The aim of the NZ Association of Scientists Inc (NZAS) is to promote science in New Zealand. It publishes the New Zealand Science Review which focuses on science and science policy in New Zealand.

While generally The Association does not scientifically scrutinise or evaluate specific scientific issues, the process by which any such evaluation of major issues is carried out in New Zealand is a concern. Our concern is not only with the consequences of particular issues, but also with the image and role of science in New Zealand.

Following enquiries into the Application to import Rabbit Calicivirus Disease, the NZAS Council at its 17/10/96 meeting decided to prepare a submission. Given the wide ranging implications of the Application for New Zealand, I as Association President was designated the task of co-ordinating preparation of the submission.

The reason for Association concern is that we believe the application to import RCD virus preparations and use them as a biocontrol agent promotes premature action where scientific knowledge for necessary safety and effectiveness is lacking and the controversial and compromising process of evaluating it threatens the credibility of all scientific input into crucial issues.

The NZAS position

We strongly oppose approval of the Application on three major grounds:

1.That the approval process is inappropriate, secretive, and its credibility is seriously compromised with the general scientific community and the public; it lays the Government open to potentially massive open-ended liabilities in the event of major adverse consequences, and it exposes New Zealand to risks far outweighing the benefits potentially gained.

2.That the Application (IIA) itself is unbalanced and deceptive, arguing by special pleading in crucial areas of uncertainty resulting in spurious assurances of safety.

3.Currently, RCD is totally unsuitable as a potential biocontrol agent:

(i) because of a lack of critical knowledge about the virus itself, its role in rabbit haemorrhagic disease, its virulence mechanisms, its capacity to evolve and the consequent risk of this to animal and human health;

(ii) because of its uncertain effectiveness and ecological effects

(iii) because if any of the major risks to humans, farm animals or wildlife based on these uncertainties proved to eventuate, the consequences for New Zealand could be catastrophic.

The NZAS recommendations.

The NZAS recommends that the application to import RCD preparations be denied.

Any approval in the future should involve wide public discussions such as under the Hazardous Substances and New Organisms Act 1996.

Further steps should be taken to reduce the risk of unauthorised importation of the virus accidentally or intentionally.

It further recommends that, unlike Australia where the virus is now endemic, in New Zealand the appropriate stance is to delay the issue of import and release of RCD until research removing the major uncertainties regarding its properties and effects becomes established. Its effects in Australia should be monitored for a period sufficient to establish absence of harm (at least 10 years) and research established into related caliciviruses and their vectors which are potentially present in New Zealand fauna.

Improved protocols for vaccinations, eradication and followup targeted release following an unauthorised release should also be developed.

The New Zealand Association of Scientists requests the opportunity to present their position at public hearings.

Amplification of the reasons for our position follows.

Dr C. H. Sissons

President
NZ Association of Scientists

REASONS FOR THE NZAS POSITION

1. THE APPROVAL PROCESS

The decision to import and use an exotic lethal virus such as RCD has widespread implications for all New Zealanders. It is a political decision, not a technical decision on a simple industrial matter. It is an abdication of political responsibility to require the Director-General of MAF to make such decisions on behalf of Government and New Zealanders under outmoded and outdated legislation (the Biosecurity Act 1993 and Animal Biocontrol Act 1967) which is quite unsuitable for considering pathogen and virus importation and which is in the process of being superseded by the Hazardous Substances and New Organisms Act 1996 giving much wider public scrutiny to such issues.

The original decisions taken by MAF to keep secret the identity of the so-called “peer” reviewers and reviews and then restrict public access to the review process we believe to be unacceptable. To change the review process two weeks before submissions closed and to refuse to extend the deadline - all this on an issue of national significance during an election campaign and in a period of caretaker government - seems to us to be quite improper, indeed unconstitutional as proper political oversight and decision making on a highly contentious process will be minimised.

Direct input into the debate over the issues seems to have been confined to anonymous MAF-co-ordinated officials and scientists and not extended even to the wider scientific community. Neither the New Zealand Microbiological Society nor the Royal Society has been involved. The Health Research Council Virus Research Unit has not been approached despite the fact that its former Director, Dr A.J. Robinson, is heavily involved in issues arising from the Australian outbreak as Chair of the Science Subcommittee Monitoring and Surveillance Program. I have directly contacted him on the issue of potential virus host-switching.

The MAF processes and analysis have been compromised, not only by its restrictive attitude to public debate and lack of encouragement of free scientific debate, but also by the appearance of close collaboration with the Applicant group, the suggestion that criticisms by the original reviewers were not deemed substantial enough to require rebuttal in the final Application, and the dismissal rather than rebuttal of key concerns raised by experienced US Calicivirus researchers Professors Smith and Matson, on the grounds of being disagreed with by “independent” unnamed virologists (J.Guthrie, Chair of the Rabbit Biocontrol Advisory Group, Evening Post, 2/10/96).

Even after requesting the names of the reviewers, with just one exception, only the names of their organisations were released, providing no evidence for the competence of the reviewing panel. The statement by Debbie Gee, Communications Manager, MAF, on Radio New Zealand News Review (17/10/96), that the original reviewers if quoted would be removed from the review process, indicates an extraordinary willingness to suppress potential sources of criticism. Even worse, this has in fact been followed by the removal from the review process of all of them and their replacement by four still anonymous reviewers. The role now being fulfilled by the other government departments involved in the “peer review” process (Environment Conservation, Te puni Kokiri and Health) is still unclear.

This history has destroyed the credibility of the MAF evaluation process, despite claims that the issue is still open. The only way to restore credibility is to refuse the Application and take steps to properly and openly establish the scientific background necessary to evaluate and control the disease.

The liability which government faces from possible catastrophic consequences of unknown risk level if it authorises an ill defined, exotic, lethal virus in the face of strong opposition and warnings, may be extremely high - particularly if the lethal factors get transmitted to stock or humans. Defences against liability under the Biosecurity Act 1993 are unlikely to protect against a biased or negligent approval process.

2. THE APPLICATION

The Application Group Document despite its volume and coverage has in places either glossed over concerns based on crucial gaps in knowledge of virus evolution and of RCD, its properties and potential effects, dismissed them without substantiating evidence or by use of special pleading. Where there are genuine uncertainties, the Application Document tends to describe unfavourable possibilities for the Application as “speculative” and favourable possibilities as having “no evidence of” harm. Consequently the Application is unbalanced and in some instances quite misleading. These areas mainly relate to gaps in knowledge concerning RCD properties, its virulence mechanisms, its potential evolvability and associated risks. There are issues which need to be clarified as a prerequisite to safe use of any biocontrol agent, especially viruses. Some areas where knowledge of RCD is lacking or where there is serious disagreement among virologists are described below (3).

3. GAPS IN KNOWLEDGE OF RCD MAKING IT UNSUITABLE FOR USE AS A BIOCONTROL AGENT.

3.1 The virus itself is ill-defined, its role in rabbit haemorrhagic disease is not finally established, it causes death by unknown virulence mechanisms.

There are still questions over the role of RCD as an etiologic agent of rabbit haemorrhagic disease, mainly because of the inability to culture it in vitro in cell lines to confirm the criteria embodied in Koch’s principles of causation. The infectious agent is prepared from rabbit liver extracts by relatively non-discriminatory techniques and are therefore of uncertain purity. The disease-causing agent could in fact be another, adventitious virus.

The tests for lack of adventitious viruses outlined in the Disease Risk Analysis involved testing for cytoxicity in cell lines and clinical disease in other animals. RCD itself is negative in these tests and other similar viruses might also prove negative. The ability to culture RCD in cell lines is necessary for its characterisation, purification and determining its role in disease etiology. It is also necessary for determining its genetic variation and for preparing vaccines useable in non-rabbit species and to develop better analyses of replicating virus to test for infection of non-rabbit species.

Mechanisms causing death are described at the (histo) pathology levels only in the application. These are acute liver necrosis and catastrophic the disseminated intravascular blood clotting. How the virus causes these effects is not even discussed although an analysis of virulence mechanisms was required of the Application. This is a key issue. Virus replication in the liver may be the cause of the liver damage but the mechanism causing disseminated blood clotting is unknown. Other diseases (eg. Facial eczema) causing acute liver damage do not cause such dramatic coagulation defects so very likely there is a specific virus- induced mechanism involved.

The RCD genome codes for a small number of genes, including the capsid protein, a helicase, an RNA polymerase, a cysteine protease and at least one protein of unknown function and effect. The protease and unknown protein may have a role in the clotting defect. Whichever gene causes the fatal disease, if it transfers to other viruses, as viral genes sometimes do, it could potentially cause the fatal disease in other organisms in the second virus’s host range without host-switching of the RCD virus itself. To suggest for this unknown virulence mechanism that “there is no reason to suppose it would cause the same [lethal] pathology in humans or other species” (Application p52) is quite fallacious. There is every reason to suppose that it might. This is one of the examples in the Application of special pleading.

Rabbits younger than about 8 weeks do not succumb to the virus, again for completely unknown reasons which also need to be established as they impact directly on the virulence and effectiveness of the virus.

3.2 Implications of the sudden and unknown origin of virulent RCD.

The origin of rabbit haemorrhagic disease is unknown, arising in China in 1984 after transfer there of European rabbits from Germany. It either switched host from an unknown host reservoir in China or spontaneously increased virulence from a non-virulent form of RCD (if it is RCD) in Chinese rabbits or by transition between rabbit strains of the non-virulent form now found in rabbits in Europe.

Hence the present virulent RCD must have arisen either by host switching or virulence-switching mechanisms which are mechanistically equivalent - mutation combined with recombination. The significance of this sudden and recent event is that it does in fact prove a degree of instability of RCD and raises a question over the future stability of the virus. This is glossed over in the Application.

3.3 Evolvability, Host-range and Host-switching.

The safety of the virus clearly depends on its being and staying specific to rabbits. This involves issues of mutability, virus evolvability, host-range and host-switching of virus and virulence genes. These are areas where there is major disagreement between virologists and the evidence necessary to resolve the issues is lacking. This is particularly true of viral evolution issues. The dismissing of these controversies and consequent unsound assurances of safety is not only a feature of the application but of the debate from official NZ sources in the NZ media.

