The Pharmaceutical Journal
Vol 269 No 7217 p459-460
28 September 2002

This article
Reprint
Photocopy

World Congress of Pharmacy and Pharmaceutical Sciences 2002 summary

The importance of standardisation techniques for herbal medicines

Standardisation techniques for herbal medicines was the theme of a symposium jointly organised by two FIP sections — the Medicinal and Aromatic Plants section and the Laboratories and Medicines Control Services section — on 4 September. Co-chairmen Professor Peter Houghton, King's College, London, and Dr Frans van der Vaart, Royal Netherlands Medicines Laboratory, stressed the importance of harmonised and validated methodology and unambiguous labelling for herbal medicinal products (HMPs).

Officially recognised standards

Dr Keith Helliwell, William Ransom, UK, referred to the categorisation of three types of extract according to correlation with their clinical activity. These are (1) herbs such as senna, where the characterised constituents of HMPs are solely responsible for the therapeutic and clinical effects with a dose related response to quantified constituents, (2) herbs such as St John's wort, where the characterised constituents are not solely responsible for the therapeutic and clinical effects, and (3) those other herbal materials, such as valerian, where there were no characterised constituents responsible for the therapeutic and clinical effects.

He differentiated therapeutically active herbal materials, with examples, within five chemobotanical families: anthraquinone-based (eg, senna), a few solanaceous alkaloids, certain other alkaloids (such as cinchona and opium), tannin-containing plants, and a miscellaneous category that included digitalis and capsicum. For quantification, Dr Helliwell exemplified three alternative methodologies used in European Pharmacopoeia monographs: solvent/solvent extraction followed either by titration [as for Belladonna] or by spectrophotometry [cascara, cinchona] or using liquid chromatography [capsicum, opium]. He proposed five "ideals" for standardisation:

1. The quantification method should be applicable equally to initial herbal materials, to the primary extraction product and to the final dosage form

2. Methodology as simple as possible to achieve the required conditions

3. Realistic levels and limits of quantification

4. Methods to be stability-indicating

5. The results of the quantification should reflect the therapeutic effect

He illustrated the applicability of his five ideals to plant sources, and extract and dosage form, of four typical families of herbal material. For ipecacuanha, the solvent method for the various herbal starting materials differs from PhEur and British Pharmacopoeia monographs for the liquid extract and the tincture, and from that used by industry for the final dosage form. The method was suited to realistic levels, and was stability indicating, but the reproducible quantitation does not necessarily match therapeutic activity because alkaloids are measured in total and there could be a significant difference from batch to batch in the ratio of the two main alkaloids, emetine and cephaeline.

For cascara, the official assay, which comprises extraction/hydrolysis/extraction followed by colour development and spectrophotometric measurement, is tedious and needs careful training; however, it eliminates most congeners and there is a reasonable correlation with therapeutic activity.

He instanced three distinct approaches to the assay of capsicum fruit: (i) total pungent constituents (capsaicin, dihydrocapsaicin and nordihydrocapsaicin), (ii) total capsaicinoids by UV spectrophotometry (cf British Pharmaceutical Codex), and (iii) individual capsaicinoids by HPLC, as in the PhEur monograph on capsicum. These methods were reasonably stability-indicating and correlated with therapeutic effect. Explaining that capsicum fruit normally contains less than 5 per cent nonanoyl vanillylamide, he said dosage forms with much larger amounts (up to 50 per cent) of this synthetic capsaicin point to adulteration.

His fourth example was opium, which reflected a "a saga of more than 30 years". Historically, there has been diversity of assay methods for different dosage forms. Since 1993, the classic PhEur/BP gravimetric assay for medicinal opium has been superseded by HPLC, whereas for various formulations (opium tincture BP, camphorated tincture BP and squill linctus opiate BP [Gee's linctus]) that assay has been replaced by the Radulescu (nitrite/ammonia) reaction. The current PhEur method is generally applicable to all forms, has simple extraction and chromatography, reflects realistic limits of quantification, is stability indicating and provides a reasonable correlation with the therapeutic effect.

