Summary: The increased focus on quality control in the herbal industry and the publication in 2003 of proposed Good Manufacturing Practices (GMPs) by the Food and Drug Administration has generated increased interest in laboratory methods to determine the identity, purity, quality, and potency of botanical raw materials as well as finished botanical products. Laboratory methods include microscopic analysis of botanical characteristics of herbs as well as analysis of the plant material’s chemistry. Analysis of chemical constituents of herbal materials is one of the primary areas of interest among quality control technicians. Various chemical methods are available to detect the presence or absence of “marker” compounds and/or active compounds in herbal extracts, as well as to detect the presence or absence of potential adulterants. In order for various chemical methods to function, the laboratory must have access to high quality “reference substances,” including pure “reference standards,” the pure compounds used to calibrate the analytical machinery. This article discusses the definition of pure chemical reference standards and their vital role in herbal quality control.

A discussion about reference substances should begin with the concept of inherent quality. This notion is connoted in the phrase, “You cannot test quality into a product. You can only produce it.” That is, appropriate testing is a necessary tool to determine the inherent quality of a substance.

GMPs and Reference Substances

Since the publication in March 2003 by the U.S. Food and Drug Administration (FDA) of proposed new good manufacturing practices (GMPs) for herbal products,¹ there has been increased interest in all areas of the herb industry regarding issues of assuring quality and proper identity of herbal raw materials and finished products.

For many years prior to the FDA’s publication of the proposed GMPs, responsible manufacturers of botanical products had already committed to improved procedures to ensure quality control and quality assurance. Many of these procedures were beyond the formal requirements contained in existing GMPs for conventional foods, the standard which herb and dietary supplement manufacturers have had to meet. For many companies, quality testing has already become an integral part of most of the steps in the entire production process for herbal preparations. This includes harvesting and post-harvesting steps, the extraction of the herbal preparation, and finally, the finished product itself. The crucial point in quality testing is the question, “How are valid and reproducible results achieved?” It is the duty of a scientist to question his or her own results. So how can one achieve the goal of producing reliable and “true” results?

When testing parameters like length, weight, or volume, a scientist can rely on official reference sources that are directly traceable to a common verified standard. However, specifying and testing the quality of herbal products is not reducible to measuring physical dimensions; assessing herbal quality requires dealing with the inherent complexity and variability of botanicals. From a European perspective, particularly with respect to clinical trials, the active principle of an herbal product is the herb itself, a complex mixture of multiple compounds. Despite the enormous scientific efforts that have been made in past decades to determine the mechanisms of action for herbal preparations, there are still very few herbs in which the observed therapeutic effects can be assigned to the activity of one or even a few defined constituents.

Thus, the preparation in its entirety must be regarded as the active part of an herbal product. However, insofar as botanical nomenclature is used to define a plant in commerce, the use of common and scientific names for herbs is not as exact as the use of chemical names for specific chemicals and conventional single-chemical drugs. For example, everyone in the herbal field knows that the name “valerian root” is not as precise a definition of an active substance as is the name “acetylsalicylic acid” (aspirin). The term “valerian root” is a generally accepted, standardized common name referring the root (more accurately, the root, rhizome and stolons) of Valeriana officinalis L., in the family Valerianaceae. Furthermore, valerian root (often referred to in Europe by its pharmacopeial name, Radix Valerianae) is a defined entity in various pharmacopeias (e.g., the German and European pharmacopeias). These pharmacopeias contain specific botanical descriptions of the plant part used as well as the range of specific marker compounds, which are specified as preconditions for the use of the herb for official drug status. However, while “valerian root” is a precise term for a plant material, the actual chemical content of that material can vary either widely with respect to products so named in the marketplace, or within a narrow range of chemical variance for officially accepted material.

Therefore, despite the use of authenticated botanical voucher specimens to help assure proper identity (per the article on botanical reference standards in this issue of HerbalGram), modern concepts and methods relating to the quality (i.e., chemical consistency) of herbal materials and products pertain to phytochemical markers and fingerprint analyses. These markers are the threads that tie together production and quality control. The World Health Organization (WHO) states that “the purpose of having reference substances is to achieve accuracy and reproducibility of analytical results required by pharmacopeial testing and pharmaceutical control in general” and that “the general use of a reference substance should be considered as an integral part of a compliance-oriented monograph or test procedure used to demonstrate the identity, purity and content of pharmaceutical substances and preparations.”² In the United States, herbal products are regulated under different types of legal provisions than those in the European Union (EU); however, the concept of a marker-based quality assessment has become widely accepted in both markets. In the absence of knowing the identity of the compounds specifically responsible for biological activity, marker compounds have been used frequently as surrogates for quality determination.

