The "Gold Standard" of double-blind placebo-controlled studies for scientific evaluation of benefit remains generally golden. This golden tool, however, is just that-one tool or rather an outline of how to obtain the data. Fundamental reassessment and change must occur if this industry wants to provide both the products and the information about these products to consumers in this country. You will not find the specific descriptions of these fundamental changes here because the topic is far too large for one person to resolve independently. This article puts forth the view and belief of what we can and should do to attain these changes. The end result is the identification of an alternative method of clinical evaluation applicable to all healthcare products but that would directly and specifically benefit the natural products industry. Sounds like a really BIG idea, and perhaps it is. Regardless of the size of this idea, it is one that is worthy of this industry to embrace and apply in order to make a positive step forward into the future. This alternative approach to clinical evaluation is mandatory not only to accurately identify the benefits of these products, but its acceptance is the best way to assure legally acceptable means of telling consumers about these benefits.
The Evolution of Clinical Research
When we look at where we are today with the clinical evaluation of products that provide health benefits-and this includes foods, drugs and dietary supplements-we are looking at a system that has evolved over time. The evaluation of drugs (i.e., ethical pharmaceuticals) serves as the foundation model where the requirement is for a high volume of rigorous study and statistical evaluation. The determination of endpoints, identification of effect using rigid statistical models, the identification of single entities responsible for the effects, and the overall evaluation of "risk/benefit" are the hallmarks of this process. The recent "Guidance of Substantiation of Claims for Dietary Supplements" merely reinforces the application of this evaluative approach to these products.
This system of evaluation has evolved over time with the primary intent being to provide drugs to the medical consuming community that are going to do something while establishing uniform controls over the evaluation of these products. The evolutionary changes occurred naturally as well as from discovery of negative occurrences when approved drugs do things that are not positive. This evolution has gone so far that it makes less and less sense for the evaluation of natural products and likely precludes approval of some legitimate drugs. The mandates out of this evaluative system are driving what passes for "drug development" today.
The target drug of today is best identified as a designer molecule, conceived on a computer and then synthesized. The clear objective is to produce a material that has dramatic physiologic effect. These designer molecules target specific, single points within a complex organism. Frequently, some of the other systems within the complex human organism this same designer molecule affects yield negative results. These are most often referred to as "side effects." In reality they are as equal as the "positive effect" the designer molecule has-the lone difference being that these additional effects are not what was originally sought.
If we apply these same rigorous standards to botanical entities or even something simpler such as resveratrol or folic acid, the results typically do not pass muster for successful acceptance of the demonstration of health benefit. Part of the issue stems from the statistical side of the evaluation.
The statistical requirements in our current model mandate strong effects demonstrated by the subject entity. The reason for this is that in order to achieve the level of statistical significance required, something pretty dramatic from a physiologic perspective must occur. It is the form and model of the statistical evaluation that causes the "development" of strong designer molecules. Without these dramatic effects the statistical significance will not be sufficient to pass muster and gain approval. Subtle improvements in a person's condition are not enough to nudge the needle on the statistical applause meter sufficiently to provide endorsement. And the additional requirement of the total "purity" of each entity under study compounds the challenge.
The current clinical evaluation model insists that we pursue the design of a "single entity" that can be identifiably responsible for physiologic effect. This approach is pure Ehrlich in philosophy since it requires the identification of the single "magic bullet" capable of addressing the disease or symptom. The single-entity stumbling block may in fact be overcome by isolation extraction from a specific botanical, but once you have this single entity or even something manmade like folic acid, you still face the task of showing dramatic effect.
The statistical model is also partially responsible for another notable challenge in applying our fully evolved evaluation system. The subjects used in drugs trials today must be in a very, very bad way. In other words, it does little good statistically to demonstrate incremental improvement in modestly ill individuals. The statistical model won't allow for it.
This additional shortfall was perhaps best demonstrated in the initial NIH St. John's Wort study. The inclusion criteria for subjects mandated that only individuals with "significant clinical depression" were candidates. But the St. John's Wort extract that was used in the study has been and continues to be used successfully in Europe for the treatment of "mild to moderate depression." There is a vast chasm of clinical presentation between "mild to moderate" and the mental status of the patients included in the NIH trial. So the system for evaluation virtually ensured a failure of St. John's Wort in this study. The only good news for the natural products industry out of the study was that it was a three-arm study involving St. John's Wort, a prescription drug and a placebo. It seems the prescription drug did not perform well either. So it appears as though our over-evolved system impacted what amounted to a re-evaluation of an already-approved prescription drug. This did not have an impact on the marketing and/or continued approval and sale of this drug, but the negative publicity surrounding St. John's Wort continues today. The bottom line is the clinical evaluation model we have today is not designed to identify effects in people with "mild to moderate" anything.
One last challenge concerning the use of our highly evolved model for clinical evaluation is that the demonstration of a product's positive effects in the area of prevention is extremely difficult to capture. If an individual consumes a product for a sufficiently long period of time and their identical twin does not, perhaps then there might be sufficient demonstration that the consumption of this entity prevented some adverse health consequence during their lifetime. This is a thornier problem than the others identified here, but there are two important elements to remember. First, drugs are defined in our law as those entities that claim to "cure, treat, diagnose, prevent or mitigate" a disease or disease state.
