UK assessment report to market Monsanto's GM maize cross NK603 x MON810 in Europe
Text of submission to a public consultation by Eva Novotny, SGR, on 19 April 2004
 

To: gmoinfo-comments@jrc.it

Re: Comment on Assessment Report C/GB/02/M3/03

These comments concern the assessment of Monsanto’s maize NK603 x MON810, in particular, Section 5, ‘Assessment of use in animal feeds’.

Section 2.1, entitled ‘Confidentiality’, states that the animal-feed safety assessments are confidential ‘in order to protect the Company’s business and competitiveness’. It is difficult to understand why a study providing evidence that the product is wholesome, safe and nutritious should compromise the Company’s business and competitiveness; such secrecy suggests that the results were not as favourable as the Company would have wished.

In this Comment, we compare the stated conclusions of the feeding trials to the conclusions reached in similar experiments performed to assess the glufosinate-resistant maize Chardon LL. Our conclusions, after re-examining the Chardon LL data, did not agree with the those reached in the study. Several other independent scientists also disagreed with the official report. We therefore warn Ministers and their advisors that results claimed for unpublished, confidential studies on animal-feeding experiments should not be accepted without independent corroboration.

Below, we comment on specific sections of the Assessment.

Section 5.1 Safety of gene products

Any crop containing a pesticide in its cells should be suspect. For human consumption, consumers and supermarkets are increasingly demanding that no pesticide residues should remain in food; but the Bt pesticide that is built into the plant by genetic engineering cannot be washed off. It is possible that the time-scales of the animal-feeding trials may not have been long enough for health problems to develop in the test animals. Results for animals might not, in any case, be applicable to human beings.

Section 5.2 Compositional analysis

This section concludes that NK603 x MON810 is ‘compositionally equivalent to its non-transgenic control hybrid line and to other commercial hybrids.’ Yet the preceding paragraph states that ‘for a number of nutrients significant differences were found, including phosphorus levels’, although ‘the range of the values fell within the range reported for the non-GM test hybrid.’ The fact that this statement fails to say that the differences also fall within the range of the additional ‘5 different non-transgenic commercial maize hybrids’ used in the tests suggests that the differences fell outside that range. This prompts the query as to how and why the particular ‘non-GM test hybrid’ was selected to be the control variety. It is, of course, not possible to compare a GM variety with its non-GM parent when, as in this case, both parents are distinct GM varieties. It is therefore necessary to compare the GM maize with a number of non-GM varieties, since no single variety has all the properties of ‘the parent’. Apparently, NK603 x MON810 is not ‘compositionally equivalent’ to the five commercial varieties, despite the concluding statement mentioned above.

Section 5.3 Feeding studies

It should be noted that some of the feeding studies involved a parent maize variety, not NK603 x MON810. Interaction between genes could produce characteristics in the hybrid that are not present in either parent. Hence any tests on the parents are of limited value in assessing the offspring.

NK603

1. Chicken-feeding experiment

A feeding study on NK603 was made on broiler chickens and concluded that ‘No differences were observed in the growth of birds fed NK603 maize compared to the parental control or other reference maize lines.’

We are reminded of a similar statement made in a feeding study¹ on broiler chickens in connection with the GM maize Chardon LL. When we re-examined this study, we found that, although the average weight-gain of the birds fed GM maize was only 1% lower than that of the control group at the end of the study (i.e., after 42 days, as in the present case), the ‘error bars’ representing the departures from the average weight increased much more rapidly with time for the birds given the GM maize, with the ‘errors’ for this group rising to 2.5 times those for the non-GM group in the last phase of the experiment. This implies that some chickens were becoming ever more underweight while at least some others were becoming overweight. No comment was made upon this fact in the study.

Graphs showing the results for the chicken-feeding study on Chardon LL are shown on our web site². Also included there are food-consumption measurements (although food consumption was apparently not measured for NK603 x MON810). In the Chardon LL study, there is a marked drop in food intake by the GM-fed group in the last period; and the ‘error bars’, as for weights, grow much faster for the GM-fed chickens.

The similarities of the NK603 x MON810 study and the Chardon LL study suggest that the data for the present case should be examined carefully. If the ever-growing abnormality of the ‘error bars’ in the Chardon LL study is repeated, this would be an indication that certain individual chickens are reacting abnormally to the GM diet and that it may not be safe.

Other independent scientists have also expressed disagreement with the claimed favourable outcome of the Chardon LL feeding study on chickens.³

2. Rat-feeding experiment

The study on rats for NK603 paralleled that for the rat-feeding study⁴ for Chardon LL. In both experiments, 20 male and 20 female rats were used. However, the Chardon LL study extended over only 13 days, whereas the NK603 experiment lasted for 13 weeks. For NK603, the control varieties were ‘a non-GM parental hybrid or one of a number of reference hybrids.’

The conclusion regarding weights is that ‘There were occasional observations of differences in weight gain between rats fed NK603 and the parental control groups. However, where such differences were recorded, they occurred within one sex, at one intermediate observation point and at one dose level, and the weights still fell within the range observed in rats fed the reference maize diets.’ The rat-feeding study for Chardon LL concluded⁵ that ‘Mean body weights were similar for treated groups and controls. There were no differences which could be attributed to treatment with the test article.’ When we re-examined this result, however, we found the results in Table 1:

Chardon LL: TABLE 1. Weight-gain (gm) per day, between Day 1 and Day 13

MON810

The conclusions of the chicken-feeding and rat-feeding studies reported here should be treated with the same caution as those for NK603.

NK603 x MON810

Again, the conclusions of the chicken-feeding study should be viewed with caution. It is reported that ‘The protein content of the breast meat of the broilers fed the transgenic maize was lower than that for broilers fed the control maize and one of the [six] commercial lines … This result may or may not be significant, depending on how the control maize was selected. In any case, the same cautious approach to the conclusions would be judicious.

Conclusion

Our experience in re-examining the confidential chicken-feeding and rat-feeding studies submitted with the application for Chardon LL should be a warning that confidential studies submitted by GM-seed developers should not be accepted at face value but need to be independently scrutinised by scientists who have no personal interest in the outcome of the results. In the absence of such surveillance, it cannot be concluded that the outcome of the tests is as stated and that the product is safe.

Notes