In conclusion, Dr Helliwell emphasised that:

• Herbal materials have been used for centuries for their therapeutic efficacy

• Accurately controlled dosage is important to prevent discomfort or even death

• Many current methods of standardisation were developed 50 or more years ago, and in many cases they are still the most applicable

• Modern chromatographic methods are not suitable for herbal materials where the therapeutic activity derives from a complex mixture of many closely related compounds, which could still be quantified by traditional methods

Herbal extracts in the European Pharmacopoeia

Professor Gerhard Franz, University of Regensburg, Germany, and chairman of PhEur group 13B, said that "perhaps 75 per cent of the HMPs in Western Europe contain extracts, mainly in the dry form", but this "industrial reality is not completely reflected in the PhEur, which contains more than 130 monographs on herbal drugs but only eight examples of quality defined extracts and tinctures". As a consequence, the general monograph on extracts and tinctures was recently revised as a basic framework monograph and as a future guideline for the establishment of individual extract- and tincture monographs.

There is usually a lack information on active pharmaceutical ingredient (API) and relatively few pharmacokinetic studies but there are compendial controls on characteristic marker substances, inert materials, allergens and toxins, cellulose and lignin. Dr Franz recalled that problems largely flowed from the huge variety of HMPs on the European market, containing many different types of extracts, produced by several methods, and using a broad spectrum of solvent systems. Generally, the production of different types of extracts is outlined and specified according to the obvious needs in industrial practice, eg, great variation in batches of ginkgo, but it is possible to consolidate in an average standardised product. The PhEur "Production statement" concept controlled a "dry extraction ratio" (DER) reflecting polarity, concentration and volume of solvent, manufacturing process, time, pressure and temperature. An example is chamomile maceration, where choice of solvent varies the API found. Such "statements" would also relate to extraction water quality (eg, potable water for extraction).

In conclusion, Professor Franz observed that the increasing development of new and more effective synthetic drugs has not been reflected in HMPs; nevertheless HMPs are being better defined, and incorporation of similar concepts of safety, quality and efficacy are included in monographs acceptable to regulators.

Absence of known active components

Dr Ezio Bombardelli, president of Indena Spa, Italy, dealt with the controversial situation of acceptable standardisation of herbals with no known active components. He agreed that the main problem has been the enormous variety of plant products in several countries. He listed the top ranked 47 herbs in tonnage quantities in France, and in Germany, where more than 200 plants are used. Relatively few extracts have modern documentation in terms of safety, efficacy and chemical standardisation comparable with the registration requirements for any chemical drug. For plants with no known active components, the plant material must be identified and properly characterised, with well chosen markers belonging to at least two different classes of chemical substances, and analytical methods thoroughly validated with respect to a defined component mixture. He proposed that instrumental methods, for instance, nuclear magnetic resonance and Fourier transform-infrared spectrometry, should be used to provide a generally recognisable fingerprint.

The ratio between originating plant material and extract could be hugely variable. Moreover, for HMPs in popular demand, there are limited natural sources world-wide, sometimes with ecological restriction on harvesting, and time is needed for agronomic development to large-scale cultivation. He gave several examples of variable composition found in practice: there is a wide range of fatty acids and alcohols in extracts from Serena repens (used for benign prostatic hyperplasia); there is variable concentration of caffeic acids and amides in Echinacea spp; and Hypericum perforatum extracts include many inactive compounds, such as pseudohypericin and flavanoids, with range of hypericin wider in "wild" crops (0.053–0.3 per cent) but around 0.15 per cent in cultivated varieties from four countries. He concluded: "A perfect standardisation is essential to obtain biologically reproducible data in terms of safety and efficacy and the preparation must be consistent over time, stable and be devoid of unpredictable toxicity due to degradation or poor quality, or selection, of the plant material."

