, which is about a week before symptom onset (i.e., parotitis). ^, Viruria can be detected during the first 5 days of illness and can last up to 14 days. ^{, ,} The period of virus shedding in saliva does not seem to be influenced by vaccination status or symptom severity, including in those asymptomatically infected. Termination of viral shedding in saliva correlates with the local appearance of virus-specific salivary IgA and IgM, usually appearing within a few days of symptom onset. ^, Fewer than 15% of patients continue to shed virus in saliva beyond day 5 of symptom onset.

The pathogenesis of specific complications, such as nephritis and orchitis, is uncertain. Nephritis may be the result of direct viral infection of the kidney or immune complex glomerulonephritis. Although a limited number of pathologic specimens have been examined, there is evidence to suggest that immune complex deposition may have a role in some cases. In the rare cases in which renal biopsies or necropsies were performed, findings included adherent renal capsule, epithelial debris in the renal capsule lumen, interstitial mononuclear cell infiltration, edema, and focal tubular epithelial cell damage. ^,

Orchitis is thought to be a result of direct viral invasion, based on recovery of virus from the semen and testis, ^, but some cases of orchitis have been hypothesized to be immune-mediated.

The pathogenesis of mumps infection of the CNS is almost wholly based on studies in hamsters using hamster-adapted strains of the virus. Based on these studies, it appears that virus enters the CNS through the venous vasculature of the choroidal and ependymal epithelial cells. Once infected, these cells become a reservoir for continued production and release of virus, which can penetrate deeper into the brain parenchyma causing encephalitis and numerous neurological complications (see “Clinical Description and Complications”).

Mumps virus–induced deafness likely involves retrograde penetration by the virus into the perilymphatic fluid following viremia via the cervical lymph nodes leading to infection of the cochlea. Although the perilymph communicates with the cerebrospinal fluid, deafness does not occur more frequently in persons with CNS complications, suggesting that mumps deafness is not a complication of CNS infection. Damage to the organ of Corti is a common finding, with degeneration of the stria vascularis and tectorial membrane being reported in some, but not other, studies. Hearing loss caused by indirect effects of virus infection (e.g., immune-mediated damage) also has been suggested.

Modes of Transmission

The demonstration by Johnson and Goodpasture and by Henle and colleagues that the virus can be transmitted to humans by nasal or buccal mucosal inoculation suggests that natural infection is initiated by droplet spread. Mumps virus appears to be capable of being transmitted transplacentally and infect the fetus, but this has been reported only rarely. ^,

The incubation period is typically 16–18 days after exposure, but considering that virus is shed in saliva as early as 1 week before symptom onset it is perhaps not surprising that outbreak investigations performed during the prevaccine era indicate that most transmissions occur before symptom appearance or during the first few days of symptom onset, highlighting the limitations of preventing virus transmission by case isolation.

Incidence and Prevalence

In the years before the introduction of mumps vaccine in the United States, mumps was most commonly reported among young school-age children, with more than half of cases being reported in 5–9 year olds and a quarter of cases reported in children 0–4 years of age. These data are consistent with prevaccine era serosurveys from around the world. A seroepidemiologic study in St. Lucia demonstrated that 70% of children were seropositive for mumps at 4 years of age. In The Netherlands, peak acquisition occurred between ages 4 and 6 years and in the United Kingdom and Spain between ages 5 and 7 years. ^,

Table 41.2 shows annual incidences in various countries in the prevaccine era, averaging approximately 300 per 100,000 population based on passive reporting. The actual annual incidence in the general population is likely much higher, as indicated in limited prospective studies. ^{, ,} It has been estimated that less than a third of mumps cases are reported.

Hope-Simpson demonstrated in household contact studies that mumps is less infectious than measles or varicella. Secondary attack rates among susceptible household contacts younger than 15 years were 76% for measles, 61% for varicella, and 31% for mumps. Some of the observed differences were undoubtedly a result of a higher rate of subclinical infection with mumps, but the higher average age at infection observed for mumps also supports the less-efficient transmission of mumps virus.

