Import Impact Assessment and Application to the Director General of Agriculture to approve the importation of Rabbit Calicivirus under the Animals Act 1967 and to issue an Import Health Standard under the Biosecurity Act 1993

prepared by the RCD Applicant Group, June 27, 1996, and transmitted to you then. I will call this "the Application Document".

My letter addresses the following issues:

1. Likelihood that the agent of RHDV is restricted to European rabbits on historical grounds.

2.Validity of conclusions from review of the scientific literature that the agent of RHDV is restricted to European rabbits.

3.Validity of conclusions about species-specificity of the agent of RHDV based upon experiments conducted at CSIRO.

4.General comments upon the quality of the scientific work done by the RCD release advocates.

You note that I speak of the "agent of RHDV", rather than "rabbit calicivirus". The use of the term "rabbit calicivirus" is inappropriate, in essence an ambiguous neologism.

Issue (1)

RHDV was recognised initially in 1984 in China among European rabbits there. Within four years the virus had spread at least 7000 miles (southern Spain; page 107 of the Application document) with no evidence of direct spread of rabbits themselves along the path of agent spread (minimum speed of transmission almost 5 miles per day). For transmission to have occurred from rabbit to rabbit, infected rabbits would have had to move 5 miles per day while incubating a lethal disease. A continuous chain of transmission from China to Spain would require ALL of the infected rabbits in the chain moved a MINIMUM AVERAGE of 5 miles per day while incubating a lethal disease.

In the Application Document, pp. 44-45, "factors potentially relevant to the spread of RCD" are discussed. These "factors" include "biting , chewing, sucking insects", "scavenging insects", carrion birds, carnivorous animals, and grazing animals. Nowhere in this discussion is any comment on whether the "transmission medium" (sic) was infected. For "biting insects" or carrion birds or any combination of the "transmission media" to move a minimum average of 5 miles per day in a continuous chain of passive transmission of particles not infectious for the "medium" stretches believability. A far more likely scenario is that an endemic host, such as a bird, travelling while asymptomatically excreting the virus, brought the virus to its epizootic host, the rabbit. Long term asymptomatic excretion is known for feline, pygmy chimpanzee, and SRSV caliciviruses.

The importance of these thoughts about transmission of RHDV are accentuated by the statement on page 45 of the Application Document that "it will be impossible to contain (i.e.limit the spread)" of RHDV. Why New Zealand would want to release something that would be impossible to contain is beyond me.

CSIRO scientists cite testing for antibody to RHDV of rabbit sera collected in Europe before 1984. Some of those sera are said to have contained antibody to RHDV. CSIRO cites this as evidence that RHDV existed prior to 1984 in Europe (see page 107 of the Application Document). CSIRO persisted with this explanation despite my personal caution to Harvey Westbury that other interpretations were possible, namely that the assay may have been detecting group-specific antibody. My caution was validated at the recent Calicivirus Meeting in Reading, UK, in September, at which it was reported that RHDV and the European Brown Hare syndrome virus (EBHSV) share epitopes.

The Application Document uses these questionable antibody results to suggest that a "non-virulent form" of RHDV existed in Europe prior to the recognition of the disease in 1984. This non-virulent form would have, then, transformed (by mutation) into a virulent strain in 1984. (pp. 108 and 143 of the Application Document). Elsewhere in the same document (p. 141), this type of transformation is described as “rare in nature”. When it suits the Applicant Group, things “rare in nature” occur if they support the Applicant Group’s agenda.

Others have summarised the historical information more cogently than I can. Everything taken together, there is no likelihood that RHDV was an endemic disease of rabbits before 1984.

Issue (2)

On page 152 of the Application Document is the statement

"In no case was evidence found for viral replication or clinical symptoms or tissue pathology arising from rabbit calicivirus infection in any of the non-target test animals."

This statement is given in bold lettering, as if we are to be convinced of its truth by its emphasis.

Actually, few data from the literature about testing for host range of RHDV are provided in the Application Document. Much of the previously published information is summarised, although not always accurately, in "Rabbit Calicivirus Disease. A report under the Biological Control Act 1984. August 1996," from the Bureau of Resource Sciences of Australia. Between pp. 22 and 30 of that document a table lists results from several prior publications. Highlights of that table, as I interpret it, include:

a) In the study by Fan et al. antibody responses were noted in inoculated goats, cattle, ducks, cats, dogs, Zao fong birds, red billed leiothrix, fish, guinea-pigs, hamsters, and pigeons. Viral antigen was detected in internal organs in fish and pigeons.

b) In the study by Schirrmeier et al. antibody responses to RHDV were detected in 26 of 135 European brown hares, one cat, all tested rats, all tested mice, and all tested guinea pigs.

c) In the study by Simon et al. faeces from inoculated dogs infected rabbits.

d) In the study by Leighton et al. foxes fed RHDV developed antibodies.

