Response to a feeding study in rats by Zdziarski et al [1]

(August 2018)

Zdziarski et al published a paper in Food and Nutrition sciences on a 26 week feeding study in rats using GM corn. The aim of the study was to investigate the effect of ingesting a triple-stacked GM corn variety (NK603 X MON810 X MON863) on the digestive tract of rats, specifically the stomach mucosa.

Two groups of 10 male Sprague Dawley rats were fed a diet containing either 60% GM corn (sourced from the United States) or 60% non-GM corn (sourced from Australia) for 26 weeks. Rats were monitored daily and weighed weekly including at termination. The stomach was removed immediately post-mortem and weighed then samples removed and prepared for light microscopy, immunohistochemistry and electron microscopy.

The authors reported that the GM corn diet had effects in the rat stomach including alterations to tight junction apposition, gland dilatations with epithelial elongation and dysplasia. It was suggested that these effects may have health implications.

FSANZ response

This study was not a guideline study or conducted according to Good Laboratory Practice (GLP). It contains a number of significant flaws in study design, conduct and reporting that make it unsuitable for regulatory purposes, including:

Animal handling and selection: 

• The source of the rats was not stated and there was no description of screening for general health, pathogenic bacteria or viruses.  
• The authors have not stated whether rats were housed together or individually.
• Only ten male Sprague Dawley rats were used for the GM-fed and non-GM groups. This is a low number of animals for a study of this type. 
• Although it is claimed that all animal husbandry was carried out under blinded conditions, it is
• unclear how this could have been achieved given the test diets had to be re-pelleted during the course of the study. 
• It was not stated whether the animals were fasted prior to necropsy and the time of necropsy was not stated.  
• Only one time point was investigated (terminal necropsy) such that a temporal relationship could not be established.could not be established. 

Test diet  

• The control diet did not include an isogenic or near isogenic corn variety.
• The method of preparation and storage conditions of the test diet to prevent mould and insect colonisation were not reported.
• Compositional analysis of the two diets (GM; non-GM) was not conducted and/or reported. Rather, the nutritional components of a standard rat diet appear to have been included in the report. If the compositions of the original GM and non-GM test diets were comparable to the standard diet, it is unclear why the reported moisture content of the diets differed significantly. This should have been addressed by the study authors.
• Chemical and microbial contamination of the test diet do not appear to have been adequately controlled. In a serious omission because the GM corn came from the USA, while the non-GM corn was grown in Australia, and the mycotoxin contamination could be quite different.
• Water consumption does not appear to have been monitored or reported and may have differed significantly between GM-fed and non-GM-fed animals, particularly given the authors suggestion that food consumption of the two groups was different and less than anticipated due to low moisture in the test diets.
• Only a single high dose level (60% in the diet) was used. Therefore it is not possible to demonstrate a dose-response relationship in order to establish whether there was an indication for a causal effect.

Results

• No mortality or clinical signs were reported. The authors reported that food consumption was not as high as expected and that the GM-fed rats were eating less than non-GM-fed rats. It was suggested that this may be due to low moisture content in the test diets. However, no data were provided on body weight, body weight gain, food consumption, water consumption or even how long it took to reformulate the test diets. Failure to adequately control or report test diet intake, diet composition, body weight and food and water consumption are significant flaws in this study.  
• There was no significant difference in stomach to body weight ratio. Stomach weight per se was not reported. There was no evidence of significant treatment-related cellular damage, degeneration or necrosis in the non-glandular or glandular stomach. Dilated gastric glands are a common age-related change in rats and not considered to be adverse. The authors state that the most striking finding was “the loss in tight junction apposition between mucus-producing cells of the fundus in the GM-fed group”. However this was also observed in 50% of non-GM fed controls and no historical control data were provided on background incidence. There was also no evidence that these changes were related to adverse or functional outcomes.
• The study design did not include an investigation of other endpoints typically required in repeat-dose studies such as haematology, clinical chemistry, urinalysis, or examination of other organs or tissues at necropsy that may be used to establish the biological relevance of the observed findings. Further, it is noted that a range of factors such as nutrients, hormones, soluble and cellular inflammatory mediators, and bacterial adherence may all regulate tight junction function and permeability. For example, a significant omission was the failure to analyse the diet for mycotoxin contamination. A number of mycotoxins have been shown to alter the gastric mucosa in rats in a variety of ways, including tight junctions, and in the absence of a mycotoxin assay of the two diets, the possibility that any subtle differences in gastric histology are mycotoxin-related and not treatment-related cannot be eliminated.  

FSANZ conclusion

There are a number of serious flaws in the design, conduct and reporting of this study that make it unsuitable for regulatory purposes.