Marker Compounds

In the EU the marker-approach is obligatory at least for those herbal products that are registered medicinal products. According to the EU’s Committee for Proprietary Medicinal Products (CPMP),³ herbal medicinal products must contain only herbal drugs or herbal drug preparations as active substances. For each herbal drug or herbal drug preparation, a comprehensive specification established on the basis of relevant scientific data must be submitted. The requirements are laid down in the CPMP Note for Guidance on Specifications: Test Procedures and Acceptance Criteria for Herbal Drugs, Herbal Drug Preparations and Herbal Medicinal Products.⁴ The specification must provide details of the characteristics, identification, and purity tests. Quantification of substances with known therapeutic activity or markers is obligatory. When pharmacopeial monographs with defined markers are not available, the applicant should select and identify suitable markers (“or other justified determinations”). The choice of markers should be justified.

In the U.S., herbal products can be sold as conventional foods (e.g., spices, teas), drugs, or dietary supplements. The regulations are set down in the Federal Food, Drug and Cosmetic Act (FDCA), with the provisions laying the groundwork for dietary supplement regulation provided in the amendment to the FDCA, which is called the Dietary Supplement Health and Education Act of 1994 (DSHEA). The product category depends on intended use of the product. A few herbs are still approved as non-prescription, over-the-counter (OTC) drugs, e.g., the stimulant laxative senna (Senna alexandrina Mill., Fabaceae; syn. Cassia senna L.) and the bulk laxative psyllium seed (Plantago arenaria Waldst. & Kit., Plantaginaceae.) Under the provisions of the FDCA, package labels are required to accurately reflect the contents of the package. There is no regulation or legal requirement that herbal dietary supplements or extracts used in dietary supplements should be characterized by declaring the content of marker compounds on the label.

According to FDA’s recently published guidance document for Botanical Drug Products,⁵ label declarations of marker compound content are not required for a botanical preparation that is intended to be evaluated for approval as a drug. FDA has written, “In the initial stage of clinical studies of a botanical drug, it is generally not necessary to identify the active constituents or other biological markers or to have a chemical identification and assay for a particular constituent or marker. Identification by spectroscopic and/or chromatographic fingerprinting and strength by dry weight (weight minus water or solvents) can be acceptable alternatives.” With respect to combinations of herbs, FDA says, “When multiple active constituents or markers are known, they should be chemically characterized and their relative amounts should be defined.”

However, a botanical drug product would usually have to conform to the criteria set forth in a monograph published by the United States Pharmacopeia, which may include a marker or group of markers and/or a fingerprint as a standard. Currently, almost all of the monographs published in the Dietary Supplement Section of the USP have incorporated within them an assay for a marker or markers. If a botanical product were approved as a drug by the FDA, the USP would likely move the monograph from the Dietary Supplement section to the United States Pharmacopeia.

In the EU many phytomedicinal products are produced in accordance to standards that are established to allow the product to be licensed as a drug. Accordingly, there is currently a significant tendency by European phytomedicine manufacturers to use a variety of phytochemical marker substances for the characterization of a particular herbal material. In the U.S., where herbal dietary supplements are often sold as dry powders of the raw material or as tinctures, such characterization is not required, although the identification of various marker compounds and the use of chemical reference standards in validated analytical methods can provide an additional measure to help ensure quality of the botanical material.

Many botanicals manufacturers and publishers of botanical monographs (e.g., the USP, European Pharmocopoeia [EP]) have set standards for minimum content of certain natural phytochemicals in raw materials and finished products. When these compounds have known therapeutic activity, assuring minimum content can assure therapeutic quality of the material. Unfortunately, the constituent or constituents responsible for the therapeutic activity of most botanicals is still unknown. In many of these cases, one or more natural constituents have been selected by manufacturers or monograph writers as surrogates for determination of quality of materials. These surrogate quality indicators are referred to as marker compounds. Generally, from a European phytomedicinal perspective, marker compounds should exhibit the following characteristics:

• be characteristic or unique for the herbal material or herbal preparation;

• be a substance with an established chemical structure;

• be present in the herbal drug preparation and finished product in sufficient amount to allow development and creation of a scientifically valid analytical method for both;

• be accessible to quantification with common analytical equipment (e.g., gas chromatography [GC], high-performance liquid chromatography [HPLC], high-performance thin-layer chromatography [HPTLC]) to keep costs of routine analysis moderate;

• be sufficiently stable under the conditions of storage and use of the final product;

• not be present in other herbs contained in the finished product when it is necessary to selectively quantify the content of one herb or herbal preparation in a multi-component product; and

• be commercially available or able to be isolated by the company in its own laboratory.