In most of the studies, excluding those above, clinical disease only was the outcome measured. While I have doubts about the quality of a few of these studies, especially because most were conducted prior to the molecular characterisation of RHDV and resulting sensitive assays, and several of the studies were limited to a search for symptoms, it is wrong to describe these studies as "in no case was evidence found".

Issue (3)

The Application Document is notable for its lack of primary data. Some primary data from animal challenge experiments conducted by CSIRO are provided in the BRS report cited above (the publication’s Tables 5 and 6). Let’s consider some facts about those primary data:

a) The indirect ELISA for detection of antibody to the agent is insensitive. The assay is described on pages 161 and 162 of the Application Document. It is unclear from the assay description whether the test antigen was virion purified from infected rabbit liver or recombinant particles purified from baculovirus-infected cells. In any case, the cut-off values for the assay are given as absorbence < 0.3 = negative, 0.3 < absorbence < 0.5 = doubtful, and absorbence > 0.5 = positive. Anyone who has developed ELISA assays will recognise that these cut-off values are absurdly high. Simply speaking, it’s a lousy test. Cut-off values for well designed ELISAs run around 0.05 to 0.1 and cut-off values for baculovirus-derived antigens are uniformly low. A critical reviewer sees that the author of the Application Document is hedging when describing the problem of this cut-off value (third paragraph from the bottom of page 162).

Missing from the assay description is how the cut-off values were obtained. Simply stated, please convince me that the author of the Application Document did not have access to AAHL scientists who developed the indirect ELISA.

One gets the feeling that the people who developed this assay don’t have much experience with solid- phase ELISA’s. [I did my Ph.D. beginning in the 1970’s on the development of solid-phase ELISAs.]

In summary, all of the results utilising the indirect ELISA can be labelled uninterpretable.

Taking the values and cut-off point given, how does one explain the positive test result in chicken #4? Was the blocking experiment conducted to determine whether free antigen could block the assay result?

In reality, cut-off values of the assay should be determined for each species evaluated. In the absence of this proper method one can use the pre-challenge value of an animal as the baseline for comparison with that animal’s subsequent values. After all, that is one reason why one obtains the pre-challenge specimen and why one tests all of the specimens from an animal on the same plate. One is interested in the intra-animal pattern.

When one looks at the indirect ELISA results in this way, 3 of 4 mice have a greater than 3 fold rise in antibody level, of which 2 have a greater than 4 fold rise in antibody level; 2 of 4 rats have a 3-fold or greater increase in antibody level; 2 of 4 bush rats have a 2-fold or greater rise in antibody level; and one spinifex hopping mouse has a 7-fold rise in antibody level. Whether the corresponding titres rose more or less is not known.

b) The results from the competition ELISA also are interesting. For this assay, described on pages 163 and 164 of the Application Document, the cut-off value is given as < 30% = negative and > 30 % = positive. What one is supposed to do with a test value of 30% is not explained.

The assay description does provide information on how the cut-off value was obtained. A total of 263 negative rabbit serum samples [we don’t learn how the sera were known to be negative] were tested. These provided a mean inhibition off 11.69% with a standard deviation of 4.94%. The authors of the Application Document apparently do not understand the concept of "significant figures". The cut-off value is given as three standard deviations above the mean, but 3 * 4.94 = 14.82 and 14.82 + 11.69 = 26.51, so the arithmetic suggests the cut-off value should be 27% inhibition (rounding to the nearest number) instead of the 31% they utilised.

A difference of 4% in the cut-off value will not change the interpretation of many samples, but the apparent arbitrary change from 27% to 31% is unexplained. In addition, using a two sided mechanism for determining a cut-off point, when test results will be interpreted only in a one-sided manner, is not proper statistical method. A result will not be called positive if there is an augmentation of the reaction three standard deviations from the mean.

The Application Document contains results from the competition ELISA testing for only two species, the kiwi and the New Zealand short tailed bat (page 151). It is useful to examine the results from other species, as given in the BRS report. In that report, 1 of 4 brown falcons and 1 of 4 blue- tongue lizards have inhibition values higher than the cut-off point in the pre-challenge serum. Was the blocking experiment conducted to determine whether free antigen could remove the inhibiting antibody ?

Similarly, using the CSIRO cut-off point, 1 of 4 brown falcons (a different individual animal), 2 of 4 silver gulls, and 4 of 4 kiwis develop positive test results after challenge.

If the cut-off point was 27%, as CSIRO said it was calculated, then 2 of 4 blue tongued lizards have pre-challenge positive test results and 1 of 4 long billed corellas, 2 of 4 pigeons, and 1 of 4 wombats developed positive results after challenge (in addition to those animals listed above).

