Emerging Pattern of Rabies Deaths and Increased Viral Infectivity

Sharon L. Messenger, Jean S. Smith, Lillian A. Orciari, Pamela A. Yager, Charles E. Rupprecht


Emerging Infectious Diseases. 2003;9(2) 

In This Article

Abstract and Introduction

Most human rabies deaths in the United States can be attributed to unrecognized exposures to rabies viruses associated with bats, particularly those associated with two infrequently encountered bat species (Lasionycteris noctivagans and Pipistrellus subflavus). These human rabies cases tend to cluster in the southeastern and northwestern United States. In these regions, most rabies deaths associated with bats in nonhuman terrestrial mammals are also associated with virus variants specific to these two bat species rather than more common bat species; outside of these regions, more common bat rabies viruses contribute to most transmissions. The preponderance of rabies deaths connected with the two uncommon L. noctivagans and P. subflavus bat rabies viruses is best explained by their evolution of increased viral infectivity.

Bites by rabid dogs are the source of 35,000-50,000 human rabies deaths each year globally,[1] yet most human rabies deaths in the United States are attributed to unrecognized exposures to rabid bats. Particular attention has focused upon two relatively rare bat species (Lasionycteris noctivagans and Pipistrellus subflavus) because rabies variants associated with these species account for approximately 70% of human cases and 75% of cryptic rabies deaths.[2,3,4,5,6]

Molecular typing (i.e., phylogenetic analysis of DNA data) has shown that rabies viruses associated with insectivorous bats (L. noctivagans and P. subflavus variants in particular) are the culprits in what otherwise would have been unsolved cryptic human rabies deaths. However, phylogenetic analyses of human rabies cases have not provided insights into why an unexpectedly large proportion of human rabies deaths involve the uncommon L. noctivagans and P. subflavus variants. Passive surveillance systems used by state public health departments confirm that human encounters with Eastern Pipistrelle bats (P. subflavus) and Silver-haired Bats (L. noctivagans) are rare. Neither species exceeded 5% of all bats submitted for rabies testing in the southeastern United States, and Silver-haired Bats account for <12% of all bats submitted in the Northwest.[7] Moreover, the prevalence of rabid individual bats within each species is also low and similar to the estimated prevalence in other, more common, bat species.[8,9] While these surveillance studies are known to produce biased estimates of the true prevalence of rabies in natural bat populations, they should accurately reflect the prevalence of rabies in bat species encountered by the public. In addition, we have determined that although other bat species occasionally are infected with L. noctivagans and P. subflavus variants, the frequency of such spillover is low.[6,10] A survey of rabid "house bats" (i.e., Eptesicusfuscus and Myotislucifugus) in the United States revealed that only 2 of 117 E. fuscus and 4 of 15 M.lucifugus were infected with L. noctivagans and P. subflavus variants.[6] The sample size of M. lucifugus is notably small because this species is rarely found rabid, despite submissions of thousands of individual bats each year.[11] Thus, all available data suggest that L. noctivagans and P. subflavus variants are rare.

Given that L. noctivagans and P. subflavus variants appear to be infrequently encountered, two explanations have been proposed to explain their prevalence among cryptic human rabies deaths associated with bats. The small vector hypothesis suggests a failure to recognize that a bite has occurred when a small bat is involved.[7,12] Absence of a bite history may result from inaccurate documentation when a patient could not be questioned directly or was not lucid. In 18 of 34 cases with no bite history, however, documented contact with a potentially rabid animal could have contributed to an unrecognized bite (i.e., physical contact with bats in 11 cases, bats in residence in 4 cases, and contact with a known sick domestic animal in 3 cases). Among these contacts, 11 involved L. noctivagans and P. subflavus variants. In addition, bites by smaller bat species, such as Eastern Pipistrelles, may be more likely to go unnoticed than bites by larger bats, which may explain why more cases are associated with this species. Wounds from the teeth of small bat species are not easily seen without careful examination,[13,14] possibly leading to the impression that a bite has not occurred.

A second hypothesis was suggested by results from experimental data comparing rabies virus isolates from Silver-haired Bats with those from domestic dogs and a coyote.[15,16] These data showed that, although both viruses replicated equally well in neuroblastoma cells, rabies viruses from Silver-haired Bats replicated to higher titers in fibroblast and epithelial cells, particularly at low temperatures 34°C. Such growth characteristics might facilitate successful infection after a superficial bite. Thus, superficial contact may occur frequently between bats and terrestrial mammals (including humans), but may be unlikely to result in a productive infection unless L. noctivagans and P. subflavus variants are involved (the increased infectivity hypothesis). Although these experimental data suggest that L. noctivagans and P. subflavus variants have evolved increased infectivity, comparison between viruses associated with Silver-haired Bats and dogs (including one coyote) is not sufficient to pinpoint whether increased infectivity evolved in the common ancestor of L. noctivagans and P. subflavus variants and, thereby, is relevant to the noted prevalence of these variants among human rabies deaths.

We proposed a novel test of the increased infectivity hypothesis using a comparative phylogenetic approach in which we characterized transmission patterns from bats to nonhuman terrestrial mammals. Transmission of bat rabies to terrestrial mammals can be viewed as a natural experiment that removes the confounding effects of vector size and, therefore, can be used as a control in understanding transmission patterns between bats and humans. That is, such comparisons take advantage of a fundamental difference between rabies exposures in humans and rabies exposures in other terrestrial mammals (nonhuman terrestrial mammals cannot initiate postexposure prophylaxis even if they are aware of a bite). Given that size of the vector species is not a factor in terrestrial mammal deaths, we can control for the effect of bat vector size in our comparison. Therefore, a disproportionate number of L. noctivagans and P. subflavus cases among terrestrial mammals would be consistent with increased infectivity of the L. noctivagans and P. subflavus variants. If these variants are not overrepresented in terrestrial mammal rabies cases, we can reject the increased infectivity hypothesis, leaving the small vector hypothesis as the most plausible alternative.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: