A non-estrogenic alternative to Bisphenol A at last?
Sarah Vogel, Ph.D., is Vice-President for Health.
Last week a new study was published showing promising results for a non-estrogenic alternative to polymers based on bisphenol A (BPA) used to line the inside of food cans. The paper, in Environmental Science & Technology (ES&T), evaluated the estrogenicity of an alternative to BPA— tetramethyl bisphenol F (TMBPF) — and its final polymer product developed by Valspar, a major paint and resin company. The authors found that, unlike BPA and some of its analogs that have been used as substitutes, TMBPF exhibited no signs of estrogenicity.
This was an unusual paper on a number of fronts—how the material was selected, how it was evaluated and by whom. In this post I’m going to explore who was involved, what testing was done and what this might mean for the BPA alternatives market.
But first a bit of context because you may be asking, aren’t there already non-BPA-based polymer alternatives for use in food cans? What about those cans that say “BPA-free”? The reality is that despite serious, decades-long concerns about the estrogenicity of BPA, the chemical continues to be widely used to make linings of food cans due to the considerable versatility of the material, which has made it challenging for alternatives to penetrate the market. There are a few alternatives available but none have offered a total replacement of BPA-based coatings used in nearly all food cans today.
The new research described in the ES&T paper was funded by Valspar, which immediately raises concerns about conflicts of interest, given the considerable financial stakes it has in finding an effective and safer replacement to BPA. But here is where it gets interesting in terms of who was involved in this project and how conflicts of interests were managed.
Who did the research?
To assess the estrogenicity of TMBPF, Valspar worked with leading researchers who have been publicly outspoken about the adverse health effects of BPA and other endocrine-disrupting chemicals. Dr. Ana Soto, the lead author, runs a laboratory at Tufts University which has been researching BPA and other endocrine-disrupting compounds for decades. And co-author Dr. Maricel Maffini, who also worked at Tufts in Dr. Soto’s lab, has an extensive publication record on the adverse effects of BPA. One could argue that they have an interest in finding a safer alternative to BPA-based resins.
What we know from the disclosure of competing financial interests included in the paper is that Valspar provided Dr. Soto’s laboratory with “unrestricted funds” to conduct tests of the estrogenicity of TMBPF and the final polymer. Valspar also contracted with Dr. Maffini to serve as technical consultant on the project. In addition, a Valspar employee and a former Valspar employee were also co-authors.
(For full disclosure on our part, Dr. Maffini is also a consultant with EDF. We would not financially benefit from the successful marketing of this material.)
Why TMBPF?
The selection of TMBPF for evaluation arose from a reframing of the design problem. Rather than searching for a novel or alternative polymer material, the problem was turned on its head: can the same polymer be used – but employ a safer bisphenol building block? More specifically, how could the highly effective 4,4’-Methylenediphenol polymer chain used in food can linings be made without using an estrogenic bisphenol, like BPA?
The authors note that they selected TMBPF because predictive modeling suggested it was unlikely to bind to the estrogen receptor. They also note that TMBPF was preferred because it could be employed in a two-step process that created an intermediate and hence avoided TMBPF being directly used in the final polymerization process; this would reduce the likelihood of unbounded monomer remaining in the final product—a problem with the BPA-based polymers.
What testing did they do?
The first step was to evaluate the migration of TMBPF from the polymer coating in order to assess the likelihood of exposure in food. No migration of TMBPF was detected above the level of detection of 0.2ppb well below reported levels of BPA migrants from epoxy can linings.
The authors then selected a series of in vitro and in vivo tests that assessed TMBPF’s and the final polymer’s estrogenicity. The in vitro tests included the following:
- E-SCREEN: This uses a human breast cancer estrogen-sensitive cell line called MCF-7. Both TMBPF and migrants from the final polymer were evaluated, along with positive controls (BPA and estradiol) and negative controls. Neither TMBPF nor the migrated material resulted in an increase in proliferation of the cells—which would indicate estrogen receptor activation.
- Real-time quantitative PCR: This test evaluated gene expression of four genes known to be regulated by estrogen receptors. TMBPF was evaluated and compared with a positive control (estradiol) and negative control. TMBPF did not activate expression of any of the genes.
Moving up in biological complexity, the researchers conducted two in vivo whole-animal studies:
- Uterotrophic assay: This test, which followed the Environmental Protection Agency’s (EPA) Endocrine Disruptor Screening Program (EDSP) protocol, evaluated the effect of TMBPF exposure on the development of the uterus in immature female rodents.