Most epidemiologic reports suggest an interepidemic period for mumps of approximately 3 years, ^{, ,} as portrayed by Barskey and colleagues who summarized the number of reported cases of mumps in the United States in the prevaccine era ( Fig. 41.2 ). In temperate zones, mumps exhibits seasonality, with peak incidence during the winter and spring and a nadir in the summer. Seasonality of mumps has not been noted in tropical areas.

In 1989, the Advisory Committee on Immunization Practices (ACIP) recommended all children receive two doses of measles vaccine. Because measles vaccine is generally given in the United States as MMR (measles, mumps, and rubella trivalent vaccine), implementation of this recommendation has meant that most children receive two doses of mumps vaccine. Fig. 41.3 shows the effect of implementation of a two-dose vaccination schedule on disease incidence in the United States. A dramatic reduction in disease incidence was apparent in the first four decades following the 1967 licensure and use of a mumps vaccine, but a trend reversal started in 2006. Importantly, despite this resurgence, the average number of reported cases over the past decade remains orders of magnitude below levels experienced in the prevaccine era. The more recent data showing the occurrence of large mumps outbreaks in highly vaccinated populations indicates complete protection against mumps using current vaccines and vaccination schedules may not be feasible (see “Effectiveness in Field Use and Duration of Immunity”).

High-Risk Groups

High-risk groups include international travelers, college students, military personnel and other persons who reside in close-contact environments, and healthcare personnel born during or after 1957.

Reservoirs of Infection

Although numerous mammalian species can be experimentally infected with mumps virus, humans are the only natural reservoir and only source of transmission. A virus with >90% sequence identity to mumps virus has been identified in bats and has been determined to be serologically closely related to human mumps virus, but it is unclear if this virus can be transmitted to humans. ^,

Significance as a Public Health Problem

Prior to widespread use of mumps-containing vaccines, annual incidences ranged from 100 to 1000 per 100,000 population, and by adolescence, nearly everyone had serological markers of prior infection. Mumps is particularly a problem in large groups of people in close, frequent contact, such as the schools and the military. In a 1968 editorial surrounding discussion on whether a mumps vaccine is warranted, the Georgia Department of Public Health Chief Epidemiologist stated “As a cause of continuing school absenteeism and year-long disruption of teaching schedules, mumps has no equal.” The problem for the military was discussed earlier (see “Why the Disease Is Important”). The burden of disease in the absence of vaccination is staggering when considering direct costs for outpatient and inpatient visits, costs of outbreak control, and productivity losses incurred by days of missed work, including for provision of care for sick children and family members. Zhou and colleagues estimated that more than 2.3 million cases of mumps were prevented in the United States by vaccination in 2009, equating to a savings of greater than $1.4 billion in direct costs and more than $2.3 billion in societal costs. Interestingly, whether or not to routinely vaccinate against mumps was a subject of great debate in the late 1960s, with some citing the general benignity of mumps, hazards of occult contaminating agents that might be present in vaccine, sensitization and oncogenicity that might result from use of live virus vaccine, and concerns over use in young children. In retrospect, these concerns were for naught.

History of Vaccine Development

Following his 1945 discovery that mumps virus can be cultivated in embryonated hens’ eggs (also reported by Levens and Enders the same year) Karl Habel of the U.S. Public Health Service produced the first experimental vaccine (inactivated) and tested its effectiveness in 1946 among 2825 West Indian workers on a sugarcane plantation in the Everglades region of Florida. Forty cases developed among 1344 vaccinees between weeks 3 and 16 postvaccination compared to 106 cases among 1481 unvaccinated controls, for a vaccine efficacy over that time period of 58%. Although inactivated vaccines were somewhat effective, the duration of immunity was short, and their use was abandoned in the United States and elsewhere in the 1950s ^, in favor of live attenuated vaccines, which were being developed at that time.

All mumps vaccines in recent use are live viruses ( Table 41.3 ). Some strains are produced by multiple manufacturers, which may result in slight differences in their safety and effectiveness. Strains have been adapted to embryonated hens’ eggs and numerous cell types as indicated in Fig. 41.4 , which gives the history of passage during the attenuation of some of the more prominent strains listed in the table.