Table 1 shows some popular herbs and their typical marker substances. The marker substances can be designated as such based on their use of quality control purposes and/or because they are known active compounds.

Table 1: Selected popular herbs and their typical marker substances

* NF21 = National Formulary 21st ed. (published by the United States Pharmacopeia)

† AS = proven active substance; M = marker substance; FM = foreign toxic compound

[Note: Some compounds have proven activity and can also be used as markers due to their relative rarity or uniqueness.]

‡ Various Chinese herbs have been adulterated and/or substituted with herbs of the genus Aristolochia, containing the nephrotoxic, carcinogenic compound aristolochic acid. Such herbs that have been sometimes adulterated include Clematis armandii and Stephania tetrandra, but these herbs do not contain aristolochic acid.

Herbal Reference Substances

Herbal reference standards can be produced in two ways: (1) by isolation from natural sources from the plant material and (2) by partial or total synthesis, i.e., artificially via chemical processes. Generally speaking, any compound that can be separated from its natural matrix with sufficient chromatographic resolution on an analytical scale can be isolated from the same source in sufficient quantity to be useful as a reference standard. But the low concentration of many substances of interest in their preferred herbal source may make isolation prohibitively expensive and thus not economically viable. In some cases, though, a higher content of the same compound may be found in herbal sources other than the herb of interest.

In principle, chemical synthesis is an alternative. But many herbal compounds have very complex structures and a high specificity with regard to possible stereochemical isomers, which may make synthesis complicated and costly. Therefore, despite the enormous progress in synthetic chemistry, synthesis is—for economic reasons—still restricted to very few cases as long as a given compound is needed for analytical purposes only. Recent examples for synthetically derived reference standards are rubiadin and lucidin for noni juice (Morinda citrifolia L., Rubiaceae). These toxic anthraquinones occur in parts of the noni plant, but not in the fruits, and are therefore used to check the quality of the production processes and the finished product. They are easily accessible by synthetic pathways, though isolation is difficult because of the small quantity appearing in natural sources. Also, for one of the widely known reference standards hypericin from Hypericum perforatum L. (see Table 1), synthesis is a good alternative to isolation.

Viewing reference substances as a basis of ensuring the quality of analytical measurements (and by proxy, quality of the herbal material), it is evident that the quality of the reference standards themselves is of crucial importance to ensure a reliable and documented level of quality, safety, and efficacy of herbal raw materials and products. Although many herbal substances are commercially available today, the commercial analytical products are not always made from high quality materials. For example, Figure 1 shows two commercially available batches of the “reference standard” hyperforin (the phloroglucinol marker compound from St. John’s wort)—both labeled for a content (by HPLC) of greater than 90%—as received in the authors’ laboratory. By comparing the HPLC chromatograms of both, it is obvious that only one batch (blue line) is sufficiently pure to serve as a reference for the quantification of hyperforin in an herbal preparation. The second batch (green line) is useless since its labeled content is not its true content, as evidenced by the presence of numerous additional peaks in the chromatogram. The reader will notice that the putative hyperforin peak is not even the largest among them. It is possible that an inadequate analytical method was used to establish purity or that the batch was not tested regularly to assure that it did not decompose over time.

When it comes to an appropriate handling and use of chemical reference standards, it is necessary to have an appreciation of the different substance types that are commonly available. This appreciation is necessary for both economic reasons and to ensure that the reference standard is of sufficient quality to meet its intended use. In short, some testing processes or procedures may not require a material that is 99.9% pure, while others might. It bears mention that purity and costs of reference standards are directly related: the more pure the material, the higher the cost. So using a more pure standard when a less pure material might do is not an economically sound decision. In contrast, using a less pure material when a more pure compound is appropriate could invalidate an entire series of expensive analyses.

According to the WHO, there are two different types of chemical reference substances that can be distinguished in connection with the quantification of compounds in an herbal product: primary and secondary reference substances.²

“A designated primary chemical reference substance is one that is widely acknowledged to have the appropriate qualities within a specified context, and whose value is accepted without requiring comparison to another chemical substance.”² This means that only highly purified compounds, with defined and documented identity, purity, and content can be used as primary reference substances. Essentially, the primary reference substance establishes a fundamental value for a specified analytical method, on which all other definitions of content depend.

