LCMV is endemic in wild M. musculus populations across the United States and throughout the world. Sporadic human LCMV cases occur after contact with infected wild house mice, but the virus has the potential to cause large epidemics when it enters high-density rodent populations, as in the outbreak of this report. Previous outbreaks of human disease have been linked to contact with pet hamsters and laboratory animals (7–11). In the current outbreak investigation, we found that nearly one third of rodent facility employees tested had LCMV antibodies. This overall attack rate is consistent with previous outbreaks of LCMV in hamsters and nude mice at research facilities (10,12,13).

Employees of rodent breeding facilities of all kinds should be aware of the risks posed by exposure to rodents infected with LCMV, and monitoring programs should be in place to detect and control infections in rodents. Commercial laboratory rodent breeding colonies have developed management practices to avoid contact between wild mice and colony animals; the US Department of Health and Human Services and the Federation of European Laboratory Animal Science Associations recommend routine virologic and serologic monitoring to detect pathogens, including LCMV (2,14). Facilities producing rodents for the pet and feeder-rodent industries should adapt these practices to avoid such outbreaks. When LCMV or LCMV antibodies are detected in rodents or employees of a rodent breeding operation, all animal-handling personnel should wear protective equipment, including a respirator. After personal protective gear and training in its proper use were provided to facility employees, LCMV infection was not laboratory confirmed in any additional employees of facility A throughout the depopulation process, which implies that use of such equipment can reduce the risk for infection.

In facility A, many mice had LCMV antibodies, and several were viremic at the time of the investigation. The rodent testing results fit with the human testing results; that is, working in building 2, which had the highest prevalence of LCMV antibody-positive and RT-PCR–positive mice, was significantly associated with employee infections. Of the ≈13,000 adult mice in room 1 of building 2, ≈131–1,179 were viremic. In such a high-density environment, the virus can be present in aerosol form (15), which explains why many employees of facilities A and C had detectable antibodies. In facility B, which only transiently held live mice for further distribution, only 1 (8%) of 13 employees had detectable LCMV antibodies, suggesting that less intensive exposure to mice put employees at lower risk for infection. Smoking, a risk factor for many bacterial and viral infections, including tuberculosis, pneumococcal pneumonia, and influenza (16), was associated with LCMV infection in facility A employees. The structural changes to the respiratory epithelium and modulation of immune function hypothetically impair the smoker’s immune response. The physical act of smoking also might facilitate transfer of pathogens from the hands to the mouth.

The proportion of infected employees in whom clinical illness developed varied by facility. In facility A, 69% of employees who had recent infection reported illness, and 33% had symptoms severe enough to cause them to seek medical care. Conversely, all facility C employees appeared to have asymptomatic infections, except for 1 with nonspecific illness. Previous LCMV outbreak investigations have found various rates of clinical illness (9,10,13), with asymptomatic infections of 25%–55%. Because employee age and sex did not differ among facilities, the reason for the difference in disease manifestation remains unclear. The longevity of LCMV IgM detectable in peripheral blood is not well-established for humans; thus, some employees with IgM who did not develop symptoms might have been infected previously.

In immunocompetent adults, the neurologic form of LCMV infection classically has biphasic features consisting of a nonspecific initial phase, with fever, myalgia, and headache most commonly observed, and can include nausea and/or vomiting and retroorbital pain (8,17). Symptoms may subside after several days to be followed by a neurologic phase comprising meningeal symptoms with fever, headache, nuchal rigidity, vomiting, and light sensitivity. In LCMV patients in whom aseptic meningitis is diagnosed, CSF characteristically shows a lymphocytic pleocytosis, and elevated protein and decreased glucose also might be present (1,3); in this report, all 4 case-patients on whom lumbar punctures were performed had lymphocytic CSF with elevated protein. Symptoms frequently reported by case-patients in this outbreak included headache; fever; and abdominal symptoms, such as nausea, vomiting, and diarrhea. Such nonspecific symptoms obscure the clinical diagnosis. A thorough clinical history covering relevant animal contacts remains vital to determining the source and appropriate etiologies to test for in work-ups of aseptic meningitis.

Although all case-patients in this LCMV outbreak recovered from their illness, the specter of more severe disease manifestations remains a cause for concern. Infections during pregnancy can result in spontaneous abortion or characteristic congenital defects, such as chorioretinitis, microencephaly or macroencephaly, and hydrocephalus (17). Mental retardation, vision deficits, cerebral palsy, and epilepsy are potential lifelong manifestations (18). Since 2005, five outbreaks of LCMV have occurred after organ transplantation, resulting in the death of 14 (82%) of 17 organ recipients (19–22). Concern about these severe forms of LCMV infection led the outbreak response team to recommend depopulating the mouse breeding facilities and to conduct a trace-forward investigation of live mice that were sold and shipped from facility B (6).

LCMV is transmitted horizontally and vertically in affected rodents (16,23). Horizontal infections, acquired through direct contact with infected rodents or indirect contact with contaminated fomites, can lead rodents to shed infectious virus for a few weeks to a few months. When mice are exposed in utero to LCMV, they become persistently infected and shed the virus throughout their lives, including to all offspring, which also will be persistently infected (2,24,25). Infections in rodents are inapparent. As a result, when LCMV is introduced to a high-density environment, such as the breeding colonies of this outbreak, the number of infected rodents can silently reach very high numbers and thus pose a risk to the humans coming in contact with them.

The primary goal of the rodent sampling scheme was to detect LCMV infection in the colony; only adult mice were sampled because we assumed that they would have the greatest likelihood of having detectable antibodies. If younger mice had been tested, the seroprevalence is likely to have been lower and the proportion of viremic animals higher. In this investigation, 21% of adult mice tested had antibodies; this prevalence is higher than that found in wild mouse populations not associated with human infections (2%–9%) (26–28).

These molecular investigations demonstrated a unique strain of LCMV, 201202467, in facility A mice, which suggests a single introduction and transmission event throughout the breeding facility. Although we were unable to sample mice from facility C, it is likely that the outbreak strain was the same as in facility A and was introduced during the wild mouse infestation that had occurred. Molecular analysis also shows the close relationship of strain 201202467 with another LCMV strain isolated from a mouse found in an infested house in Michigan in 2005 (29). The geographic distribution of this LCMV strain throughout North America is not known (7).

As in previous outbreaks, interactions between wild mice and colony rodents frequently introduce LCMV into breeding colonies (16). In the absence of control measures and monitoring, movements of live mice among breeder mice can contribute to the spread of potentially infected mice. No human or rodent LCMV vaccine is available, and no treatment exists. Therefore, prevention measures are necessary and must rely on wild rodent exclusion, infection control, and microbiological monitoring (2). When LCMV antibodies are detected in colony mice, transmission must be assumed to be ongoing, and all possibly infected or exposed rodents should be removed from the colony by euthanasia and disposal to mitigate human risk (23,30).

In conclusion, laboratory and epidemiologic investigations effectively identified a large outbreak of LCMV in 3 commercial mouse breeding facilities and associated infections in several employees. The presumptive source of virus introduction was contact between wild mice and colony mice, and the outbreak spread among facilities when mice were transported for use as breeding stock. The breeding colonies were depopulated to prevent further human infections. Future outbreaks can be prevented with strict biosecurity and microbiological monitoring, and employees should be made aware of the symptoms of LCMV infection and prevention measures.