Memorandum by Novartis UK Ltd
1. Working practices and environmental safety
of work with GMOs in research is governed by three basic EU directives
89/769, 90/219 and 90/220 and many derivatives of these.
89/769 deals with workers safety;
90/219 deals with the environmental safety of
work done with GMOs in contained environments: the laboratory,
the glasshouse, animal houses, factories, hospitals, etc;
90/220 deals with environmental and human health
safety aspects of work done on GMOs outside the commercialisation
of GMOs. This is not correct. Most of the regulatory work generated
by 90/220 covers research done with GMOs in the field.
2. FIELDS OF
Medical research involving genetically
modified organisms. A key case is the use of genetically modified
mice as disease models for human diseases.
Medical research on xenotransplantation.
Pharmaceutical research on human,
animal and plant pathogens.
The basic study of human, plant and
Research on plants, animals and micro-organisms
involved in agriculture and food production. The most visible
targets of such research are agricultural crops, but micro-organisms
such as yeast, bacteria and fungi are as important targets.
3. All the above are fields of applied research.
Most of the use of genetically modified organisms in a research
setting happens in basic research projects, the only aim of which
is to obtain insight into the development, behaviour, ecology
of living organisms.
4. A special case is the use of genetically
modified organisms in education. When 90/219 was developed (1988-90)
it was almost inconceivable that experiments with GMOs would soon
become part of the basic curriculum in higher education in life
sciences. Undergraduate training in biology without experiments
using GMOs is almost unthinkable today. All training experiments
of course are done with micro-organisms at the lowest risk category.
But because European regulation is based on the premise that the
process of rDNA techniques might induce unique risk factors not
encountered with any other form of biological experimentation
(against the scientific consensus) there is no risk-free
category for GMOs. This meant that in practice work with micro-organisms
that had been the standard work horses for biology training suddenly
was declared potentially dangerous, against all evidence to the
contrary, and governed by new rules.
5. In most EU member states, awareness of this
situation was so low that it took years for government and academic
authorities to realise that much of their science training had
suddenly become technically illegal. They had not foreseen budgetary
measures to implement the safety standards set up by the new regulation.
Most EU member states have not foreseen any training of their
new students in life sciences on the regulatory framework under
which they will work during their professional careers.
6. Many of the safety measures prescribed in
90/219 and its successors have had relatively little impact on
the work done in contained environment in a company setting. The
containment measures and working procedures prescribed in this
directive have to a large degree followed those already known
to scientists such as the GLP (good laboratory practices) and
GMP (good manufacturing practices) regulations in force in the
pharmaceutical industry. For work with potential dangerous organisms
such as lethal pathogens, the isolation requirements closely follow
those already used successfully in earlier legislation. It has
been asked why there had to be an additional layer of regulation
for these activities, since they basically lay down an additional
layer of paperwork covering the same measures.
7. It is different for the application of 90/219
in academic research. The bulk of this research has traditionally
been done with organisms that are generally considered safe to
work with, such as laboratory strains of the bacteria Escherichia
coli and the bakers yeast Saccharomyces cerevisiae. Suddenly much
of this research became technically dangerous, not because there
had been an incident or any other indication of possible risk,
but because there had been a new regulation, which is based exclusively
on a hypothetical risk evaluation not supported by any experimental
evidence and an extreme interpretation of the precautionary principle.
8. Most experimental releases in the field in
an industrial setting involve the development of genetically modified
crops, with the ultimate goal of developing new commercial varieties.
Traditionally, new genotypes are tested extensively against relatives
of proven performance. The introduction of directive 90/220 introduced
new rules on isolation of the material, on its disposal and on
post experiment monitoring. This has made experimental evaluation
of genetically modified crops much more expensive. It also required
that agronomic research stations that accepted GMOs in their work
had to develop new safety infrastructure.
9. The problem is compounded by the fact that
much of the field research on the agronomic behaviour of crops
is not done by the companies themselves, but has traditionally
been subcontracted to private or public experimental stations
that were not technically ready to work under the new regulations.
10. Field research with GMOs in public research
institutions has been severely affected by the directive 90/220.
When this directive was negotiated little attention was given
to the fact that much of the early research in the field was done
in public research institutions. Legislators worked from a silent
assumption, never publicly challenged, that almost all this work
would in the future be done by companies that are used to work
under the severe constraints imposed by a restrictive regulatory
framework. These companies often have specialists in place to
handle the administrative load of compliance with regulations,
including the monitoring of compliance with requirements.
