Dr. Amy Roe: Botanical dietary supplement use in the United States is widespread and growing with thousands of products available in the marketplace. New data from the 2018 Council for Responsible Nutrition's (CRNs) Consumer Survey on Dietary Supplements found that 75% of U.S. adults take dietary supplements. Although vitamin and mineral supplements remain the most popular categories, the overall use of herbals and botanicals has significantly increased in the past 5 years.
These data are also consistent with the annual report on herbal supplement sales in the United States, provided by the American Botanical Council. This year, the ABC report indicates that for 2017, herbal supplement sales in the United States increased 8.5%, topping $8 billion. This was the strongest sales growth in more than 15 years, and there appears to be no slowing down in sight.
There is an interesting dichotomy, however, in that despite the significant expansion in the marketplace of dietary supplements, there is an accompanying public concern over the quality and safety of many of these products. The FDA's creation of an Office of Dietary Supplement Programs is a welcome response to the rapidly expanding dietary supplement industry and should make available more resources to address the public's quality and safety concerns.
Likewise, I think great strides in ensuring the quality of dietary supplements have been made by responsible manufacturers working through trade associations such as the CRN and others, to participate in programs including the Online Wellness Library (OWL). Participation in the OWL program shows manufacturers' commitment to transparency in product labels, and involvement in the mainstream supplement marketplace.
From a safety standpoint, the regulatory framework for botanical dietary supplements in the United States is grounded in the 1994 Dietary Supplement Health and Education Act, or DSHEA, which does not require premarket safety evaluations for ingredients that were in the marketplace before 1994.
New dietary ingredients require notification to the FDA, including history of use information or other evidence of safety. However, there are no specified requirements for premarket testing. Instead, manufacturers have been provided with the draft guidance document that roughly outlines safety tests to perform. The FDA then evaluates the submitted package to determine whether the sponsor has provided adequate substantiation that the new dietary ingredient is reasonably expected to be safe.
Preclinical safety data often provide the best indication of the new dietary ingredient's safety. However, it is only available for a small number of botanical ingredients. I would like to pause at this point and open the roundtable up to questions and discussion, and I would like to begin with Dr. Swift of the FDA.
Dr. Swift, since there are no specified requirements for premarket safety testing of dietary supplements, what is the nature of most of the preclinical safety packages that you see?
Dr. Sibyl Swift: Thank you, Dr. Roe, for that introduction. Most of the packages that we see rely upon in vivo safety studies. There is a combination of some in vitro but primarily it is mostly in vivo data.
Dr. Roe: So, animal studies have been the historical standard for assessing toxicity, but the number, complexity, and variable nature of botanicals seems to preclude their use as a pragmatic solution. Dr. Rider, I know that the National Toxicology Program has substantial experience in conducting animal studies with botanical ingredients. What difficulties have you experienced conducting such studies with these complex botanical mixtures?
Dr. Cynthia Rider: The National Toxicology Program has really faced three major challenges in the areas of test article selection, comparing our animal study data to actual human exposure levels, and in identifying the active constituents. To go into some detail on these, for test article selection, there are numerous versions of any one botanical ingredient available in the marketplace and we have to select a single sample to assess in our toxicity studies because of time and resource limitations.
We have to consider many different things in test article selection; such as which material is the most representative of that botanical ingredient? Which is closest to what people are actually being exposed to, so which one has the greatest market share? What material represents the worst-case scenario? Because our goal is to protect public health, these are considerations that go into test article selection.
Once we generate data with a test article, we want to be able to compare our findings in animals to human exposure scenarios. We have to use absorption, distribution, metabolism, and elimination (ADME) data to make that comparison. When we are doing that, it is not always clear what the active constituent of the botanical is, and which marker compounds we should be following in those studies.
That leads to the third challenge of identifying the active constituent. It is very helpful to know what the toxic constituent is, which facilitates our mechanistic understanding and also might facilitate any regulatory action on a botanical.
Most of the work in the area of identifying active constituents has been either from a drug development perspective or looking at efficacy. It is important to do this kind of work on identifying active constituents from a toxicological perspective because it is not always the pharmacologically active constituent that is toxic.
Dr. Roe: That is interesting, and I would like to talk more about some of those topics that you raised. Before I do, I want to give an opportunity to our industry representatives to add further input before we move away from talking about animal studies. Could you share some of the hurdles that you all have encountered with conducting in vivo studies? And, from an industry perspective, what do you see as the greatest challenges in using animals for testing botanical ingredients?