WHO states that “a secondary chemical reference substance is a substance whose characteristics are assigned and/or calibrated by comparison with a primary chemical reference substance.”² This comparison allows for the definition of the content of a particular substance without generating the complete data about identity, purity, and content for the substance itself. This is only possible when a primary reference substance is available for calibration. In comparison to the field of measuring physical dimensions, one ounce (avoirdupois) can only be defined as 0.0283 of a kilogram when the mass of a kilogram is itself defined. Secondary reference substances are often called working standards—a term that indicates their importance in “everyday” laboratory work.

The USP and the EP offer primary reference substances for use with methods specified in their respective monographs. From a regulatory point of view, despite their high purity, these pharmacopeial standards are not valid for analytical purposes, other than their use in methods described in the respective monographs.

Finally, any other chemical substance with known structure can be used as a standard for identity testing, although its content may not be defined. For example, in fingerprint analysis by thin layer chromatography (TLC) the comparison of a preparation with marker substances shows whether or not the desired compounds are present.

Defining a Chemical Reference Substance

At this point it is necessary to discuss the determination of the content of a reference substance. The knowledge about how the content of a given reference substance is established is an essential prerequisite for its appropriate use in quantitative analyses. Unfortunately, many users of reference standards are not aware of the danger of using a reference substance—the content of which is not clearly defined—for the quantification of a component.

The first step in defining a chemical reference substance is always its identification. Today, nuclear magnetic resonance spectroscopy (NMR) and mass spectrometry (MS) are the most important tools for structure elucidation (the determination of the actual three-dimensional chemical structure of a specific molecule). The second step is to examine its purity. Again, this requires a combination of different methods. High performance liquid chromatography (HPLC) or gas chromatography (GC) are typical techniques used to search for organic impurities, as is thin layer chromatography (TLC). The content of water and residual solvents needs to be established, and elemental analysis would reveal whether or not there are inorganic impurities present that have to be considered. In particular cases, purity tests may include other methods, such as optical rotation or melting point. Then, as the third step, the content is typically determined using HPLC or GC methods. However, the reference substance user should be aware that a test for content based on a single HPLC or GC method is not sufficient because these analytical methods do not provide for an absolute content value, i.e., detect all possible organic impurities. To overcome this problem it is widely accepted to establish the content based on at least two independent methods to ensure that the assigned value is the “true” one. Until the total amount of impurities is subtracted from this value, it is not possible to speak about the defined content of a reference standard (e.g., for primary reference substances).

Because of these complex, time-consuming measurements, “the production, validation, maintenance and distribution of chemical reference substances is a costly and time-consuming undertaking.”² For the user of reference substances this statement simply means that reference standards, especially verified reference standards, are expensive materials and it is of high importance to optimize their use.

In most day-to-day lab work, the ideal procedure for quantification of a single compound in a mixture—comparison with a verified reference standard—is impractical due to the high costs of the standards. Therefore, the most common procedure is to buy a small amount of a commercially available primary reference substance, or generate one’s own primary substance, and to calibrate the content of cheaper bulk batches (i.e., less pure) of the same substance by comparison with the primary substance, thus producing secondary or working standards. This process provides a defined content for the working standard. The approach saves money because the amount of primary reference substance needed for comparison is lower and because only a single analytical method is generally required to assign the content value to the secondary standard.

But some pitfalls must be considered carefully. Usually, reference substances are used over a long period of time; for example, in routine analysis or stability testing. But the content of a reference substance depends on variable factors:

• Decomposition may occur, especially if the substance is not properly protected against oxygen and/or light.

• The water content and the content of other residual solvents may change if the reference substance is not stored in an airtight vessel.

• Consequently, storage of reference substances under defined conditions in a cool, dry, dark place is essential, but it does not eliminate the need to retest regularly.

However, sooner or later a batch will be completely depleted. In view of this inevitability, every new batch of a reference substance should be tested by direct comparison with the preceding batch. Further, it is important to order new reference material before the previous batch is exhausted. In the authors’ opinion, this is the only way to provide comparability and traceability of content values in the long-term.

Most often, users are left to discover all these requirements concerning the methodically correct use of reference substances on their own. In the authors’ opinions, it should be possible for users to choose whether they want to perform all the procedures described above on their own, or whether they will simply purchase a working standard from a reputable supplier who offers long-term traceability of the calculated content values.