Quality control of isoflavones in soya

Professor Maria da Graηa Campos, University of Coimbra, Portugal, referred to the current interest in the role of isoflavonoids as phyto-oestrogens, particularly soya products also present in infant formula products. These are of interest in hormone replacement therapy in women and for treatment of prostatic cancer in men. Genistein is less effective for hot flush relief but no side effects have been reported other than allergenic response in asthma patients to dust from extract of soya protein.

She emphasised the need for rapid, sensitive and precise assays to analyse the compounds involved in pharmaceutical formulations, mostly tablets, with a minimum of manipulation, and to compare the results with those from extracts prepared from soya seeds. She advised that processing of the extracts may alter the distribution of the forms and can result in the loss of some isoflavones through leaching and through removal of undesirable fractions. To avoid this problem, the isoflavones were analysed using reversed-phase HPLC, with diode array detection, under gradient conditions, and without prior separation from the complementary bioactive compounds present in the tablets. She had confirmed by liquid chromatography/mass spectrometry that those bioactive impurities were eliminated in the first step of the HPLC gradient without producing any interference in the analysis. Coloured components were removed with a C18 pre-column and the isoflavone markers were stable.

The results that her group obtained with a series of tablets available on the Portuguese market suggest that there is no well-established dose/activity relationship, because the amount of isoflavones recommended is different for each product, with complex HPLC profiles exhibited by different herb species. She proposed to follow up this work with a more detailed project, involving wider collaboration — including the Lisbon national laboratory — with the objective of studying isoflavonoids obtained from soya seeds from different botanical sources, and also assessing these compounds for their pharmacological and toxicological effects in laboratory rats. Chemical studies would be performed to analyse isoflavonoid compounds from tablets and extracts from soya seeds. In parallel, pharmacological and toxicological studies, using rats, would evaluate the effects on oestrogen receptors, and the toxicity effects of these compounds in relevant tissues.

Preparations on the Belgian market

Dr Jozef Corthout, Belgian Medicines Control Laboratory, Brussels, described some problems they had encountered in evaluating hypericum. His laboratory undertook post-market quality control of branded drugs sold in public pharmacies, including herbal drug preparations. A general problem was the different ways that a plant could be used — harvested from a different part of the plant, or as crude drug , or different kind of extracts, not necessarily in a standardised preparation or normalised to a certain quantity of active substances or markers. A further problem was that the product could be standardised with respect to a single constituent (active or not), or to a group of metabolites — for example, hydroxyanthracene glycosides, or total cascarosides, or specifically cascaroside A.

He preferred a colorimetric assay for analysing a group of related compounds, whereas the determination of a single constituent required the separative power of a chromatographic method such as TLC, GC or HPLC, with various options for detectors. Their analytical method was then validated for linearity over a 70–130 per cent range, for precision as repeatability of 18 replicates; and for accuracy by determination of recovery of a spiked analyte with 70, 100 and 130 per cent of the nominal amount. There were also problems with labelling as to content, the form used (such as salt, ester or ether), whether they were using a drug or a crude extract, and whether it was "normalised" to a fixed content.

Dr Corthout then discussed specific problems encountered when examining eight preparations of St John's wort (two tablets, three coated tablets and three capsules), all claiming a certain content of hypericins. Using PhEur definitions, and standard methods where available, his group verified a series of parameters and confirmed compliance with uniformity of mass, identification and microbial quality for all the products. However, for one capsule preparation, the disintegration test did not comply with the PhEur and the total hypericin content varied according to assay method — 13–85 per cent of the label claim by HPLC or 27–176 per cent by colorimetry. He commented that these results showed the importance of a detailed and accurate label description combined with a defined assay method.

Dr Corthout concluded that quality control of HMPs is much more complex than for synthetic substance analysis; it is often uncertain what extract has been used in the preparation, or which assay has been relied upon. Label claims can be misleading.
—Contributed by Professor Geoffrey Phillips, a former secretary of FIP section for Laboratories & Medicines Control Services.