Dr. Daniel Marsman: Thank you for giving us this opportunity. The current advancements in our understanding of both botanicals and toxicology make this a very important question. Animal studies have been a critical tool for advancing our understanding of chemical hazards. However, modern science has provided many in vitro and in silico tools that allow us to take a much more mechanistic approach to the understanding of toxicology. Where these tools have incorporated human cells and tissues, we have the added potential for inference for our human safety risk assessments.
Complex mixtures, on the other hand, have always been a challenge for the toxicologist. However, with current in vitro and computational systems and our analytical sophistication, we have the ability to understand botanicals at the constituent level, and the potential to understand the role of constituents or groups of constituents with much greater thoughtfulness.
Given the time, cost, and complexity of animal studies, coupled with bioethical considerations, 21st century toxicology tools allow us a much greater opportunity to understand complex botanicals.
Dr. Roe: Dr. Marsman, as you mentioned, 21st century tools offer great promise in bridging some of the data gaps that we have with botanical extracts and perhaps more so for toxicity endpoints over others.
There are still numerous uncertainties in the appropriate application of these alternative methodologies in the evaluation of the safety of botanical ingredients. So, how is the Botanical Safety Consortium structured to address some of these key data gaps and toxicity endpoints?
I will start with Drs. Swift and Rider, since you are both leading this Consortium, if you would like to begin by talking about the overall structure of the Consortium and how it might begin to use some of these alternative methods to address safety endpoints for botanicals?
Dr. Rider: Thanks for the question. The Consortium is comprised of a good balance between government, industry, and academic partners because we all bring our different perspectives to the challenges. We have selected a number of different endpoints based on both adverse event reporting and signals of toxicity in the peer-reviewed literature, so that is what has driven our endpoint selection.
Dr. Roe: Do you think that some of these endpoints will be easier to address by in silico or in vitro tools that are already available, or do you think that you will have to really develop these tools to address them?
Dr. Rider: I think it will be a combination of both identifying existing tools and also identifying areas where more development is needed. For example, hepatotoxicity and genotoxicity might have a lot of established methods that are tried-and-true, whereas we might need to do more development work to find effective in vitro models for cardiotoxicity.
Dr. Roe: Thanks. I would like to hear more on which endpoints the group believes are important, because I'm wondering whether across government and industry there is an agreement in terms of what some of the most critical endpoints are. Perhaps one of our industry representatives could speak to this?
Dr. Joseph Dever: When we scan the globe and take stock of the primary concerns from regulators, one of the main targets is certainly the liver. Many of the testing requirements in place today focus on the liver with kidney a close second.
To go back a bit to what Dr. Rider was saying and your question about existing tools versus new tools, I actually think that there are many existing tools that can be applied. The trick is to learn how to apply them with botanicals. And botanicals are a bit different because they are a complex mixture, but the tools are there and that is the exciting part of this.
With regards to endpoints, one of the critical endpoints that needs to be addressed is repeat dose toxicity. Supplements are typically taken repeatedly, for a long period of time. It is important to develop or apply in vitro technologies that take into account that sort of exposure. So, that is going to be a very important outcome and type of endpoint; focusing on systemic toxicity.
Dr. Rider: Let me add briefly some detail on the structure of the Consortium. We have a steering committee, which is comprised of this panel, focused on the direction we are moving in. In addition, we have multiple subgroups that are dedicated to making decisions about which are the appropriate endpoint assays and which botanicals should be included in our evaluation. The subgroups include chemical characterization, hepatotoxicity, genotoxicity, reproductive and developmental toxicity, cardiotoxicity, and systems toxicity, which Dr. Dever just mentioned.
Dr. Swift: You asked if we have noticed that there are different endpoints of interest between industry and the regulators. And the answer is no, actually; we all have one primary concern, which is that we want to ensure that the products consumers have access to are safe. Therefore, we don't have different endpoints of interest. That has not been an issue as of yet. In fact, we have identified quite a few commonalities that we did not realize previously.
Dr. Roe: Great. Dr. Rider, you mentioned the challenges that National Toxicology Program (NTP) has dealt with. For some of the challenges mentioned, such as identifying the active constituent and toxicity associated with various constituents, do you see being able to address these with alternative methods, perhaps in silico or in vitro methods?
Dr. Rider: Yes, I think they hold great promise for contributing to the overall portfolio of tools that we have available to assess the safety of botanicals. One example of that can be in our test article selection phase. We are only going to select one test article, as I mentioned, if we are doing a 2-year cancer bioassay in mice and rats, for example. We really want to understand how that test article relates to other things that are available in the marketplace if we are going to extrapolate our findings to real-world scenarios.
That concept of sufficient similarity between different samples, or phytoequivalence, is one where we can systematically show that our test article looks like other products on the marketplace and actually have sound science to support that. I think using in vitro methods is really helpful because we can test multiple different lots and look across the biological activity patterns and use that to provide context in our work.
