Probable carcinogenicity of glyphosateBMJ 2019; 365 doi: https://doi.org/10.1136/bmj.l1613 (Published 08 April 2019) Cite this as: BMJ 2019;365:l1613
All rapid responses
Rapid responses are electronic comments to the editor. They enable our users to debate issues raised in articles published on bmj.com. A rapid response is first posted online. If you need the URL (web address) of an individual response, simply click on the response headline and copy the URL from the browser window. A proportion of responses will, after editing, be published online and in the print journal as letters, which are indexed in PubMed. Rapid responses are not indexed in PubMed and they are not journal articles. The BMJ reserves the right to remove responses which are being wilfully misrepresented as published articles or when it is brought to our attention that a response spreads misinformation.
From March 2022, the word limit for rapid responses will be 600 words not including references and author details. We will no longer post responses that exceed this limit.
The word limit for letters selected from posted responses remains 300 words.
In his editorial “The probable carcinogenicity of glyphosate,” Kogevinas  engages in the standard apologetics used by IARC’s defenders to reject any substantive criticism of the Agency’s process and specific judgments.
The essential elements of these apologetics are: 1) deflect attention from the most central and glaring facts that should be kept in view; 2) fail to make any distinction between existing studies (or analyses of said studies) on the basis of quality; 3) attempt to discredit any criticism by implying that it is motivated by corporate interests; and 4) repeat the rhetorical assertions of IARC’s transparency and authority.
Kogevinas writes that, “…substantial published evidence from human, animal,
and mechanistic studies at the time of the IARC evaluation indicated that adverse effects from exposure to glyphosate could be classified as probable.” This formulation glosses over at least two crucial facts. First, the IARC 2A classification was dependent on the conclusion that there was sufficient evidence that glyphosate was carcinogenic in rodents. We now know that the rodent studies relied upon by IARC showed no consistent or robust evidence of increased tumor yields in exposed animals. Robert Tarone has shown that the Working Group selected a few positive results in one sex and used an inappropriate statistical test to declare some tumor increases to be significant. Comparable inverse associations, some statistically significant, were ignored. Neither of Tarone’s two peer-reviewed papers [2, 3] is cited by Kogevinas.
Second, the sentence quoted above includes the words “evidence from human,…” [studies] “at the time.” But IARC was aware of updated NHL results from the National Cancer Institute’s Agricultural Health Study (AHS), which were ignored in IARC deliberations because the results for herbicides were inexplicably excluded from a 2014 paper published in PLoS One . The characteristics and methods of the AHS are widely known, and the details of the statistical methods applied in the analysis of NHL were described in the published paper. Thus, the decision to exclude updated results for glyphosate and NHL, while strictly consistent with the IARC policy of not relying on unpublished results, is questionable. A later paper examining associations between glyphosate and cancer incidence in the AHS reported no significant increases for more than 20 solid or lymphopoietic cancers (including many lymphoma and leukemia subtypes) . The phrase “at the time of the IARC evaluation” gives away the whole game. By torturing the data from rodent studies and ignoring the updated data from the AHS, IARC was able to push through a misleading evaluation in line with its predetermined objective of drawing attention to the “probable carcinogenicity of glyphosate.”
Few scientists would dispute that further high-quality studies of glyphosate and glyphosate-containing products might add to our knowledge. But IARC should be held accountable for its cynical misrepresentation of the existing evidence. The Agency needs to own up to the egregious errors in its glyphosate evaluation, rather than arguing that scientists who draw attention to errors and to ignored evidence, and that other agencies which have found glyphosate to be safe at levels to which most people are exposed, can be dismissed as being subservient to corporate interests.
Kogevinas concludes by warning ominously against “attacking authoritative science.” Since IARC stands alone in concluding that glyphosate is likely to cause cancer in humans, “authoritative science” presumably refers to the conclusions of the IARC Monograph 112 Working Group regarding glyphosate exposure. Invoking authority to allow IARC to ignore legitimate criticism of the seriously flawed classification of glyphosate as a probable human carcinogen does a disservice to science.
 Kogevinas M. The probable carcinogenicity of glyphosate. BMJ 2019;365:l1613 doi: 10.1136/bmj.l1613 (Published 8 April 2019)
 Tarone RE. On the International Agency for Research on Cancer classification of glyphosate as a probable human carcinogen. Eur J Cancer Prev. 2018 Jan;27(1):82-87. doi: 10.1097/CEJ.0000000000000289.
 Tarone RE. Conflicts of interest, bias, and the IARC Monographs Program.
Regul Toxicol Pharmacol. 2018 Oct;98:A1-A4. doi: 10.1016/j.yrtph.2018.09.005. Epub 2018 Sep 5.
 Alavanja MC, Hofmann JN, Lynch CF, Hines CJ, Barry KH, Barker J, Buckman DW, Thomas K, Sandler DP, Hoppin JA, Koutros S, Andreotti G, Lubin JH, Blair A, Beane Freeman LE. Non-hodgkin lymphoma risk and insecticide, fungicide and fumigant use in the agricultural health study. PLoS One. 2014 Oct 22;9(10):e109332. doi: 10.1371/journal.pone.0109332. eCollection 2014.
 Andreotti G, Koutros S, Hofmann JN, Sandler DP, Lubin JH, Lynch CF, Lerro CC, De Roos AJ, Parks CG, Alavanja MC, Silverman DT, Beane Freeman LE. Andreotti Glyphosate use and cancer incidence in the Agricultural Health Study. J Natl Cancer Inst. 2018 May 1;110(5):509-516. doi: 10.1093/jnci/djx233
Competing interests: No competing interests