Expert Review Under Attack: Glyphosate, Talc, and Cancer

Conclusions as to the causation of adverse effects by environmental and occupational agents may have powerful societal consequences, leading to measures to control exposure through removal of products from the marketplace, changes in design and manufacturing processes, regulation, and litigation. We are at a moment when massive and widely publicized litigation is underway in the United States related to carcinogenicity and causation of cancer by two agents, talcum powder and the herbicide glyphosate, sold as Roundup. Verdicts amounting to hundreds of millions of dollars (as much as $4 billion in the case of talc and ovarian cancer) have been reached: for ovarian cancer in talcum powder users and for non-Hodgkin’s lymphoma in persons exposed to glyphosate. Conclusions of the World Health Organization’s International Agency for Research on Cancer (IARC) have figured in this litigation. Through its IARC Monographs, the agency offers evidence-based classifications of the strength of evidence for carcinogenicity of selected agents and exposures.

In this issue of AJPH, Rosner et al. (p. 969) comment on how exposure to talc contaminated with asbestos might have occurred, thus leading to increased risk for ovarian cancer. The manufacturers’ strategy was to label talcum powder as having “nondetected” amounts of asbestos, leaving the possibility that cancer-causing asbestos fibers were present. Today’s litigation may be a legacy of this strategy.

Here, I focus on the IARC classification of the herbicide glyphosate as “probably carcinogenic to humans” (Group 2A) and the aftermath of the classification.¹ IARC Monographs have also covered both talc and asbestos and ovarian cancer. Monograph 93 classified talc applied to the perineum as “possibly carcinogenic to humans” (Group 2B) while Monograph 100C concluded that asbestos causes cancer of the ovary.

There are critical lessons learned from the IARC glyphosate classification and its aftermath around transparency, journals, and unrevealed conflict of interest (COI), and on potential consequences for those serving on expert panels. First, I state my own potential biases and COIs; for more than three decades, I have participated in and chaired multiple IARC Working Groups, including chairing the group that recently revised the Preamble to the IARC Monographs. I support the program and have defended it and called out its accomplishments.² My views are shared by many.³

UNPAID VOLUNTEER EXPERTS

Globally, the IARC in Lyon, France, has maintained a program since 1971 that classifies the strength of evidence for carcinogenicity of agents (e.g., asbestos) and exposures (e.g., shift work). The agents to be reviewed are identified through nominations and the deliberations of an international advisory group convened every five years. The IARC Monographs, developed by volunteer working groups and IARC staff, comprehensively cover exposure, mechanistic evidence, animal bioassays, and human studies and reach consensus on the strength of evidence for carcinogenicity following principles in the revised Preamble that guide the process. The work of assembling the monographs is demanding of volunteers and staff; it begins with literature searches, premeeting drafts that cover the components of the volume, a meeting at IARC lasting eight days, and postmeeting work to complete the Monograph. As IARC is a World Health Organization agency, IARC’s classifications have global implications as they may affect regulation, policy measures, public health recommendations, and litigation. The connection of IARC classifications to litigation is particularly tight in the United States. Not surprisingly, those entities involved with producing or using the specific agents addressed in the Monographs have a high level of interest in an IARC classification.

Remarkably, IARC is able to recruit unpaid volunteers to participate in the working groups. They commit to at least two weeks of professional time, given the tasks of literature review and writing of drafts before the meeting, along with time spent at the working group meeting in Lyon. Generally, as experts involved with agents reviewed, they understand the array of stakeholders, their interests, and the consequences of potential classifications.

Measures by IARC and other groups convening expert panels to guard against potential COIs have long been in place. The revised IARC Preamble describes the steps taken to ensure that working group members and the IARC Secretariat reveal potential COIs:

Before a meeting invitation is extended, each potential participant, including the IARC Secretariat, completes the WHO Declaration of Interests form to report financial interests, employment and consulting (including remuneration for serving as an expert witness), individual and institutional research support, and non-financial interests such as public statements and positions related to the subject of the meeting. IARC assesses the declared interests to determine whether there is a conflict that warrants any limitation on participation.^4(p6)

Two months before a meeting, IARC releases the names of meeting participants with a summary of declared interests. There is a further updating at the time of the meeting. These systematic procedures cannot guard against all possibilities, including future activities of working group members. In the instance of the glyphosate review, an invited specialist became involved in litigation against Monsanto.

