The Audit Committee was established by the Director of the Rowett Research Institute in August 1998. Its remit was to determine whether statements made by or attributed to Dr Pusztai during a World in Action programme broadcast on the 10^th August 1998 were justified on the basis of the experimental evidence available at that time and to report its findings to the Director. The Audit Committee met on the 21^st August and reported shortly thereafter.

The relevant statements taken from a transcript of the World in Action programme prepared by Tellex Monitors Ltd were:

The three statements implicit in this exchange were: the completion of long-term experiments (110 days) made with two different kinds of GM potatoes, the retardation of growth and the reduced response to a challenge to the immune system. These three points were found by the Audit Committee to be untrue or not supported by the evidence.

1. Feeding studies were made only with one type of GM potato, that expressing the gene for the lectin GNA.
2. There was no statistical difference in the growth rate of rats fed cooked transgenic GNA potato over the 110 days when compared to the parental contols. Similarly rats fed cooked potatoes spiked with added GNA and a lactalbumin diet similarly spiked with GNA grew at the same rate as rats fed the same diet without the lectin.
3. No test of immune function in rats fed cooked GNA transgenic potato for 110 days had been made at the time of filming or broadcast.

The Audit Committee were also concerned to examine all of the data available in case a simple transposition of data had led to an inadvertent misunderstanding. Accordingly the Audit Committee considered in addition to the single long-term experiment made with GNA transgenic potato, a long-term experiment made with conventional potatoes to which a second type of lectin (ConA) had been added as a "spike". Data from ten short-term (10 day) trials were also studied. Only two of these ten short-term experiments involved the feeding of rats with transgenic potato (Expts D227 and D242) although a third in progress at the time of the Audit was subsequently completed (D249).

Dr Puszrtai was provided with a copy of the Audit report by the Director and was invited to respond. This he did in October 1998, commenting both on the conclusions of the Audit Committee and on the data from experiment D249, an experiment not started at the time of the recording of the World in Action programme and only completed after the Audit process. A copy of his response was sent to all members of the Audit Committee.

In his report Dr Pusztai takes issue with the Audit committee on three broad counts:

their comments on the analytical data;

their failure to take account of differences in organ weights in animals fed GM potatoes compared with parental controls;

their conclusion with respect to immune function.

Dr Pusztai also rejects the conclusion of the Audit Committee that the existing data (data available in August 1998) did not support any suggestion that the consumption by rats of transgenic potatoes expressing GNA has an effect on growth, organ development or immune function.

Each of the cells comprising an organism are likely to be subtly different and to carry small differences in their genetic composition. This is the source of somaclonal variation referred to by Dr Pusztai. Since any GM plant derives ultimately from a single cell in which the transgene has been successfully incorporated it is to be expected that the GM plant will differ from that of the parent line. In commercial practice this difference is minimised by crossing with non-GM plants until a plant expressing the transgene is obtained which also has all of the other features considered desirable. This is not normal practice with potato because of its complex genetics. Thus for both experimental and commercial purposes it is common practice to select directly the GM potato lines which are "true to type" with respect to the parent plant with the exception of the trait or feature that has been modified by insertion of the target gene. Given the very small number of lines analysed by Dr Pusztai this type of selection was not possible.

The Audit committee was concerned that the inevitable variation between transgenic and parental cell lines was not so great as to make any subsequent interpretation of data difficult. The committee was of the opinion that although the expected differences were present, they were not of sufficient magnitude to seriously undermine the value of the nutritional experiments. The greatest difference found, that of the protein content of line 74/2T, could be compensated for in terms of amount, if not quality and availability, by manipulation of the diet. However, the committee were concerned that no data on the individual and total glycoalkaloid content of the tubers was available since significant differences in glycoalkaloid composition could have unrecognised nutritional consequences (Audit report 2.5).

Contrary to the views of the Audit Committee, Dr Pusztai appears to suggest that his experimental material was not "substantially equivalent" to the parent lines. If this is the case then it is difficult to see how any meaningful conclusions based on a comparison of GM and conventional potatoes can be drawn from his data.

There were only two occasions when differences in growth rates were observed which could possibly be ascribed to the presence of foreign lectin. Feeding of raw and cooked potato diets to which ConA lectin had been added at some 5000x the concentration found in transgenic tubers resulted in a small reduction in growth rate, a response was not replicated in the short-term when ConA was included at lower concentration ((Audit Report 3.3). The second case involved an apparent increase in the growth rate in rats fed uncooked GNA transgenic potato when compared to the appropriate control (Expt D242). This was subsequently attributed to the accumulation of food in the digestive tract; the empty body weights at the end of the experiment were not significantly different. This accumulation of digesta was not observed in a second similar experiment.

The Audit Committee was not able to offer an explanation for this observation (Audit Report 5.3) which was presumably related to poor starch digestion. However raw potato is not an ideal basis for any mammalian diet and some digestive disturbances might well be expected with diets of this kind. The Audit Committee remains of the opinion that there is insufficient evidence to ascribe this effect to any single cause and to suggest that the presence of the transgene is a significant factor would be premature.

Other than these two cases, no statistically significant differences in growth rates attributable to the presence of GNA or ConA lectins could be found.

Dr Pusztai suggests that the Audit Committee did not give due consideration or emphasis to differents in organ weights which were sometimes observed. In his report, Dr Pusztai draws attention to changes in the weight of "many essential body organs", highlighting possible effects on liver function and the digestive system in D227 and other organs including the brain in D242. However, comparisons between data are difficult as values for D249 are provided only wet weights and data are given only for selected organs.

The Audit Committee did consider all of the organ weight data available, making comparisons on relative dry weights ( i.e. independent of body weight) to remove any variation which might be ascribed to differences in body mass. These data, which give a complete account of all body tissues measured, is reproduced hereas Tables 1-5 together with an appropriate statistical analysis. These tables allow ready comparison of the differences between rats fed diets containing foreign lectin and the appropriate controls. Although occasional statistically significant differences were observed, there is no consistency in response. In addition, many effects to which Dr Pusztai draws attention are not significantly different when results are expressed as relative dry weights (see Table 4, Expt. 249). The Audit Committee concluded that, contrary to Dr Pusztai's view, there was no evidence of a consistent effect of foreign lectin, whether supplied as a spike or as a gene product, on organ development.

The Audit Committee recognised that, in some cases, the lymphocyte proliferation data showed a statistically significant reduction in the response of lymphocytes from rats exposed to GNA compared to lymphocytes from rats which had not been exposed. However, statistical significance does not automatically imply biological significance. The Audit Committee remains firmly of the opinion that the extreme level of variation between the responses of replicated cell populations and the inconsistent response to stimulation within and between experiments means that these data cannot be used to make substantial claims about immune responsiveness in these studies.

The Audit Committee has taken cognisance of Dr Pusztai's views but remains of the opinion that there is nothing in his report or in the data seen by the Committee which would lead them to modify the conclusion expressed in the original audit report.

Professor F J Bourne

Dr A Chesson (Chair)

Professor H Davies

Dr H Flint

16 February 1999

• Rat Studies - Transgenic [GNA-expressing] POTATO
• Studies with Transgenic [GNA-expressing] POTATO
• Tables