Opinion of the Scientific Panel on Food Additives, Flavourings, Processing Aids and Materials in Contact with Food on a request from the Commission related to an application on the use of cassia gum as a food additive QUESTION No EFSA-Q-2003-134 Adopted on 26 September 2006

The European Commission has asked the European Food Safety Authority to provide a scientific opinion on the safety of cassia gum when used as a gelling agent and as a thickener. Cassia gum is the flour from the purified endosperm of seeds from Cassia tora and Cassia obtusifolia. Cassia gum is comprised primarily of a linear chain of 1,4-beta-D mannopyranose units with 1,6-linked alpha-D galactopyranose units attached to every fifth mannose. Galactomannans are recognised as components of dietary fiber and are resistant to digestive enzymes in the gastrointestinal tract. The absorption and distribution of cassia gum have not been studied. It is expected that cassia gum would be excreted unchanged. Fermentation of cassia gum by gut microflora may occur to a small extent. However, the panel notes that any hydrolysed material would represent oligo- or monosaccharides that can be expected to be absorbed and metabolised in normal biochemical pathways. The petitioner requests the authorisation of cassia gum for use in ice cream and frozen milk desserts for stabilising and controlling overrun, for the improvement of water retention in certain baked goods (including cheese-creme filling in pastries and other cheese-creme filled desserts), for use as a thickening agent for soup mixes, sauces and selected oil-free salad dressings, and for the improvement of texture and water retention in yogurt and yogurt drinks, sausages, corned beef, and canned poultry meats. Cassia gum is intended to be used in sausages, corned beef, and canned poultry meats at levels up to 1.5 g/kg and in all other applications at levels up to 2.5 g/kg. Based on the petitioner's proposed use levels of cassia gum as a food additive and on conservative assumptions daily exposure to cassia gum was estimated to be 2.1 mg/kg bw/day at the mean, and 4.9 mg/kg bw/day at the 90th percentile (US data for consumers only). Most acute, sub-chronic and reproductive and developmental toxicity studies reported for cassia gum have been performed with a cassia gum preparation with specifications different from those defined in the present application. These studies used an older test sample of cassia gum containing approximately 70 mg/kg of total anthraquinones. The petitioner has recently implemented an isopropanol extraction purification step that reduces the total anthraquinones levels from 70 mg/kg to below the 0.5 mg/kg detection limit, the latter being in line with the specifications of the present submission. Cassia gum containing approximately 70 mg/kg anthraquinones has been tested in acute, sub-chronic, reproductive and developmental toxicity studies and it does not have significant toxicological properties. The Panel notes that the margin between the no-observed-adverse-effect level (NOEL) in the 28-day rat study of 230 mg/kg bw/day and the exposure estimates of 2.1 mg/kg bw/day at the mean, and 4.9 mg/kg bw/day at the 90th percentile, based on the proposed levels of use, is between 50 and 110. However, the Panel also notes that the slight biochemical and haematological effects occurring at the higher dose levels in this 28-day rat study were generally not dose related and of limited toxicological significance. Furthermore, in a 90-day dog study and a 13-week cat study there were no effects up to doses of 3500 and 1250 mg/kg bw/day respectively. Cassia gum of the old specifications was not mutagenic or clastogenic in mammalian cells. Based on the results of recent genotoxicity studies, cassia gum, prepared by the newly defined production method, did not increase the number revertants in any of the four Ames tester strains (S. typhimurium), nor in the E. coli WP2uvrA test strain both in the presence and absence of S9-metabolic activation. It is concluded that cassia gum complying with the newly defined specifications does not give rise to safety concern with respect to genotoxicity. Long-term carcinogenicity studies on cassia gum were not available. Other related galactomannan gums, including locust (carob) bean, guar gum and tara gum were not carcinogenic when fed to mice and rats. Given that cassia gum is not genotoxic, and that many other related galactomannan gums are not carcinogenic, the Panel does not consider long-term carcinogenicity studies essential for the safety assessment of cassia gum. The toxicological data on cassia gum are insufficient to establish an acceptable daily intake (ADI). On the other hand, the existing data do not give reason for concern. The Panel wishes to stress the importance of inspection of the seeds for cassia gum preparation for the presence of seeds of C. occidentalis which has to be less than 0.1% by selection based on color and shape. Given these results from the toxicological studies, the very low absorption of cassia gum and the fact that, if hydrolysed at all, cassia gum would be degraded to compounds that will enter normal metabolic pathways, the Panel concludes that the use of cassia gum complying with the newly defined specifications as an additive for the proposed food uses is not of safety concern.