Br J Ind Med: first published as 10.1136/oem.50.4.340 on 1 April 1993. Downloaded from 340 British Journal of Industrial Medicine 1993;50:340-348 A review of potential human carcinogenicity of the chlorophenoxy herbicides MCPA, MCPP, and 2,4-DP GG Bond, R Rossbacher Abstract Sweden that had linked use of 2,4,5-T to increased For the purpose ofassessing the human carcin- risks of soft tissue sarcoma and malignant lym- ogenic potential of the chlorophenoxy herbi- phoma.'4 cides MCPA, MCPP, and 2,4-DP, the relevant More recently, the focus on chlorophenoxy com- epidemiological and toxicological evidence is pounds has shifted to 2,4-dichlorophenoxyacetic reviewed. These compounds have not pro- acid (2,4-D) owing to findings of an association duced tumours in animal studies conducted between its frequent use and non-Hodgkin's lym- under current test guidelines, giving no reason phoma.5 A panel of experts was recently convened to predict that they would be carcinogenic to to review the extensive toxicological and humans. Epidemiological studies have been epidemiological data on 2,4-D, and concluded that conducted on three continents; greater whereas there was suggestive evidence of a link with emphasis is placed on the studies reported non-Hodgkin's lymphoma, it was far from con- from western Europe, however, as this has clusive (Harvard School of Public Health, unpub- copyright. been the area of more use. Although several of lished 1990). these studies provide suggestive evidence of By comparison, little attention has been given to associations between exposure to chloro- the other members of the chlorophenoxy family phenoxy compounds and increased risks for including 4-chloro-2-methyl phenoxyacetic acid some uncommon cancers, it is inconsistent and (MCPA), 2-(4-chloro-2 methylphenoxy) propionic far from conclusive. None of the evidence acid (MCPP), and' 2-(2,4-dichlorophenoxy) pro- http://oem.bmj.com/ specifically implicates MCPA, MCPP, or 2,4- pionic acid (2,4-DP). Yet MCPA, MCPP, and 2,4- DP as human carcinogens. DP continue to be commercially important her- bicides with widespread use, particularly in western (British Journal of Industrial Medicine 1993;50:340-348) Europe. This paper reviews the available epidemiological evidence relating to MCPA, MCPP During the past decade, the potential for chlorophen- and 2,4-DP, in context with the relevant animal oxy herbicides to cause certain forms of cancer in toxicological data, to arrive at a weight of the humans has come under increasing scrutiny. The evidence evaluation of the potential for these com- on September 30, 2021 by guest. Protected initial focus was principally directed at one member pounds to cause cancer in humans. of this family of herbicides, 2,4,5-trichloro- phenoxyacetic acid (2,4,5-T), presumably because of Patterns of use for MCPA, MCPP, and 2,4-DP its low level contamination by 2,3,7,8-tetrachloro- The chlorophenoxy compounds act as selective her- dibenzo-p-dioxin (2,3,7,8-TCDD), which had been bicides and have been used worldwide for the control shown to be an animal carcinogen.' In 1979, the US of unwanted vegetation. Their individual patterns of Environmental Protection Agency cancelled most use differ somewhat between North America and registrations for 2,4,5-T citing as partial justification, western Europe. For instance, whereas 2,4-D is the a series of case-control epidemiological studies from major herbicide used to control weeds in small grains in the United States, it has a much smaller market share in western Europe where MCPP is currently the principal product of choice for this purpose. In North The Dow Chemical Company, 1803 Building, America, 2,4-DP is used primarily on Midland, MI 48674, USA non-cropland (for forestry purposes and to clear G G Bond rights of way) whereas it is used extensively on small Department of Toxicology 470, BASF Aktiengesells- grain crops in Europe; MCPA is used in similar ways chaft, D-6700 Ludwigshafen, Germany on both continents, but has traditionally had a much R Rossbacher larger market share in Europe than in the United Br J Ind Med: first published as 10.1136/oem.50.4.340 on 1 April 1993. Downloaded from A review ofpotential human carcinogenicity of the chlorophenoxy herbicides MCPA, MCPP, and 2,4-DP 341 States. These patterns of use suggest that, among cinogen, as mechanisms other than a direct attack on studies that have not specified a principal phenoxy genetic material may be responsible for tumour product, those conducted in Europe should be given development. For these reasons, it is essential to greater attention when assessing the potential for interpret mutagenicity as only a single component of MCPP, MCPA, and 2,4-DP to cause