Phosphine use as a pesticide in fumigation of grain and as a rodenticide
COM statement COM/01/S1 - March 2001
Update of COM conclusions reached in 1997
Introduction
1. Aluminium and magnesium phosphides are used in pesticides to fumigate stored products (principally grain). In addition aluminium and zinc phosphide are used in rodenticide products. The active compound in both cases is phosphine gas which is liberated when the phosphides come into contact with moisture. These compounds have been used for this purpose in the UK for over 25 years. In view of their high acute toxicity, their use is limited to trained operators.
2. The COM provided advice on the mutagenicity of phosphine to the Advisory Committee on Pesticides (ACP) in 1997. The text of the COM advice to the ACP is reproduced below as background information. The COM concluded that there was some limited evidence (from human monitoring studies) to suggest that phosphine was an in-vivo mutagen, and suggested that further investigation in pesticide workers would aid in reaching a definite conclusion.
Background: COM conclusions reached in 1997
3. The Committee reached the following conclusions with respect to the mutagenicity of phosphine and metal phosphides.
i) Aluminium and magnesium phosphide are used in pesticides to fumigate stored food products (principally grain). In addition, aluminium and zinc phosphide are used in rodenticide products. The active compound in both cases is phosphine gas which is liberated when the phosphide comes into contact with moisture. These compounds have been used for this purpose in the UK for over 25 years. The Advisory Committee on Pesticides (ACP) requested the advice of the COM on the mutagenicity of these pesticides. The Committee reviewed all the available published and unpublished data forwarded by the Pesticides Safety Directorate and placed a particular emphasis on the studies in humans. The Committee agreed the following conclusions which are applicable to phosphine and all the above mentioned metallic phosphides.
In-vitro studies
(ii) Phosphine (or zinc phosphide) has consistently given negative results in assays for gene mutation in bacteria (Salmonella typhimurium). There was some concern that the high toxicity of phosphine could mask potential mutagenic effects. Phosphine has however been shown to have clastogenic potential in cytogenetic studies in mammalian cells and zinc phosphide has given positive results in the mouse lymphoma assay. In both cases activity was seen in the presence and absence of an exogenous metabolic activation system. Phosphine and metallic phosphides therefore have mutagenic potential in-vitro.
In-vivo studies in animals
(iii) Phosphine has been extensively studied in-vivo using bone marrow/peripheral blood assays for clastogenicity. Although some inconsistent data were obtained, overall it can be concluded that negative results were obtained. Negative results were also obtained in a liver UDS assay and in a dominant lethal assay in mice, although this latter assay was of limited quality. Overall, the Committee concluded that there was no convincing evidence that phosphine and metallic phosphides had shown mutagenic activity in-vivo in rodents.
Studies in humans
(iv) The Committee considered three published reports in detail.
(v) Garry VF et al (Science, 246, 251-255, 1989) studied small groups of grain workers in Minnesota. They reported an increase in total chromosome aberrations (excluding gaps) and of deletions, breaks and complex aberrations in a group of 9 individuals with reported exposure to phosphine alone compared to a control group of 24 individuals with no exposure to pesticides and a control group of 15 grain workers employed in the inspection and processing of grain. Limited personal sampling data were available to show exposures up to 4.1 ppm phosphine in enclosed space applications and up to 0.64 ppm for open-air applications. The Committee noted the small number of phosphine workers examined and the limited details available regarding matching of control and exposed groups for smoking habits. The Committee observed the considerable overlap in the range of results for pesticide applicators exposed to phosphine and grain worker controls. Overall it was concluded that pesticide applicators exposed to phosphine alone had elevated levels of chromosome damage in this study.
(vi) In a subsequent study the same group of authors (Garry VF et al Cancer Epidemiology, Biomarkers and Prevention, 1, 287-291, 1992) reported similar findings in a group of 6 pesticide applicators exposed to phosphine compared to a control group of 26 individuals. The Committee agreed that the results, ie increased levels of chromosome damage were compatible with the earlier study by these authors.
