PLANTS IN CANCER TREATMENT

Plant materials have been used in the treatment of malignant diseases for centuries; a comprehensive survey of the literature describing plants used against cancer listed over 1400 genera (Hartwell, 1967–71, see Lloydia, 1971, 34, 427 for index). Recent phytochemical examination of plants which have a suitable history of use in folklore for the treatment of cancer has indeed often resulted in the isolation of principles with antitumour activity. Podophyllum was used over 2000 years ago by the ancient Chinese as an antitumour drug, and resins from the root of the plant Podophyllum hexandrum (syn. P. emodi) and the related American species, the May-apple (P. peltatum) have yielded a number of lignans and their glycosides having antitumour activity. Although the major constituents from these two species, podophyllotoxin and the peltatins, are unsuitable for systemic drug use, two semi-synthetic derivatives of podophyllotoxin, etoposide and teniposide, gave particularly good results in clinical trials. Etoposide is currently available for the treatment of small-cell lung cancer and testicular cancer, and teniposide is used in paediatric cancers, though both compounds have a similar anticancer spectrum. Other podophyllotoxin-related analogues have been developed and tested. Podophyllotoxin itself may be used topically, and is most effective in the treatment of venereal warts. From the time of Galen (about AD 180), the juice expressed from woody nightshade (Solanum dulcamara) has been used to treat cancers, tumours and warts, and references to its use have appeared in the literature of many countries. The active tumour-inhibitory principle has been identified as the steroidal alkaloid glycoside β-solamarine. Various lichens, e.g. species of Cladonia, Cetraria and Usnea, also have a history of use in folk medicine against cancer since about AD 970. These are all rich sources of usnic acid, a compound which has been recognized for many years as an antibacterial and antifungal agent, but only more recently as an antitumour compound. Similarly, many centuries ago, the druids claimed that mistletoe (Viscum album) could be used to cure cancer; protein fractions with marked antitumour activity have been isolated from mistletoe extract. Mezereon (Daphne mezereum), despite its toxic properties, has also been used in many countries for the treatment of cancer. The active antitumour constituent of this plant has been identified as a diterpene derivative mezerein, which is structurally very similar to the toxic principle daphnetoxin.

Very successful higher plant materials used in cancer chemotherapy are the alkaloids of Catharanthus roseus. Research on this plant, the Madagascan periwinkle, was stimulated by its mention in folklore,not as a cure for cancer, but in the treatment of diabetes. No hypoglycaemic activity was detected, but treated test animals became susceptible to bacterial infection, and this led the researchers to undertake extensive examination for possible immunosuppressive principles causing these effects. A number of bisindole alkaloids showing antileukaemic activity have subsequently been isolated and two of these, vincaleukoblastine (vinblastine) and leurocristine (vincristine), are now extracted commercially from Catharanthus roseus and used, either alone, or in combination with other forms of therapy for cancer treatment. Another important, more recent, addition to the list of anticancer drugs is paclitaxel (Taxol), a diterpene derivative isolated initially from the bark of the Pacific Yew, Taxus brevifolia. Although reportedly used by Native North Americans for various conditions, it does not appear to have had any traditional cancer usage and was obtained as part of the random collection programme of the US National Cancer Institute (NCI). Taxol and related taxanes are treated below.

METHODS OF INVESTIGATION

An intensive survey of plants, microorganisms and marine animals (starfish, corals, etc.) for antitumour activity began in the late 1950s, mainly because the United States National Cancer Institute (NCI) instigated and funded a major screening programme. A random- selection screening programme was adopted, since novel compounds may be found anywhere in the plant or animal kingdom, and it is known that some natural products are restricted to a single genus, or even species. Random mass-screening is naturally an expensive operation, and probably only justified in certain areas, where our present range of drugs is seriously inadequate or inefficient. Cancer was considered to be in this category.

Since the beginning of the programme, which continued until 1983, a vast number of extracts from various sources has been tested for antitumour activity. About 4% of the extracts tested have shown reproducible activity. Over about 25 years, some 114 000 plant samples representing 40 000 species were tested. Different parts of a plant—seeds, leaves, roots, etc.—were separately examined wherever possible. It is estimated that between a quarter and half a million plant species exist worldwide, and thus the plant kingdom still represented a vast untapped source of material.

The isolation of biologically active constituents, probably minor constituents, from a crude plant extract involves techniques differing from those of conventional phytochemical evaluation. With these, it was customary to study those chemicals which were most easily separated from a plant extract; these were usually those present in the largest quantities and which crystallized readily, or those which represented the researcher’s field of interest, e.g. alkaloids, terpenoids, phenols, etc. Only after characterization of their structures were such compounds subjected to biological testing, e.g. for hypotensive, antibacterial, anticancer activities, etc., and this would depend on sufficient material being available. Countless medicinally useful compounds have been missed in this type of approach.

In the more recent systematic studies for useful plant constituents, every portion of the plant and every fraction of the extract is tested biologically before any constituent is isolated and characterized. Usually only those fractions showing biological activity are studied further. Thus, one may isolate almost any class of compound as an active constituent, and it may not be one traditionally associated with a particular plant family. Even procedures involving continuous monitoring of fractions for biological activity are not free from anomalies. It is quite well known that isolated constituents of a plant drug may not give the same clinical response as a crude preparation of that plant drug. Very often, the total therapeutic activity is greater than, or different from the therapeutic activities of the individuals. Synergism or antagonism (see Chapter 7) due to the complex nature of the extract are probably the causes of such observations. It is thus possible that a fraction from a plant extract, although showing significant biological activity, possesses no single constituent with this activity. Conversely, a fraction showing no activity may still contain an active constituent. A further complication is that crude fractions may contain additional substances with delayed toxicity, causing test animals to die at about the same time as control animals.

