To the Editor: We are aware of Capucci and Lavazza's excellent work. Indeed, one of the best characterized calicivirus genomes is that detected in rabbit hemorrhagic disease (RHD); however, the virus' infectivity, pathogenesis, modes of transmission, reservoirs, survival in nature, host of origin, virulence factors, number of neutralization serotypes, and multispecies infectivity are poorly characterized. Propagating this virus in vitro could provide insight for addressing questions relevant to caliciviruses that cannot be propagated in vitro.

We are unclear about the confusion regarding Norwalk virus and feline calicivirus (FCV). Both are caliciviruses. Norwalk virus is a human pathogen. FCV is in a different genus (1) that includes strains infecting humans (2). We know of no documented FCV infections in humans nor of detailed studies to search for such occurrences, although some evidence suggests the possibility (3).

Capucci and Lavazza's remaining questions address the etiology of RHD, diagnostic reagents, and possible human infection. They report nine laboratory workers as antibody negative but do not report test results on persons at high risk, such as rabbit farm workers, nor do they mention having positive control human or primate sera. Koch's postulates have been fulfilled for RHD: a parvovirus was isolated in vitro and was cell-passaged 15 times; at a second laboratory, the parvovirus was identified in materials causing RHD (4,5). In Europe the parvovirus etiology for RHD was deemed hypothetical but has not been refuted on a scientific basis. The calicivirus consistently identified in European materials has not been isolated in vitro, and Koch's postulates have not been fulfilled. Are the parvovirus-associated outbreaks of RHD in Mexico and China (4,5) and the calicivirus-associated RHD outbreaks in Europe identical disease manifestations of two different viruses? Is RHD multifactorial requiring two or more agents? Is RHD caused by only a calicivirus or only a parvovirus? A calicivirus and a parvovirus can be isolated in vitro from the same fecal sample of a sick rabbit (N. Keefer, D.E. Skilling, A.W. Smith, unpub. data).

Our comments on RHD diagnostic assays referred to those used in Australia (6,7) to screen humans and experimentally infected animals to support legalizing the spread of RHD in Australia and New Zealand.

Public health protection requires prudent avoidance of pathogens associated with risk of adverse outcome, not necessarily proof of causation (8). In this context, human health risk for RHD goes largely unaddressed. The deliberate introduction of a new disease agent (RHD) known to cause death in mammals requires prudence rather than proof of human illness, especially when the scientific literature includes reports that the agent has induced antibody reactions in a wide range of mammalian and avian species (6).

Mead et al. (9) conclude, "No significant association between exposure to RCV and subsequent bouts of sickness could be demonstrated." Their recorded data do not support a statistically significant risk of illness because sample sizes in the monthly groups were too small for any meaningful interpretation. Mead et al. (9) state a "lack of any serologic reaction of the respondents," but a 50% cut-point was used for the competitive ELISA test, and some individual sera were repeated up to six times with percent inhibition reactions ranging from approximately 1% to 44% in one instance and 12% to 100% in another. Results were selected from these laboratory data and reported "lack of serologic reaction."

We derived our findings from data obtained under a freedom of information request. Mead et al. used the same data to support an opposite conclusion. Opposing conclusions "red flag" the quality of the study. In summary, the reporting of negative results of such a study cannot be used to support the important biologic, health, and political conclusion that humans are not at risk from infection with RHD.

We encourage a well-designed longitudinal study of persons at high risk of RHD exposure to answer conclusively whether RHD has infected humans. If "the rule" means that most humans exposed to RHD would become infected, we agree with Dr. Capucci "that infection is unlikely to be the rule," but transmission of equine morbillivirus, Rift Valley fever, and H5N1 influenza to humans is also "unlikely to be the rule" (10).

Top

Related Links

• More Letter Articles