Technically it is almost impossible to distinguish a cryptic host range from host switching, especially in cross-sectional studies. The history of such studies demonstrates this. Even feline parvovirus, thought to be an example of host switching to dogs, apparently could always replicate in dogs, albeit poorly. A mutation allowed the virus to increase its virulence such that the virus could maintain itself in dogs.(A.J.Robinson, personal communication). If this situation of low level replication in other species pertained to RCD, the current host range testing carried out almost certainly would not have picked it up. However, with this caveat, we agree it seems likely that the current host-range of RCD is confined to the European rabbit and possibly the blue fox (farmed Alopex lagopus) most of which have antibodies against RCD (A.J.Robinson, pers. comm. describing unpublished studies of D. Gavier-Widen).

The evolvability threat of RNA virus is a particular area of dispute. They mutate at a rate of 1 in 10 to the power of 4 base pairs, i.e. about one mutation every replication - the eigen limit for avoidance of error catastrophe. This is a rate at least 10 to the power of 3 higher than DNA viruses and about 10 to the power of 4 times higher than rabbits. It is simply untrue to portray the genetic volatility of Calicivirus as the same as other viruses (Application 6.10.9,p.144). The consequences of this high mutation rate are still in the realm of speculation and it is necessary for the applicant group to demonstrate safety not demand proof that unfavourable hypothesis e.g. induction of host-switching, are correct (p142) . In discussing the genome variability they have also omitted to mention that clonal diversity of a pathogen is minimised during an epidemic phase, as is the case with RCD in Europe and now Australia. The particular selective pressures which conserve RCD protein sequences are not known; even the virulence factors are not known, so the effect of the high mutation on virus genome variability is not predictable. Likewise the consequences of genome variability on factors such as virulence and host-switching is also an area of controversy where the evidence is simply lacking.

A paper of direct relevance, "evolvability of emerging viruses," was presented by Col.Donald S. Burke, Assoc. Director for Emerging Diseases and Biotechnology at Walter Reed Army Institute of Research at the MZMS-ASM Conference (1/10/96). He analysed the mutation and recombination rates of different RNA viral groups of viruses to identify special evolvability threats to human health. To quote the abstract:

"RNA viruses pose an important emerging disease threat to humans, largely because they are highly evolvable. Mutation rates of RNA viruses are a million fold higher than that of their hosts. All known RNA virus families evolve through mutation with intermittent recombination of genetic information."

Dr Burke suggested, as have others, that some RNA viruses can cross high level phylogenetic boundaries (vertebrates, insects,plants). Influenza virus is an extreme example, being a chimera including bird, pig and human genes. Recombination is an especially important strategy allowing phylogenetic broad-jumping and adaptation. His position on Caliciviruses was that not enough is known to evaluate their recombination rates and hence evolvability hazard.

The inadequacy of current knowledge in the area of viral evolution where host-switching obviously does occur means that assurances of safety in respect of host-switching are spurious. They are simply best guesses of undefined probabilities (Application 6.10.9,p.142). Further knowledge may indicate, for example,that we do have significant cause for worry about feline calicivirus (Application p.144) which is already endemic in NZ. RCD is an exotic, lethal virus that we do not have to introduce.

3.4 Protection measures.

The known extremely high stability of viral infectivity in the environment, the unexpected escape from containment in Australia and the wide and rapid spread of the virus in Australia, prove an inability to either contain it or to eradicate it from a large feral rabbit population. Therefore, there is no feasible control of the virus once released. Paradoxically, this is a major argument used for importation; that these properties mean that NZ cannot keep the virus out when in fact they are a major reason for keeping it out.

There is currently no vaccine apart from rabbit liver purified virus which is unacceptable in a non-rabbit species and especially humans. A useable vaccine would seem to be an absolute prerequisite for contemplated use of a biocontrol lethal virus. The high rate of viral variability may compromise the longterm or even short-term effectiveness of such a vaccine. This situation alone makes RCD unacceptable in a biocontrol agent, especially one based on a lethal virus.

3.5 Ecological consequences, effectiveness as a rabbit biocontrol and priorities.

The ecological effects of the disease, especially prey-switching, are uncertain. We believe that these issues have been discussed in detail by other organisations.

Although the Application claims, apparently conservatively, about a 40% overall rabbit kill rate to be likely, this is the rate that is assumed in Australia after a year’s experience of release. It is likely to be lower in NZ as the large warrens facilitating transmissions in Australia are mainly replaced by a less dense mode of living in undergrowth in NZ. The Australian experience is of highly patchy effects with greatest effectiveness close to bodies of water, conditions uncommon in the most rabbit prone areas of NZ. The effects on reducing Australian rabbit populations are estimated to last about 5 years before resistance restores the populations (S Thomas presentation NZMS/ASM Conference). Rabbits younger than 8 weeks are not killed enhancing the rate of development of resistant rabbit populations. Rabbits breed rapidly. This type of biocontrol has inevitable needs to be combined with other control and land management to maximise its limited effects.