Agriculture Reference
In-Depth Information
evaluation of GM-based food or feed. One of
the major problems concerning the adapt-
ation of the test is the fact that feeds (diets)
are comparable between animal groups
(control or treated) in 90-day rodent studies,
whereas in food safety assessment studies,
the feeds themselves are tested. In most
GM-based food or feed studies, 33% of GM
feed is included in animal diets (see
recommendations of the French Agency for
Food, Environmental and Occupational
Health and Safety, ANSES, 2011); the
remaining part provides a balanced diet.
Nevertheless, it should be kept in mind that
90-day animal feeding studies are designed
to expose any change (e.g. a compositional
change) in the GM feed, whether linked
directly to the genetic modii cation (i.e. the
transgene) or not (e.g. due to other genetic
dif erences between plant varieties). Such a
broad method leads to a drawback in which
these studies may not actually be able to
detect weak ef ects, as stated by the expert
panel, 'It is unlikely that substances present
in small amounts and/or with a low toxic
potential will result in any observable
unintended ef ects'
(EFSA, 2008). For the
testing of a specii c molecule, it is possible to
increase the dose in order to observe the
biological ef ect; however, this is impossible
for feed tests, as it would compromise the
balance of the diet (see Chapter 5). h erefore,
it is benei cial to examine whether tests
beyond a 90-day rodent feeding study are
needed.
Moreover, the EFSA states that:
In these specii c cases, such late ef ects may
not be detected by 90-day rodent feeding
studies. h erefore, it is interesting to
examine whether long-term studies (per-
formed over a period longer than 90 days)
and multi-generational studies can detect
unintended ef ects and whether the i ndings
between such studies and 90-day feeding
studies dif er.
In this chapter, we assess critically
recently published studies on the long-term
ef ects of GM plants, i.e. studies signii cantly
longer than the 90-day subchronic tests
(17 studies), as well as multi-generational
studies (16 studies). h e GM feeds were
derived mainly from marketed insect-
resistant (Bt) and herbicide-tolerant
varieties, while some studies tested experi-
mental GM lines. We examine whether these
publications reveal adverse ef ects. h ese
long-term studies and multi-generational
studies are compared to 90-day studies that
have already been performed. h e possible
need to update the current regulatory frame-
work is discussed (see Chapter 5 for types of
studies).
Long-term studies (longer than the classical
90- to 96-day feeding trials using rodents)
are listed in Table 8.1 and discussed below.
h e duration of GM-based diet feeding
varies between 110 days (16 weeks) and 765
days (110 weeks). Long-term studies that
also involved mating/reproduction/of -
spring will be examined in the following
section on multi-generational studies.
Various animal models have been used, such
as mice, pigs, cows, quail, macaques and i sh;
however, rodents were the predominant
models. Several criteria have been evaluated
(body and organ weights, haematological
analyses, enzymatic activities, histo-
pathological observations of organs and
detection of transgenic DNA). More details
can be found in Snell
et al
. (2012). h e
studies presented in Table 8.1 concern
h e subchronic, 90-day rodent feeding study
is not designed to detect ef ects on
reproduction or development, other than
ef ects on adult reproductive organ weights
and histopathology. h us, in some cases,
testing of the whole food and feed beyond a
90-day rodent feeding study may be
recommended. In cases where structural
alerts, indications from the subchronic study
or other information on the whole GM plant
derived food and feed are available that
suggest the potential for reproductive,
developmental or chronic toxicity, the
performance of such testing should be
considered.
(EFSA, 2008)