The first widely used live mumps vaccine, Leningrad-3, was produced in the former Soviet Union in 1956. The vaccine was derived from five wild-type isolates obtained from saliva and CSF specimens from children with mumps. Each virus isolate was passaged in embryonated hens’ eggs before being pooled for further passage in eggs to produce vaccine. In 1961, the vaccine was further passaged in guinea pig kidney cells and quail embryo fibroblast cells, with final product produced in guinea pig kidney cells. ^, Serial passage across multiple cell substrates likely selectively enriched the virus population, as later sequencing did not reveal data suggestive of the copresence of multiple different virus strains.

The Leningrad-3 vaccine was obtained by the Institute of Immunology in Zagreb in former Yugoslavia in the early 1970s and passaged in chicken embryo fibroblast culture and renamed L-Zagreb.

In the United States, in 1965 a vaccine strain was developed from virus isolated from the throat of Jeryl Lynn Hilleman, daughter of Maurice Hilleman who developed the product at Merck & Co. The isolate was attenuated by passage in embryonated hens’ eggs and whole chick embryo cell culture and aptly named Jeryl Lynn. The vaccine was licensed in the United States in 1967 and continues to be the only mumps vaccine approved for use in the United States and is the most widely used vaccine strain globally. The vaccine has been demonstrated to contain two distinct but genetically related viruses identified as JL-major (sometimes referred to in the literature as JL-1 or JL-5) and JL-minor (sometimes referred to in the literature as JL-2), existing in an approximate 5 : 1 ratio, respectively. These two viruses differ from each other at 414 nucleotide sites (3%), leading to 87 amino acid substitutions. GlaxoSmithKline (SmithKline Beecham Biologicals, at the time) clonally isolated JL-major from Merck’s JL vaccine and passaged it two times in chick embryo fibroblast cells to obtain a preparation designated RIT-4385, which was used as the virus seed for production of their mumps-containing MMR vaccine, Priorix. Based on prelicensure and postlicensure experience, both vaccines have similar safety and immunogenicity profiles. Other JL-based vaccines have also been produced by other companies and are in more limited use (see “Description of Vaccines and Producers”).

The Urabe Am9 strain was developed by the Biken Institute in Japan from an isolate obtained from the saliva of a child named Urabe. , The virus was isolated in primary human embryonic kidney cells and passaged in primary green monkey kidney cells followed by passages in the amniotic cavity of developing chick embryos. Virus was then plaque purified in quail embryo fibroblast cultures and one of the purified clones, clone 9, was selected and passaged in the amniotic cavity of developing chick embryos to produce Urabe Am9 vaccine.

Like JL, the Leningrad-3, L-Zagreb, and Urabe Am9 strains also have been shown to contain variant virus populations, but the level of variation is not as great as with JL, and they do not represent distinct virus strains, but are rather part of the quasispecies formed by negative-strand RNA viruses. The Urabe Am9 vaccine from two manufacturers has been demonstrated to contain at least two major variant virus populations existing in a 1 : 3 ratio. ^, At least two major variants were also identified in the L-Zagreb vaccine and in the Leningrad-3 mumps vaccine. ^, The clinical significance of variants is not known.

The Rubini mumps vaccine virus was derived from a mumps isolate obtained from the urine of Carlo Rubini in Switzerland in 1974. The isolate was passaged in WI-38 human diploid cell culture, followed by passages in embryonated hens’ eggs before adaptation to the human diploid cell line MRC-5, which was also the substrate for vaccine production. Interestingly, the Rubini vaccine strain SH gene sequence (a hypervariable 316 nucleotide stretch that is used for molecular epidemiological purposes) is 99% identical to the tissue culture-adapted Enders strain prepared in the United States in 1945 and to three wild-type strains isolated in Germany from 1987 to 1992, ^, As described later, this vaccine was found to offer virtually no protective efficacy. In 2001, the World Health Organization (WHO) recommended against its use. The product is no longer manufactured.