Dr. Stefan Gafner: I'd like to add something regarding identifying active constituents, which is an approach that I find particularly interesting. It is an approach published by the University of Illinois at Chicago, providing what they call “knockout extracts.” The researchers use an herbal extract where they have selectively eliminated certain compounds, and test those in the same assays.1,2 That can sometimes give you an idea if the compounds that have been eliminated are actually responsible for a given activity or toxicity.
Dr. Roe: That is another interesting point that you make, Dr. Gafner. You bring up the great analytical power that we have access to. And bringing together the analytical chemists, the natural product chemists with the toxicologists is absolutely crucial to these complex botanical extracts, in terms of understanding what we are testing and how to test.
And that is going to be a great challenge among the various subcommittees associated with this Consortium—how do you decide from an in vitro standpoint what to test. Do you test whole mixtures? Do you break down the extracts into individual phytoconstituents and test those? I think the example you mentioned may be one process or method that could help us. Several of you have mentioned the complexity issue, and clearly that is going to be a challenge with some of these in vitro assays. Would anybody like to add to that?
Dr. Rider: Yes, I will add that our chemical analysis group will work with the different subgroups because that is such a key issue that needs to be addressed across the board. Hopefully, the chemical analysis group will provide recommendations on approaches for characterizing a test article and the minimum requirements for analysis.
Dr. Dever: Complex mixtures including botanicals have been frustrating for toxicologists because there is usually an unknown fraction, and there exists the possibility for additive or synergistic interactions between constituents. That being said, with the right tools, it is certainly possible to understand the safety profile of a botanical even if you don't know exactly what constituents are responsible for an effect. The goal of the consortium is to bring experts together to more clearly establish what data are needed, and what data may not be needed, for completing botanical risk assessments.
Dr. Roe: Before we leave the topic of botanical raw materials, I wanted to follow-up with Dr. Gafner. I know that the American Botanical Council has been instrumental in forming the Botanical Adulterants Prevention Program that focuses on accidental adulteration that occurs as a result of human error or poor Quality Control procedures, as well as the intentional adulteration of plant-based products for financial gain.
Are there other issues beyond that, that you would like to see addressed by the Consortium as it relates to adulterants and contaminants?
Dr. Gafner: There are a number of publications that suggest certain well-established ingredients to be potentially liver toxic. I take the example of skullcap (Scutellaria lateriflora, Lamiaceae), where there are a number of case reports in the published literature that suggest it might be liver toxic.3,4
On the other hand, it is well known that there has been a problem with substitution of skullcap with materials from plants from the genus Teucrium.5,6 This type of adulteration has been later established as the reason why we have the case reports of skullcap toxicity.7,8
For the Consortium, it might be interesting to look at not only the authentic plant species used in herbal medicine, but also to use some of the known adulterants and see if there is any implication for liver toxicity.
There are a number of other issues that we see with herbal ingredients. Contamination is one. Lead contamination may be something that in certain instances can cause toxicity that otherwise the plant does not have. All in all, there are a number of quality issues that play a big role in how a plant may be safe or not. Looking at those issues will be challenging, but helpful in many cases.
Dr. Marsman: I would like to build on that by coming back to our topic about analytical characterization, because I think test substance characterization will be a critical step in the process of the Botanical Safety Consortium (BSC). For botanical toxicologists, this is not unique to this initiative, as test substance characterization has always been a critical step.
Through the Consortium's systematic investigation, we hope to be able to also answer some of these challenging issues, such as the safety implications of minor variations in botanical raw materials due to harvesting or processing differences or between unique, but similar botanical cultivars and species, like Dr. Gafner pointed out.
Dr. Roe: It would be great to hear how the Consortium will choose the ingredients or botanical extracts to focus on. Will each subcommittee choose their own related to their endpoint or other?
Dr. Rider: We definitely want subgroups to choose botanicals that reflect both potentially positive and negative effects on the endpoint of interest. For example, in the hepatotoxicity group, we would like them to pick some botanicals that have adverse event reporting of hepatotoxicity, or we have good animal data to suggest there is hepatotoxicity, so that we can compare our results as part of the evaluation. For the different assays, we want to compare our in vitro and in silico methods to established human and animal data.
In addition, we hope to include some examples where we vary the test article, so we can address some of those sufficient similarity questions within this context by including multiple versions of the same botanical.