The demonstration of prevention is a significant challenge. This is perhaps one of the reasons why there are so few drugs that make preventive claims and so very many that detail their effects on the symptoms of diseases. Another reason this is important is that our regulatory agencies have limited health claims (i.e., drug claims for food products and dietary supplements) by their interpreted definition as exclusively prevention claims. Following the path of least resistance is a human trait and thus the exploration of preventive entities is not often a path chosen. This is likely why there are so few demonstrations of prevention of disease through drugs; only compilations of studies indicating that consumption of some foods yield protective benefits for health consequences.
These are not simple challenges to overcome but the answer lies in one specific outcome. The first step is to identify the problem with some specificity. The problem is that the current clinical model for evaluation of health benefit is not universally applicable. The three most important parts of this lack of universality are:
1. The current clinical evaluation model mandates strong to harsh effects in order to demonstrate adequate statistical impact. This eliminates entities that exhibit subtle or moderate effects. The evaluation of effect must include conditions that are neither dramatic nor conditionally obvious. This is akin to the application of the right amount of force to achieve a goal. This exclusionary aspect must be addressed in order to better identify and characterize positive yet subtle improvements and positive outcomes from a wider array of ingredients.
2. The evaluation of entities must allow for well-recognized and identifiably consistent entities but must not mandate exclusively single chemical entities. There likely should be some allowance for the characterizations to exist within ranges of strength among individual constituents while ensuring that the characterization is sufficient to distinguish. For example, identification of literally every chemical entity within a botanical extract should not be the requirement, but definition of the extract needs to be acceptably well defined.
3. The model for clinical evaluation must include easier demonstration of preventive effects. Such demonstrations should not take lifetimes to prove nor, ethically, should they mandate that populations "treated" in the placebo group undergo disease processes that could have been prevented. There are epidemiologic approaches to this to be sure, but inclusion of these studies is problematic both in terms of acceptance and in terms of generation of data.
These three principal issues are not the only ones to be ferreted out during the solving of the challenge, but they remain the most obvious and likely most important. Therefore the questions become: "How do we address these issues in developing a clinical model?" and "What do we do with it once we have one designed?" The answer to the second question comes out of the generation of the answer to the first.
Meeting of the Minds
The solution lies in the generation of an alternative model for clinical evaluation of health benefit. This model must come from those capable of developing such an item in conjunction with those holding best knowledge of the products in question. The proposal here is to take the task head on and apply readily available resources. The construct is a symposium or gathering of a limited number of the finer minds on the applicable subjects: epidemiology, statistics, clinical study, botanicals and other nutrients.
This gathering should not be held under the auspices of any trade association, but rather under the purview and guidance of a more scientific body, i.e., the International Life Sciences Institute (ILSI). The individuals selected could, for example, participate in a weekend summit cut across academia and clinical settings, which must also include minds from industry to temper the thought to more pragmatic applications.
The designated participants should number no more than 10 to 15 of these luminaries of thought. Their status must be sufficient to garner credibility with the certainty that they hold exemplary knowledge in their chosen field and the ability to listen to the input from the other participants and imagine an approach based on this multi-faceted input. Their advance preparation is the key to the success of the symposium.
The symposium format calls for one day of direct presentation and a second day of roundtable discussion with the outcome to be a single document. All the work done in advance of the meeting would be individually available for publication by the generators of the material, while the outcome of the symposium would be a consensus document intended specifically for publication. Attendees to the symposium would be treated to the finest in thought on the subject and would become witnesses to the future advancement of health benefit identification and claim.
At the conclusion of the symposium, the findings and the generation of this "Alternative Model for Clinical Evaluation of Health Benefit" is compiled, formally written, edited and subsequently published in a peer-reviewed journal. This task is the simplest since the groundwork for the information is solid and the individual contributors now hold significant stake in the idea and potential advancement of status. It is this publication that becomes most critical to the overall success of the project. Publication after peer-review validates these alternative ideas and thoughts even within the current system.
Presenting the Idea in Washington
Once published, the data, the article, the thoughts and the "Alternative Model" are prepared for presentation to the most important audience of all: FDA. The presentation is offered as a Citizen's Petition to address the clinical evaluation of entities that purport to provide health benefit. Rather than request that the FDA generate something, the alternative is prepared for them. Much like a self-affirmation of GRAS (Generally Recognized As Safe) using independent experts in the field, the input has already undergone review by a cross-section of experts on the topics involved. It is here that the trade associations play the most important role: support.
Supporting the idea without having to generate it and using the input from the finest minds on the topics available should be a simple process. The benefits of a positive acceptance of the "Alternative Clinical Model" cuts across more than just the natural products industry and additional support from other associations would be welcomed.
The development of an "Alternative Model for Clinical Evaluation" is beyond the scope of one person's ability to perform. Critical to this project is not only the development of such a model, but also broad acceptance. While this proposal includes a good many people, it is simple in its approach.
Projects are distinguished from processes by having an endpoint and outcome. Thus the "project" of developing a proposal for an "Alternative Model for Clinical Evaluation of Health Benefit" can be swiftly accomplished through a focused group, in a focused timespan, with the outcomes published as rapidly as they are generated. As we all know, there is a new administration in place in Washington. Perhaps there is an open ear, particularly if the presentation comes from the finest minds. Now is a very good time to consider putting together such a project.