Just as for the indirect ELISA, it is correct to develop a cut-off point for each species tested. This was not done, but we can use each individual animal as a control for itself. Again, that is why one collects pre-challenge samples from animals and tests all the samples from an individual animal on the same assay plate.

Looking at the patterns within individual animals, one sees a greater than 8 fold change in antibody level in 1 of 4 emus, 2- to 4- fold transitory rises in 3 of 4 silver gulls, 2- to 10- fold rises in 4 of 4 echidnas, greater than 4-fold rises in 4 of 4 kiwis, and a 17- fold rise in the only bat that can be assessed.

c) How does CSIRO interpret these findings? Some, such as those from the emu, are ignored (and maybe they should be).For others, they simply are assumed not to be positive. For others, such as for the kiwi, the reactions simply are concluded to be "not evidence for infection"

The logic of the Applicant Group demands that positive test results in the pre challenge sera must be falsely positive and that positive test results following challenge must be reactions to inert virus.

Why not do the experiments?

False-positive, pre-challenge positive test results should remain unchanged by absorption with free particles."Reactions to inert virus" should persist if the challenge inoculum is inactivated before injection.

Again, if the hypothesis in the kiwis is that the antibody response detected in the assay is "indicative only of an immune reaction to the antigenic protein components of the virus" (page 154 of the Application Document), why not kill the virus inoculum and repeat the injection experiments using the same antigenic protein components without the chance of infection? People do experiments like this ALL THE TIME.

d) The RT-PCR and histological investigations are used to "confirm" absence of infection. Simply stated, one doesn’t accept histological examinations performed weeks after inoculation as conclusive evidence of anything. One doesn’t accept histological examinations as providing any conclusive evidence about asymptomatic infections. One doesn’t accept RT-PCR experiments on the blood 7 days after inoculation as conclusive evidence of virus absence. The proper approach is to rechallenge suspect species, such as the kiwi, harvest one to two animals per day, and test all available tissues for virus presence. The rechallenge must include a range of challenge doses.

e) The discounted results from the literature (issue #2) and the CSIRO results actually are quite consistent. Birds and tested rodents appear to be more susceptible to RHDV. Certainly birds would provide an explanation for the rapid movement of the virus and rodents would be a likely endemic host. It is to the benefit of the Applicant Group’s agenda, but not science’s, and in my opinion not New Zealand’s, to ignore these potential clues to the biology of RHDV.

Issue (4)

After just a year of exposure to the RHDV release program I am finding it hard to take the documents provided by the Applicant Group seriously. The Application Document and other documents are so full of half- truths, oversights, convoluted logic, inconsistencies, ignored literature, etc. that trying to comment on all the inadequacies is beyond anyone’s life desire. Just some examples:

a) I have been called "virophobic" [I have a Ph.D. in virology and work with viruses every day.],"ignorant of scientific method" [I have been funded by our National Institutes of Health continuously since my first year as a faculty member and have about 68 publications.], "out for the money" [Show me where there’s money in this process and I’ll go for it.], and "an expert on human caliciviruses, not rabbit calicivirus" [As if my experience with human caliciviruses is irrelevant]. After all, CSIRO’s total time of exposure to the rabbit calicivirus is half mine to other caliciviruses. [ I am a co-owner of patent rights describing the original methods for detecting caliciviruses as used by CSIRO.] These demeaning or marginalizing comments come from CSIRO SCIENTISTS. If I am so discredited, why cite my work in the Application Document?

b)Among other gems from the Applicant Group was the statement that it was safe for people to eat infected rabbits.

c)The "accidental" virus escape from Wardang Island actually was the SECOND escape from the study area. That no one was fired or publicly chastised, let along criminally prosecuted, for the second escape is a marker for prejudice against rational commentary on the RHDV release program.

"Far better a little with accurateness than a heap of rubbish carelessly thrown together." Edward Tyson, FRS (1651-1708)

I could go on, but, as you can see, my tone is becoming more polemic, and surely, therefore, less convincing. New Zealand has the opportunity to proceed cautiously. New Zealand should not be intimidated by statements that the virus will be introduced illegally, so it’s best to release it legally (page xv of the Application Document). I have laid out above some clear examples of inadequate scientific justification for the conclusion that the agent of RHDV is species-specific and given specific experiments that should be conducted to strengthen such a conclusion if true.

From the very beginning of my communications with CSIRO a year ago , I offered suggestions on how an approved release could be improved. Because I also offered criticisms of the work that had been done, CSIRO scientists decided it was more expedient to attempt to marginalise me and other credentialed critics, than to do the work properly.

The activities of the Applicant Group have been an embarrassment to science.

I apologise that this document does not have all of the citations that are needed to back up every one of my statements. I am ill and about to go and see a surgeon.