- Pubertal assay: This test also followed the EPA’s EDSP guidelines to assess development of puberty and thyroid effects. The researchers also evaluated effects on the development of the reproductive organs including the mammary gland.
The paper reports no significant effects of TMBPF on any of these endpoints. (The uterotrophic assay included a positive control while the pubertal assay did not.) Valspar also made reports of tests available on-line.
What additional disclosures would be helpful?
It would be useful to have a more complete description of the criteria used to select the different tests for estrogenicity and how these decisions were made and by whom. A discussion of the terms of the contract would make more transparent the level of independence given to the Tufts Laboratory. We strongly encourage additional disclosure of these details.
That said, the rigor of the assessment went well above anything required by current law. And that deserves praise. While the effectiveness of the final polymer derived from the use of TMBPF (i.e. shelf-life, performance, etc.) as well as additional safety assessments were beyond the scope of this paper, we are cautiously optimistic that Valspar has identified a non-estrogenic replacement for BPA-based polymers food cans.
That is encouraging news.
2 Comments
As a co-author (Mark Maier), and on my own behalf, I would like to comment on two important questions raised by Dr. Vogel in her blog post. I also wish to comment on how the ES&T article was written and edited. Please note that I am speaking on my own behalf, and not as a Valspar employee. I am now in private practice and no longer a Valspar employee, entirely for personal reasons.
Comment 1:
“It would be useful to have a more complete description of the criteria used to select the different tests for estrogenicity and how these decisions were made and by whom.“
The Valspar BPA replacement project began in 2010, about the time I joined and started as an internal product development project. It evolved into an open collaborative effort after we found a candidate molecule (monomer) that worked in the coating, and that appeared to be biologically inactive based on in vitro screening. Molecular modeling was also done to see if we could predict biological screening results for other new molecules.
Once we had a few candidate molecules, we went to every science-savvy stakeholder we could find, and asked them what they would like us to do. We asked big international brand owners, the FDA, EFSA, EPA, ECHA, Dutch G4, French ANSES via the DGCCRF, and numerous universities. Perhaps most importantly, we asked Dr. Maffini and her NGO colleagues to put Valspar to the ultimate test of completeness and transparency. Finally, we asked the people we asked, whom they would ask, and so it went. I think we shared our data with everyone that expressed an interest in the US and EU, and ended up doing every study that that was asked for, including important studies that experts like Dr. Maffini wanted, such as mammary gland development studies.
So, to address Dr. Vogel’s first comment, science-savvy stakeholders made testing suggestions and Valspar followed those suggestions entirely. Each suggestion was either important, or in the case of animal studies, required. When an animal study was required by a regulatory stakeholder, we did everything conceivable to design studies that would acquire as much information as possible from each animal.
Comment 2:
“A discussion of the terms of the contract would make more transparent the level of independence given to the Tufts Laboratory.”
Valspar broke entirely from corporate precedent on how to work with academia, because there were no contracts and no terms. The only agreement was by hand-shake, with the understanding that the university laboratory could do whatever they wished with TMBPF to evaluate its estrogenicity (and other endpoints) drawing upon unrestricted gifts to the University. Moreover, Valspar chose to transmit test materials through independent third parties for verification accompanied by chain-of-custody. It was Valspar’s policy to leave academic researchers completely alone until they were ready to talk to us. Sometimes we had no idea what was happening for months.
But here is the real differentiator, and where I believe most industry research fails on transparency. We were offered, but declined to accept any form of non-disclosure agreement with academic collaborators regarding data or materials. There is much, much more to be said about this topic and how it can be done, but this is exactly how professional trust and ethics should work. Valspar had the fortitude to refuse any legal barrier to prevent academic collaborators from speaking freely about anything.
Comment 3:
Editorial control is an important transparency consideration. Because Valspar wanted no influence over the scientific results or interpretations, it was my charge as an employee, to let the scientists who did the work, write about the work. There is little more to say than that. Dr. Maffini coordinated much of the writing and summarized the animal studies done by contract labs. Peer reviewer comments and questions were also addressed in a similar manner. My core responsibility was to write packaging context and provide segue for components of the project.
I appreciate the blog and the comment. If this info is true, it is a remarkable, admirable model for how all vested parties can collaborate. Why can we not work together for the benefit of all?