We envision that each of the subgroups will nominate their botanical ingredients that represent both positives and negatives and we will come together as a group to decide on a final list of the botanicals that will be tested in the program. Then, the final ingredient list will be run through the various assays that are recommended by the different subgroups. So, we will have this body of data and we hope to create a public-facing database for housing that data from the Botanical Safety Consortium group.
Dr. Roe: What is the dissemination plan for the key learnings, especially the implications? What is your plan for sharing that externally?
Dr. Swift: Ideally, we will be having a panel discussion of the consortium at the International Conference for the Science of Botanicals in Oxford, at the University of Mississippi, in April of 2019. After that, we will consider holding a public meeting to request feedback and encourage participation. As far as a plan to disseminate the data and key learning, we plan to let the science speak for itself by sharing the data in publications and at conferences and speaking engagements.
Dr. Rider: In terms of implications, we hope that the work of this group will really inspire the field to do more evaluation of safety. We hope to help industry, regulators and government scientists in understanding how to apply these new methods to the goal of providing high quality, safe products in the marketplace.
Dr. Marsman: We were talking before about the structure of the BSC and the endpoints of concern. As Dr. Swift mentioned, we are all seeking a common goal of protecting the consumers. From an industry perspective, we are always trying to do a holistic evaluation of the safety of our products.
Oftentimes, that evaluation starts with acknowledging that a herb may have a history of significant use. But that use may be in a particular context, and so when we are evaluating endpoints of concern, what we might lack from our history of significant use is specific targeted data on potentially unique subpopulations of individuals that might use our products. So, we pay attention to address those endpoints where the epidemiological data may be inadequate to address those points of concern.
Dr. Gafner: If I may build on that, generally, we need to distinguish between the new ingredients and those that have a long history of use. My comments here are specific to those that have a long history of use and these are based on discussions I have with health care professionals. As such, it is a little bit more focused on the real world and the concerns that they have.
What I hear from pharmacists and physicians is that they would really like to have data on the safety in specific populations, as Dr. Marsman mentioned, and, in particular, pregnant women and small children and babies, because these are people that often do not want to take conventional drugs. These patients are concerned about the adverse risks of taking these drugs and they turn to alternative medicine. But there is not much data on safety in these particular populations. I think that is where any data would be very helpful.
Dr. Dever: History of safe use in humans is probably the most important botanical safety metric. You always start there when doing an assessment. And it comes in varying degrees. Some botanicals have a very long, safe history of use, which is an important consideration, and some not so much.
So, the Consortium is trying to work within that context. When history of use is not enough, and you have additional questions, whether it be for sensitive populations or simply on the inherent safety, these tools are what we want to develop to help answer those questions in a pragmatic and meaningful way.
Dr. Marsman: I agree. And, as was briefly mentioned by Dr. Rider, we will be looking at addressing systemic concerns. ADME, in particular absorption/disposition, is an area that we want to continue to emphasize because we may have a significant history of use but it may not always be in the exposure route or a specific context that we are studying. So, being able to normalize the playing field with the use of systemic exposure would be a big advantage to the work of the Consortium.
Dr. Roe: I know that the Botanical Safety Consortium is primarily focused on the United States, but can any of you briefly describe some of the challenges that the dietary supplement industry is facing globally? Do you think that efforts of this Consortium can help address some of these global challenges?
Dr. Dever: Absolutely, and it goes back to what Dr. Swift said earlier, which is we are going to lead with the science. That is what this Consortium is about, plain and simple.
Around the globe, there is great diversity in the ways that dietary supplement safety is approached. There is not what you might call a global standard with regards to botanical safety. That is a real challenge when you are trying to bring products to market in these different parts of the world, because the requirements can vary so much.
I think the Consortium is an effort to move toward something that can provide greater alignment; letting the science lead, but also providing some processes and tools that can bring a bit more alignment to how botanical safety can be approached.
Dr. Marsman: There are many challenges, some we have discussed today, including the limited mechanistic scope and utility of animal studies, lack of uniformity in the trade in defining what adequate analytic characterization looks like, and inherent variability of complex natural ingredients, and the decades of focus of testing single chemical entities in our toxicity assays.
There is currently a lack of fit-for-purpose methodologies for addressing botanicals. When you compound this with the incredible number of botanicals that are used in consumer products around the world, it creates a dilemma. Because of these competing vectors a vast array of safety approaches is taken locally by boards of health.
From an international perspective, I think there are many of us who would welcome a scientifically credible process for the safety evaluation. Our focus is on botanicals in food supplements, but there is no reason to assume that this scientific process for botanical safety would not also have broad application outside of the food supplement industry.
Dr. Gafner: I believe the challenges for safety evaluation for herbal ingredients are the same globally. I am from Europe and I know that they are in a similar situation as the United States with regard to what is known about safety and toxicity of herbal ingredients.