A supplier of reference substances could offer (1) “primary” or verified reference substances with a defined content and complete documentation available, and (2) ready-to-use working standards for which the calibration against the “primary” substance is clearly and traceably documented.

By providing such documentation the requirements regarding (1) stability testing, (2) examination for impurities, (3) definition of storage conditions, and (4) documentation of all these procedures are minimized for users. Consequently, users finally get what they really need: reference substances, characterized through a defined content, at reasonable cost, and traceable to a basic “primary” value for a sufficient period of time. In this way, manufacturers of herbal products can help to ensure that the analytical testing—which will probably be required for many herbal products under new GMP regulations—will be based on high-quality reference substances. In the opinion of these authors, this will ultimately result in the production of more herb products of reliable quality.

Klaus Reif has a doctorate in Chemistry. A widely acknowledged expert in the field of analytical chemistry, he is the Head of the Department of Instrumental Analysis at PhytoLab in Germany, a large service laboratory for herbal products and a leading European supplier of chemical reference standards for botanicals. A member of the AOAC´s Pool of Experts, he belongs to several Expert Review Panels of the AOAC for Dietary Supplements. He has publications on various analytical subjects, e.g., in the Journal of Chromatography.

Hartwig Sievers has a doctorate in Pharmacy (Pharmacognosy) with extensive experience in regulatory affairs and quality control of herbal medicinal products (HMPs). He is the Head of the Department of Pharmaceutical Services and Research and Development at PhytoLab. His present focus is the safety and efficacy, pharmacovigilance, and development of new HMPs. He has publications in Phytochemistry, Clinical Pharmacology and Therapeutics, and Herba Polonica.

Jens-Peter Steffen has a doctorate in Chemistry specializing in synthetic organic chemistry and natural products research. He is Technical Manager of the PhytoLab Reference Standards Department. He has published papers on natural products chemistry in Phytochemistry and the Journal of Agricultural and Food Chemistry.

References:

1. Current Good Manufacturing Practice in Manufacturing, Packing or Holding Dietary Ingredients and Dietary Supplements: Proposed Rule. Federal Register Vol. 68, No. 49, Docket No. 96N-0417. Washington, DC: Food and Drug Administration. March 13, 2003;12158-12263.

2. General guidelines for the establishment, maintenance and distribution of chemical reference substances, WHO Technical Report Series, No. 885, 1999 (Annex 3). Available at http://www.who.int/medicines/organization/qsm/activities/qualityassurance/pharmacopea/who-trs-885annex3.html.

3. Note for Guidance on quality of herbal medicinal products, Committee for Proprietary Medicinal Products (EU) /QWP/2819/00. Available at http://www.emea.eu.int/index/indexh1.htm (Homepage of the EMEA Herbal Medicinal Products Working Party [HMPWP]).

4. Note for Guidance on specifications: test procedures and acceptance criteria for herbal drugs, herbal drug preparations and herbal medicinal products, Committee for Proprietary Medicinal Products (EU) /QWP/2820/00. Available at http://www.emea.eu.int/index/indexh1.htm (Homepage of the EMEA Herbal Medicinal Products Working Party [HMPWP]).

5. U.S. Department of Health and Human Services Food and Drug Administration Center for Drug Evaluation and Research (CDER), Guidance for Industry: Botanical Drug Products. Jun, 2004. Available at http://www.fda.gov/cder/guidance/index.htm.

Reference Standards

Reference standards are highly purified compounds used in the testing of various materials, whether they are herbal or conventional drugs. A small amount of a reference standard is used as the basis for the detection of the presence and quantity of the same compound in the material being tested. Examples of reference standards include specific ginsenosides (e.g., Rg1, Rb1) from Asian ginseng (Panax ginseng C.A. Meyer, Araliaceae) or American ginseng (P. quinquefolius L.), the terpenes ginkgolide A, B and C or bilobalide in ginkgo (Ginkgo biloba L., Ginkgoaceae), hypericin or hyperforin in St. John’s wort (Hypericum perforatum L., Clusiaceae), etc. Theoretically, pure reference standards can be developed by extraction and isolation in an herb company’s own laboratory, but the technology and the costs to do this may be prohibitively high. Most analytical laboratories obtain reference standards from (1) chemical supply companies that specialize in the sales of laboratory chemicals, including reference standards, reagents, and other specialized chemicals or (2) directly from the few companies that specialize in the production and sale of reference standards for herbal products. The purity of these compounds can vary unless they are produced according to exacting chemical standards.