(vii) In a separate Australian study, Barbosa A and Bonin AM (Occupational and Environmental Medicine, 51, 700-705, 1994) analysed micronuclei in peripheral blood lymphocytes taken from 31 fumigators working with phosphine and 21 control subjects. Limited exposure data were provided for 3 individuals employed as fumigators and reported exposure concentrations of 0.1-0.8 ppm for a period of 1 hour. The authors report that the groups were matched for sex age and smoking habit but few details were provided. No evidence of an increase in micronuclei in phosphine exposed fumigators was documented. No increase in urinary mutagenicity (Salmonella assay) was seen in fumigators.
Conclusions human studies
(viii) The Committee concluded that it would be prudent to assume that phosphine was a human genotoxin on the basis of the results reported in studies of Garry et al. The Committee agreed that the negative data reported by Barbosa and Bonin may be due to the fact that they studied workers exposed to lower levels of phosphine and used a methodology of lower sensitivity than Garry et al. However, there were limitations in the studies published by Gary et al, for example in subject and control selection and matching, and phosphine exposure estimations. The Committee felt that further data were needed to ascertain whether these results were reproducible (by other groups) before any definitive conclusions could be drawn. Consideration should be given to the feasibility of a study in UK pesticide applicators using modern methods for assessing genotoxic effects. The COM would be willing to advise on the design of such a study.
Referral from ACP secretariat February 2001
3. The ACP secretariat asked for further advice from the COM in view of additional data which had been submitted by industry (namely a negative inhalation carcinogenicity bioassay in the rat) and proposals outlining a strategy for assessing genotoxic effects in workers occupationally exposed to phosphine.
COM consideration of new data
4. Members reaffirmed that phosphine was highly reactive and agreed that the mutagenicity data provided some evidence for positive results in-vitro but the available in-vivo assays were negative. The Committee agreed that the negative rat inhalation carcinogenicity bioassay provided additional reassurance with regard to site of contact mutagenicity.(1) It was noted that there was no evidence that phosphine had a direct interaction with DNA, and that any mutagenic effect in-vitro might be related to the production of reactive oxides which could react with cellular components. It has come to the notice of the Committee that phosphine has recently been reported to induce oxidative DNA damage in-vitro and in-vivo. (2,3) The Committee was aware that occupational exposure to phosphine was significantly reduced in fumigators and other workers exposed to phosphine by the use of Personal Protective Equipment.(4) It was therefore unlikely that any occupational groups could be currently identified within the UK where an appropriate biomonitoring study could be undertaken. Members were aware that the use of phosphine as a pesticide in the UK was expected to increase in the future as a result of restrictions to be enforced with respect to the use of other fumigants (methyl bromide). The Committee was aware of proposals by the ACP to require monitoring of exposure in workers undertaking fumigation operations and those re-entering grain stores. The Committee noted that they would like to see a review of any results when available.
COM conclusion
5. The Committee concluded that additional new data, (namely negative rat inhalation carcinogenicity study), together with the negative in-vivo mutagenicity data in bone marrow and liver in animals provided sufficient reassurance regarding the in-vivo mutagenicity of phosphine, taken together with the information relating to very low potential exposure arising from pesticides use. Thus the COM's original suggestion for consideration of a monitoring study in UK pesticide applicators was no longer necessary. The Committee noted that use of phosphine was likely to increase in the future and asked to be informed of any results of exposure measurements made available to the UK regulatory authorities.
References (post 1997)
1. Newton PE, Hilaski RJ, Banas DA, Wilson NH, Buesy WM and Shaheen DG (1999). A 2-year inhalation study of phosphine in rats. Inhalation Toxicology, 11, 693-708.
2. Hsu CH, Quistad GB and Casida JE (1998). Phosphine-induced oxidative stress in Hepa 1c1c7 cells. Toxicological Sciences, 46, 204-210.
3. Hsu CH, Han B, Liu M, Yeh C and Casida JE (2000). Phosphine-induced oxidative damage in rats: attenuation by melatonin. Free Radical Biology and Medicine, 15, 636-642.
4. Anon. Commentary on PSD/ACP requirements for worker exposure evaluation of phosphine. 30 March 2000. (In confidence data submitted to PSD).
rophages following ozone inhalation. Mut Res, 241, 67-73 (1990).