Because of the high neurotropism of wild-type mumps viruses, preclinical neurological safety testing of candidate mumps vaccines is required by national regulatory authorities. Historically, such testing has been performed in old world monkeys (Cercopithecidae) and is based on an evaluation of mumps virus–specific neuropathology (mainly periventricular inflammation and neuronal necrosis) following intracerebral inoculation. However, two independent studies conducted in 1999 indicated monkey neurovirulence safety test results do not correlate with the risk of vaccine virus neurovirulence for humans. ^, It is therefore perhaps not entirely surprising that certain mumps vaccines presumably passing the monkey neurovirulence safety test have produced aseptic meningitis in vaccinees (see “Rare Reactions”). On the other hand, the incidence of vaccine-associated aseptic meningitis is but a small fraction of the incidence of aseptic meningitis following natural infection, and compared to natural disease with a known risk of encephalitis, postvaccine aseptic meningitis is relatively mild.

While efforts have been made to identify specific molecular correlates of mumps virus neurovirulence, none have been identified that distinguish virulent from attenuated mumps virus strains. Several groups have identified numerous genetic differences between attenuated and virulent variants of the Urabe Am9 vaccine strain, but which of these and in what combination are responsible for the observed attenuation cannot be discerned and whether any might influence attenuation of other mumps virus strains appears doubtful. Similar conclusions have come from studies of other mumps virus strains. Rather than attempting to identify point mutations that might impact virus attenuation, others have engineered mumps viruses to preclude transcription or expression of specific genes, for example, the V and SH genes ^, that encode proteins that interfere with the antiviral host immune response. In both cases attenuation in vitro and in animals was achieved with such investigational recombinant viruses, but safety or effectiveness in humans has not been reported. Other approaches toward rational design of future mumps vaccines include use of established safe strains as the genetic backbone of the virus, but engineered to express the surface glycoproteins of strains responsible for current outbreaks. This is discussed further in Future Vaccines below.

Vaccine Use

In the United States, mumps vaccine was first licensed in 1967 and has been administered as MMR since 1971. The ACIP recommendation is that MMR be included in the national immunization schedule as a single dose initially, and then as a two-dose regimen in 1989. Currently, 122 of the 194 (63%) WHO member states use at least one dose of mumps-containing vaccine in their immunization programs. Of these, 109 countries use a two-dose regimen. The distribution of countries is shown in Fig. 41.5 and vaccine use in select countries is given in Table 41.4 .

In the United States, the two-dose MMR series is administered at ages 12–15 months and 4–6 years. The second dose may be administered before age 4 years, provided at least 4 weeks have elapsed since the first dose. For international travel, the ACIP recommends vaccinating infants 6–11 months of age before departure from the United States, followed by doses at 12–15 months of age and at least 4 weeks later. For older children who will travel internationally, the first dose is recommended on or after age 12 months before departure and the second dose at least 4 weeks later. In response to widespread mumps outbreaks on college and university campuses, in 2018 the ACIP issued guidance on the use of a third dose of MMR vaccine during mumps outbreaks, discussed in Disease Control Strategies and Possible Eradication.

In other countries, with few exceptions, the first dose is administered at 12–18 months of age, but the age at which a second dose is administered varies, from as soon as 1 month after the first dose, to 10–12 years after the first dose ( Table 41.4 ). The impact of the timing of the second dose is unclear. LeBaron and colleagues found similar levels of neutralizing antibody in 17-year-old children 12 years after receiving their second dose of MMR II at 4 years of age versus antibody levels in 17-year-old children 7 years after receiving their second dose of vaccine at 10 years of age, suggesting no advantage to delaying the age at which a second dose is administered.

In most countries, MMR vaccination is not compulsory; thus, mumps virus seroprevalence varies widely between countries. In the United States, where vaccine use is a requirement for school entry (see “Public Health Considerations”), vaccine coverage is relatively high.

For the 2020–21 school year in the US, vaccination coverage among children in kindergarten was 93.9% for 2 doses of MMR vaccine, an approximate 1% decrease compared with the 2019–20 school year. This decrease is likely related to the COVID-19 pandemic, the response to which may have impacted the amount and quality of student vaccination data collected and reported by local health departments, as well as parental decision to homeschool or delay school entry.

Vaccine use is also relatively high in a number of other countries, including Finland, Sweden, The Netherlands, Denmark, and Belgium. ^, In some countries with purportedly high vaccine coverage, seroprevalence is lower than expected. ^,

Description of Vaccines and Producers