human cancer. the weight of the evidence evaluation of potential Indeed, for this reason many ofthe studies conducted carcinogenic effects in humans. outside Europe are not relevant to this review and so have been given either little or no attention. It is also critical to recognise that, compared to the CHRONIC TOXICITY AND CARCINOGENICITY STUDIES acetic acids, the racemic propionic acid derivatives MCPA were introduced into commerce recently and 2,4-DP The herbicide MCPA was administerd to groups of and MCPP were not actively marketed until the late male and female Wistar rats at doses of 0, 20, 80, and 1950s. Consequently, there has been less than 30 320 ppm in the diet for 24 months.'0 At doses of years of latency possible with these products. 80 ppm and above evidence of toxicity was found. More recently, MCPP-P and 2,4-DP-P were in- The target organs identified were the kidneys, liver, troduced in the market. Both compounds consist of and red blood cells. No substance related changes the herbicidal active D-form whereas MCPP and were found in the 20 ppm (- 1-33 mg/kg/day) dose 2,4-DP are a mixture of the L- and D-forms with a group. There was no increase in the incidence oftotal ratio of 1:1. or specific neoplasms at any dose. By contrast, MCPA and the other chlorophenoxy Similarly, MCPA given in the diet to B6C3F1 compounds have been in use more than 40 years. It is mice at doses of 20, 100, and 500 ppm for two years apparent then, that the available epidemiological produced some toxicity to the kidneys of the animals studies can only evaluate the hypothesis that 2,4-DP in the highest dose group, but no carcinogenic and MCPP are human carcinogens which act with response." short to moderately long latency. copyright. 2,4-DP Summary of toxicology Technical grade 2,4-DP was administered in the diet Evidence from toxicological studies conducted on of Fischer 344 rats at doses of 0, 100, 300, 1000, and laboratory animals or cell cultures is often useful for 3000 ppm for two years.'2 The major target organs interpreting the biological plausibility of findings identified were the kidneys and liver. No substance obtained from observational epidemiology. Par- related toxic effects were found in the 100 ppm ticularly helpful for evaluating a carcinogenic poten- (-6-7 mg/kg/day) dose group. There was no http://oem.bmj.com/ tial are studies in which animals receive the test increase in total or specific neoplasms at any dose, compound over their life span (chronic toxicity and thus showing the lack of carcinogenic potential of carcinogenicity studies). The species routinely used 2,4-DP in rats. are mice and rats. For pesticides oral administration A carcinogenicity study in mice with 2,4-DP-P is preferred to simulate the possible uptake of (optically active D-form) will be initiated after the residues in food. For the applicator it is generally completion of a respective range finding study. A accepted that dermal exposure is most relevant, then subchronic range finding study in B6C3F1 mice inhalation exposure, whereas oral exposure is indicates that 2,4-DP is less systemically toxic than on September 30, 2021 by guest. Protected negligible. MCPA." Nevertheless, oral carcinogenicity studies can also be used for the applicator, because like 2,4-D, other compounds such as MCPA, MCPP, and 2,4-DP are MCPP also hardly metabolised79; thus after dermal adsorp- The herbicide MCPP was given in the diet of male tion no difference in systemic toxicity would be and female Wistar rats for two years at doses of0, 20, expected. 100, and 400 ppm.'4 As for 2,4-DP and MCPA the Genotoxicity assays are conducted to test a kidney was a target organ. No substance related chemical's capacity to damage cellular genetic effects were noted at 20 ppm (- 1-3 mg/kg/days). material as shown by the formation of genetic There was no increase seen in total or specific mutations. A positive result in a mutagenicity test neoplasms at any dose, thus showing the lack of suggests a greater likelihood that a chemical will be carcinogenic potential of MCPP in rats. carcinogenic in intact animals, but does not provide A carcinogenic study in mice with MCPP-P definitive evidence. Indeed, there is considerable (optically active D-form) will be initiated after the published debate about the importance of positive completion of a respective range-finding study. A results in mutagenicity studies and their potential to subchronic range finding study in B6C3F1 mice cause cancer in humans. Negative results in indicates that racemic MCPP is less systemically mutagenicity tests cannot exonerate a potential car- toxic than MCPA.'5 Br J Ind Med: first published as 10.1136/oem.50.4.340 on 1 April 1993.