One of the challenges that I see—and I think where the Consortium can make a major difference—is that there is a fair amount of contradictory data published on certain herbal ingredients. That might be animal studies or cell-based toxicity studies. For a health care professional, or even for an industry group that wants to bring a product on the market, it is difficult to make sense of all the data. If we can help to get a clear indication of whether the published data are clinically meaningful or not, by having the right tools, I think that would go a long way.
Dr. Marsman: Yes, agreed. I want to add that some may have concerns about global harmonization in this area. And it is wise to respect how incredibly difficult harmonization activities can be. But given the relatively recent acceleration of botanical safety science, I believe we have the opportunity to get ahead of the curve and offer solutions that may resonate globally.
The newness of the industry relative to others may afford us the opportunity to develop standards more easily without quite the same degree of entrenched views on scientific methodology. Our goal, as Dr. Swift pointed out, is to show what good science looks like. And that suggests it could readily be adopted by other countries. So, while we do not really see harmonization as a primary objective of the BSC, we hope to arrive at a place where our safety evaluation meets the needs of other countries too.
Dr. Roe: In conclusion, I would say that resource-efficient in vitro tools are needed that can accommodate complex mixtures, represented by these botanical extracts. And, as a few of you mentioned, many in vitro models are available—for things like predicting disposition, absorption, metabolism, even bioavailability, as well as toxicity across many biological endpoints—and they could be incorporated into hazard identification and risk assessments of natural ingredients.
The availability of in vitro models with in vivo and clinical predictive power to manufacturers of dietary supplements, coupled with the regulatory acceptance of data generated from such systems, would really help ensure the safety of natural products in the marketplace.
As Dr. Swift mentioned, that is a consistent goal across the Consortium, whether it is from members of the industry or the regulators. There is that consistent dedication to ensuring the safety of these products in the marketplace.
Would anyone like to add any final closing thoughts?
Dr. Marsman: One thing I would add is the BSC is really focused on botanical safety science, however, our important emphasis on the analytical characterization could be helpful in other areas. There are multiple industry-led initiatives looking at such improvements as ingredient traceability and supply chain integrity, and the auditing standards and capability of manufacturing sites.
Adequate ingredient identification is one aspect that cuts across many of those initiatives. One of the biggest threats on the integrity of the legitimate dietary supplement industry is economic adulteration, with either direct or indirect incorporation and sale of prescription drugs claimed to be dietary supplements. While, the work of the BSC is not focused on clinical good manufacturing processes best practices for a botanical, adequate analytical characterization would be a central tenet in any good program to eliminate this concern.
Dr. Roe: And there is an analytical subcommittee as part of the Consortium.
Dr. Marsman: Yes, thanks. Its aim, in my mind, would be to focus on best practice substance characterization, to make sure that the toxicity assays that we develop and evaluate are done well, and we know what we tested. But, that same analytical characterization of the test substance materials will hopefully set the standard of what “good” looks like in terms of characterizing a botanical, like Dr. Gafner said, whether it is a new botanical or a variant of an existing botanical.
Dr. Dever: I will add that, from a safety standpoint, it's really important for people to realize that in vitro and in silico tools for botanicals have just not been tapped. It is actually quite astonishing how little they are used compared to other industries. And that is why there is such a great opportunity to make really rapid progress.
There are going to be challenges. There are going to be tricky things that have to be worked out, like there always are when you study complex mixtures. But I think we can rapidly improve the situation for botanicals by simply leveraging well-established models and tools.
Dr. Gafner: What I believe is that the Consortium is not there to have people scared of using botanicals, but to provide the science to show that they can use them safely.
Dr. Rider: I'd like to add that we are trying to engage with stakeholders who want to be involved in the effort, and so hopefully through this article, as well as our dissemination in other forms, we will engage with as many stakeholders as possible on this topic that we think is important for public health.
Dr. Roe: If there are individuals out there who read this article or hear about this, what is the best way to approach the Consortium about getting involved?
Dr. Rider: Yes, using our contact information to e-mail would be great or if you are already engaged in the International Conference for the Sciences of Botanicals, held each April at the University of Mississippi in Oxford, or other botanical meetings, we are planning to connect with people there as well.
Dr. Marsman: The only thing I would re-emphasize is that this Consortium is a science-based endeavor. We want significant industry and trade association participation. But particularly within the subgroups we want the scientists who read this to step up and be a part of those working groups because that is where the real work is going to get done, and that involves not just industry but academia and government as well.
Dr. Roe: Thank you all for your participation in this panel. This was a great discussion and we look forward to hearing the results and progress from the Consortium.