AN UNPRECEDENTED AND HARSH STRATEGY

Aggressive actions have been taken by Monsanto (now owned by Bayer), the manufacturer of glyphosate, around the 2015 IARC classification of glyphosate as “probably carcinogenic to humans.” The Monsanto strategy parallels those used by the tobacco industry and others, but the glyphosate story is notable for its intensity, its reach to the working group members, and the immediacy and scope of litigation in the United States related to non-Hodgkin’s lymphoma. (To illustrate, search online for “looking for a lawyer for Roundup lawsuit.”) The attack even reached to the funding from the National Institutes of Health for the Monographs program. A complete telling of the story can be found in Carey Gillam’s book, Whitewash, and it is not over.⁵ The glyphosate litigation illustrates the inevitability of the translation of IARC classifications into actions based on the strength of evidence. The glyphosate classification as “probably carcinogenic to humans” (Group 2A) apparently met the threshold for litigation of US lawyers.

IARC experts understand that findings of working groups are subject to review and criticism, but the consequences for individual members of the glyphosate working group are unprecedented and could affect participation in future working groups. The targeting of particular panel members by Monsanto is a tactic that could spread to other volunteer groups of scientists who provide necessary expertise on controversial issues. One unfortunate consequence could be discouragement of participation in the many venues where researchers provide advice, often on a volunteer and unpaid basis.

The IARC Monographs, like many reviews for weight-of-evidence purposes, are based on the peer-reviewed literature, excluding unpublished findings even if relevant. The pipeline to publication in the peer-reviewed literature is subject to manipulation in many ways and its integrity is based on truthfulness by authors, probing for COIs by journals, and rigorous peer-review processes.

Nonetheless, a troubling set of publications made its way through these filters with unrevealed potential COIs of authors related to funding from Monsanto. In an orchestrated strategy, the full gamut of approaches to contaminating the scientific literature can be found: ghost-written papers, unrevealed connections to Monsanto, and review of messaging by Monsanto. This strategy is not new, and can be traced back to how the tobacco industry manufactured doubt after smoking was first linked to lung cancer in the 1950s.⁶ While long recognized as problematic, unrevealed COIs persist and are ubiquitous, extending well beyond the risks of environmental agents; very recent revelations around unreported potential COIs related to publications on cancer chemotherapies are exemplary (nyti.ms/2XPBZy2).

EXPERT REVIEW UNDER ATTACK

In the United States, entities involved in evaluating potential carcinogens are also threatened: Within the Environmental Protection Agency (EPA), the Integrated Risk Information System (IRIS) Program assesses hazards and quantifies risks, and the Office of Chemical Safety and Pollution Prevention assesses hazards under the Toxic Substances Control Act. The National Toxicology Program of the National Institute for Environmental Health Sciences carries out toxicological assays and synthesizes evidence to identify hazards, both cancer and noncancer outcomes. Each of these programs has protocols for evidence identification, evaluation, and integration.

The IRIS Program has been changing its methodology following guidance from the National Academies of Science, Engineering, and Medicine (NASEM), and its progress in this transition to more transparent approaches was judged to be satisfactory by the NASEM in 2018. Nonetheless, its resources have been reallocated internally and the release of reports in progress has been slowed.⁷ The EPA is also facing personnel and procedural challenges in implementing assessments under the Toxic Substances Control Act.⁷

There are abundant lessons learned from the IARC glyphosate classification, including warnings from the tactics used by Monsanto and their implications. At a moment when evidence-based actions related to the environment are under threat, the widening of the attack to the scientific review processes and the reviewers needs to be called out.