11. It was not much noticed that there was great
excitement among public scientists about the opportunities offered
by the use of genetically modified plants and microorganisms for
basic field studies in ecology, physiology and population genetics,
among others. Academic research is ill equipped to work under
heavy regulatory constraints. Its decentralised nature makes it
difficult to develop the expertise in handling the regulatory
requirements, and the constant funding problems force hard choices
between investing in research or in regulatory compliance. The
result has been that fieldwork with GMOs in academic institutions
still represents only a tiny fraction of total research done on
them. What is done is mostly concentrated in a few research institutions
that have invested in setting up a regulatory compliance infrastructure.
12. This has direct consequences on our knowledge
of potential risks associated with GMOs. The bulk of this research
is done in contained environment, or in the form of computer models.
Very little research is actually done in the field, at least in
part because of the restrictions imposed by the regulations. This
leads to the situation where the regulations make it very difficult
to test their own validity in real life situations.
13. This is governed by 90/220, part C. No GMO
can be released on a commercial scale without a comprehensive
risk assessment. This automatically includes series of field trials
done under the requirements of 90/220 part B. Over the years several
new directives have refined the procedures associated with 90/220.
It sets up an administrative process to assess risk, based on
the scientific method.
14. However, it has procedural flaws that have
become clear over the past three years, as the first generation
of genetically modified crops have moved through the process.
After the end of the risk assessment process, there is no provision
for evaluation of its conclusion in the broader context of a cost
benefit analysis or of a comparative risk analysis.
90/220 is a risk assessment, not
an environmental impact analysis. The technical analysis of an
application for field release does not take into account the potential
environmental benefits of a new GMO.
It is scientifically impossible to
prove the absence of risk (although several environmentalist groups
are asking just that). Any application for introducing to the
market will therefore come through the scientific risk evaluation
process with a report stating that the risk is small, even negligible,
but never zero.
In the European system, any file
that has not received unanimous approval in the administrative
process (and the positions of Austria and Luxembourg make unanimity
impossible for most files) moves on to the political level for
decision making. Nowhere in this transfer is there a place where
the risk assessment can be set in the context of the total impact
of the new crops. There is in other words no provision for companies
to present their products in such a way, as the only dossiers
they can introduce in the system officially are the risk assessment
files linked to directive 90/220 and the novel food regulation
97/258 (see below).
As political decision-makers never
get to see an environmental cost-benefit analysis as a basis for
their deliberations, the only basis they have for decisions is
the concept of acceptable risk. There is also no provision
for a comparative risk assessment of the hypothetical risks of
the new product against the risks associated with existing products
15. This leads to counter productive decision-making.
For example, the present controversy over insect resistant crops
totally disregards the fact that these crops were developed to
widen the options for insect pest control, and that this in most
cases results in reductions of pesticide use. There is no place
in the European decision making process to evaluate these different
options. In fact, GMOs are treated from the premise that they
have no potential for environmental benefits at all, which is
absurd. The large-scale introduction of insect resistant cotton
in the USA has reduced the use of chemical pesticides on this
crop by one-half to two-thirds, depending on the region. Nowhere
in the European evaluation process is there room to find out if
some of the GMO crops could do similar things to European agriculture.
16. Another major flaw of directive 90/220 is
that is does not formally lay down rules for monitoring after
commercialisation of a genetically modified crop. There is a broad
consensus in the field that good monitoring makes good business
sense as well as being environmentally sound. Companies have invested
heavily in these products, and want them to be effective for a
very long time.
17. Post marketing product stewardship is not
new to the seed sector. Much of the so called traditional breeding
over the past half century focused on developing varieties with
new resistances against pests and diseases. It is well known among
plant breeders that sooner or later a disease will overcome the
resistance of the crop. This is a logical consequence of the theory
of evolution by natural selection. Therefore, breeders and seed
certification bodies have for a long time had networks for gathering
information on the behaviour of varieties, to provide them with
early warning if a resistance shows signs of weakening. This often
gives them time to introduce new resistance genes in new varieties
before the problem starts having an economic impact. There is
every indication that the same system can be used with success
for genes that were introduced through modern gene technology.
18. The absence of monitoring schemes based
on existing knowledge rather than on theoretical scenarios based
on an extreme interpretation of the precautionary principle leads
to the development of entirely new bodies for oversight of monitoring
by some member states, and to demands for "monitoring"
that have in fact no relation to the activity as such. For example,
in France a body was set up to evaluate monitoring schemes for
insect resistant maize after approval was given for commercial
production. The bulk of the "monitoring" work proposed
was in fact basic research into soil bacterial ecology, not monitoring
of insect resistance management.
19. The novel food regulation (97/258) and its
dependant directives are the logical complement of the environmental
directives on GMOs. Nationally, 97/258 applies to all novel
foods, regardless of how they are produced, but most of the
text, and almost all the public debate surrounding it indicates
that it was written essentially for GMO derived food.
20. The EU novel food regulation is based on
work of the OECD Group of National Experts on biosafety in biotechnology.
This group produced a consensus document based on the notion of
substantial equivalence. This notion, well documented in US regulations,
was used as the basis for deciding whether a GMO or a derived
product should be treated as a novel food or not, and should be
subject to additional regulation. When the notion of substantial
equivalence was introduced in the EU novel food regulation though,
it was given an interpretation that is much narrower than that
in the USA. Moreover, different standards have emerged for the
determination of equivalence in the debate of safety evaluation
(which is the basis for the novel food regulation), and in the
debate over labelling.
21. For labelling purposes, substantial equivalence
has essentially come to mean chemical identity. Moreover, chemical
identity is defined as absence of difference in gene content as
determined with state of the art technology. Analytical methodology
is an evolving field, and setting a zero tolerance limit in such
a situation is always a dangerous regulatory approach. It means
that a food producer can be selling food that is non-GMO today,
and may with exactly the same ingredients be found working with
GMOs tomorrow. The status of food products, especially processed
food with relatively long shelf life, may well change between
the day they were produced and the day they are tested, not because
the food has changed, but because the detection techniques have.
This puts food companies in a situation where they face random
public outcries about their products that have nothing to do with
safety nor with their compliance with the law.
22. The novel food legislation is unenforceable
as long as there is no agreement on threshold values. The present
situation sets a zero threshold value, and this is technically
impossible to police. Modern DNA detection techniques (especially
PCR) allow detection of the presence of a foreign gene in a product
made with say soybean, even if beans containing that gene are
only present in a concentration of one in 10,000.
23. Food is never "pure" by these
standards. In practice, the PCR technique gives many "false
positives". For example, transporting non-GMO soybeans in
a ship that contained GMO soybeans on its previous shipment will
almost certainly give a positive signal in PCR. The dust in the
ship (which is mainly composed of fragments of seed hulls) contains
enough DNA to show up in such tests as positive.
24. A further complication is that PCR, although
extremely sensitive, is not easily quantifiable. If a threshold
valuesay one per centwas introduced, PCR is not
likely to be able to distinguish between food that contain 0.5
to 1.5 per cent GMO derived material.
25. As long as no threshold values are introduced,
and reliable measuring methods agreed upon, the only way for food
producers to avoid random scandals about their products is to
label almost all their products with the label "may contain
GMO products" or some equivalent term. This effectively destroys
the basis for the labelling regulation: preservation of the consumer's
right to choose. It has become clear over the past two years that
the only practical way to provide consumers with real choice is
by the introduction of threshold values, in a way similar to what
is done with agrochemicals in organic food.
26. It has recently become clear to many associations
of organic farmers that they are themselves not able to comply
with present standards to guarantee the GMO free status of their
products. This provides an additional push towards the negotiation
of a practical threshold value, and an agreed technique for policing
27. An alternative, which is used in practice
now, is to essentially do away with controls, and provide segregated
food on the basis of certificates of origin of the raw materials.
This is a dangerous situation. A regulation with high exposure
to public concern, in which large amounts of material will be
categorised purely on the basis of a paper trail, which is not
certified by strong policing activities, is an open door to the
development of fraud.
28. Finally, threshold values are indispensable
for seed certification. Modern seed varieties are sold with an
unprecedented level of purity, even before the advent of biotechnology.
Depending on the crop certified seed has to be from 95-99 per
cent pure (i.e., of the variety mentioned on the package). Seed
lots that fail this requirement are rejected. In real life no
seed lot is ever 100 per cent pure though, nor need it be. There
are technical limits to the seed producer's capability to keep
elite seed separate from other seed lots, while maintaining a
reasonable price. But as it is impossible to deliver seed that
is guaranteed 100 per cent free from any GMOs, it is by definition
impossible to deliver food up to that standard.
29. The United Kingdom and France were the member
states that had systems in place for regulatory oversight of research
with GMOs before 1980. This is a logical consequence of the longstanding
tradition of world class research in the life sciences in these
two countries. The introduction of a common regulatory framework
for environmental and worker safety by the EU did not present
a major problem in these countries, as expert bodies were already
in place and functioning.
30. The UK, Denmark and the Netherlands were
the only EU member states with a novel food regulation in place
before the introduction of the EU directive 97/258. Partly this
was the result of British and Dutch frustration with the slow
progress made in the development of a common European regulatory
framework for novel foods. These three countries have maintained
the lead in the evaluation and introduction of novel foods containing
GMO derived raw materials.
31. Safety issues related to GMOs fall in two
distinct categories: those related to human health, and those
related to the environment.
32. Issues related to human health are essentially
universal. It would therefore be logical to ensure adoption of
common food safety evaluation methods worldwide. In practice the
ongoing controversy between the USA and EU about novel food containing
GMOs makes it unlikely that such common standards will be reached
soon. This has further consequences for international trade. It
means that the USA and the EU are at odds with each other over
food safety in the international arena. There is an urgent need
for the EU and the USA to come to common grounds on food safety
issues in general, because in this area it is all too obvious
that cultural differences and/or trade considerations are sometimes
translated into safety terms to justify different standards. Ideally,
food safety standards, with GMOs as well as with other foods,
should be the subject of world-wide legislation.
33. Issues related to the environment are more
diverse, as the behaviour of organisms is always the result of
the interaction between their genetic make-up and the environment
in which they develop. However, environmental safety issues related
to the introduction of crops do transcend individual countries.
The current dominant level of regulationthe EUseems
to be the appropriate one from the scientific point of view. It
is also the only one that would not impede the principles of the
34. The seed industry is hit with high
and unpredictable costs. When seed companies introduce their files
for approval, they are already upscaling seed production, which
is a slow process. If the application files are delayed, a full
year's seed may be produced for nothing and has to be destroyed.
It is sometimes assumed that such seed can then simply be used
elsewhere. This is usually not the case. The varieties developed
with a particular gene (eg., insect resistance) are different
for different regions, as they should be: their growing environment
is different, and they have to be locally adapted. On the other
hand, companies can not afford to delay the production of seed
until they are absolutely certain of having the permission to
sell. This would also delay sales by one or two years, and could
destroy any development lead they have spent a large amount of
money to build.
35. In the long run this cost is bearable for
the seed companies. If the current uncertainties continue to exist
for many more years, companies may become more and more reluctant
to invest in GMO seed. They will still sell seed to the farmer:
the "traditional" varieties sold today.
36. The hardest hit though is the farming
sector. EU farmers are squeezed between the high cost structure
of European farming and the demands to reduce agricultural subsidies.
In the long run, the solution has to come from a combination of
two factors: reduction of the cost structure, and creation of
higher value output.GMO crops are delivering important openings
for the achievement of both these objectives.
The introduction of crops designed
to have better pest and disease resistance, and more rational
herbicide use, have an immediate impact on production costs for
the farmer. The cost reduction is much higher than replacement
cost of the agrochemicals saved. Most of the cost of crop protection
is not in the products used, but in the investment in equipment
and time to apply them to the field. Major cost saving are possible
The introduction of crops with altered
output traits is changing several commodity agricultural crops
into groups of speciality crops. The best-known European example
has been the creation of a potato variety with a specialised starch
composition that makes it uniquely suitable as a raw material
for the chemical industry (high amylopectin potato). Scientists
of the Agricultural University of Wageningen in the Netherlands
developed the variety. Such crops automatically command higher
prices as they open new market niches. Similar projects are under
development (or already commercialised in the USA) with the vegetable
oil composition of major oil crops: oilseed rape and soybean.
The attractiveness of these crops lies in the fact that they can
prevent several expensive post harvest processing steps, and thereby
bring part of the added value that was previously created in the
factory back to the farm.
37. An "industry sector" that is often
overlooked when considering the impact of legislation on high
technology is the "research industry". As our society
moves from material to knowledge production as the engine of growth,
the research sector becomes an ever more important component of
our overall economy. Intellectual inputs can only reach their
expected levels of economic impact in terms of national growth
if they can be properly protected and valued, and if they can
be translated in economic realities in the form of improved goods
38. The first of these requirements has required
much more time in the EU than in its major competitors. The EU
directive on intellectual property rights for biotechnology has
only recently been adopted by the European Parliament, more than
10 years after its initiation.
39. For the second requirement, a stable regulatory
environment for R&D is essential. The continuous controversy
over GMOs has created great uncertainty among the scientific community
about the long-term commitment of the EU to its declared objective
of making biotechnology into one of its key development areas
over the coming decades. This leads to a brain drain as well as
an investment drain. Even today, investment in biotechnology in
the EU lags behind our main trading partners. Investment in a
knowledge intensive sector like this translates directly into
high added value employment. The employment generates the proprietary
knowledge that generates new income.
40. Especially the public research institutions
are hard hit by the complexity and instability of the regulatory
framework, since they often do not have the means to respond timely.
The effects of this are creeping, and therefore easily overlooked.
They translate into a steady reduction of field agronomic work
with GMOs. They can, if continued, also translate in a steady
loss of the best brains. Switzerland today offers a severe example
of how this can happen. The uncertainty generated by the Swiss
referendum on products of biotechnology has made it very difficult
for the highly reputed Swiss public research institutes to recruit
world class young scientists lately, a situation unheard of until
a year ago.