An effective screening procedure in which a large number of crude plant extracts are to be assayed for biological activity should fulfil several criteria. It should be sufficiently selective to limit the number of leads for follow-up evaluation yet it should be highly sensitive in order to detect low concentrations of active compounds, and it should be specific so that the assay is not affected by a wide variety of inactive compounds. The preliminary screens employed in the NCI studies changed during the lifetime of the programme to avoid detecting weakly active common plant products such as polyphenols, tannins, saponins and sterols not capable of being developed into useful drugs; such dereplication (Chapters 8, 9) is now commonly employed. The routine testing of extract fractions for antitumour activity is frequently done via an in vitro cytotoxity assay, although in vitro cytotoxicity is not always an effective or reliable means of predicting in vivo antitumour activity. However, since the in vitro cytotoxicity bioassay is rapid and inexpensive, and only small amounts of extract are necessary, it is the popular method for initial tests.

Promising chemicals are subsequently tested against a range of standard experimental neoplasms, and then considered for preclinical toxicological studies if these results are sufficiently encouraging. At this stage, relatively large amounts of material will be required, and larger-scale extractions and fractionation may be necessary. Very few compounds will reach clinical trials. A low or very narrow therapeutic index (the ratio of maximum tolerated dose to minimum effective dose), undesirable side-effects or high toxicity can outweigh beneficial tumour-inhibitory activity. From some 25 000 screens conducted annually by the NCI (including both synthetic and natural materials), only eight to twelve compounds are likely to be selected for preclinical testing, and perhaps only six to eight go on to clinical trials. Slightly less than half of these may be plant-derived.

Should a plant-derived natural product or derivative be considered worthy of development as a drug, the availability of future supplies of the plant becomes critical. Collections from the wild may be exploited if the plant is common, but mass cultivation should be considered. With slow-growing crops, e.g. trees, this could mean a considerable delay before significant supplies are available. Alternative plants might be richer sources of the compound, or be more accessible; other species of the same genus or closely related genera from the same family should also be analysed. Thus, wild sources could not supply the huge amount of Taxus brevifolia bark needed to satisfy demand for the manufacture of taxol but the discovery of baccatin III and deacetylbaccatin III (readily convertible to taxol) in the leaves of the common yew (T. baccata) has ensured the future supply of the drug. For commercial exploitation, agreements giving some slice of any profit arising from the sale of a product derived from wild plant material must be arranged with the country of origin.

Plant tissue cultures might provide a reliable source. If total synthesis of the active chemical is feasible, this will always be the preferred option. However, the extremely complex structures of most bioactive natural products frequently preclude satisfactory commercial syntheses.

Although the random-selection screening programme for natural products was terminated by the NCI in 1983, the number of cytotoxic and antitumour agents identified was enormous, and these have increased our understanding of the cancer process, and the mechanisms of action of the agents. Synthetic work has enabled structure–activity studies to be undertaken, and there is every hope that synthetic or semi-synthetic analogues may, in time, be developed and become useful drugs.

Over the years, separation techniques (TLC, HPLC, chiral chromatography, etc.) and chemical structure determination (MS, NMR, UV, X-ray crystallography) have reached high levels of sophistication. For the examination of large numbers of samples, high-throughput screening is now employed (Chapter 9).

Following on from the above, partly as a result of new screening techniques, the NCI in 1986 revived its collection of plants, focusing on tropical and subtropical species which had local medicinal use.

Considerable recent research has been, and still is, carried out by National Cooperative Drug Discovery Groups (NCDDGs); these groups result from cooperative agreement awards funded essentially by the NCI, which supports all aspects of preclinical anticancer drug discovery and treatment strategies. In 2003, there were 13 funded groups of which five were natural-product based. Such groups may involve the participation world-wide, of universities, research centres and industry, together with the resource countries of plant or marine materials. The complex nature of such groups is evident from the number of authors involved in ensuing publications. For updated reports from a number of such groups, see ‘Further reading’. For an article giving explanatory details of the NCDDGs project, see Y. F. Hallock and G. M. Cragg, Pharm. Biol., 2003, 41(supplement), 78.

Another multinational group involving collaboration of seven S. American countries and supported by the Organization of American States has prioritized 314 Latin American plant species for screening for cytotoxic properties. Results for some 70 species from 40 families tested against breast, lung and central nervous system human cancer cell lines have been reported (A. I. Calderon et al. (11 authors), Pharm. Biol., 2006, 44, 130).

For cytotoxicity testing, human cancer cell lines obtained from the NCI are frequently employed. Plant selection for screening can be assisted by initially choosing those from an ethnomedical database and submitting them to a search for biological and chemical information in the libraries held by NAPRALERT (see Chapter 2), industry or other research centres.

Research groups in which there is a commercial interest should establish agreements with the countries of origin of the plant material giving the latter a fair share of any profits that might arise from the marketing of any successful products. This follows from the 1992 UN Convention on Biological Diversity, which calls for recognition of the sovereign rights of countries to control the utilization of their natural resources and genetic materials. Further, some countries, e.g. the Philippines, have instituted their own regulations concerning bio-prospecting.

ESTABLISHED NATURAL PRODUCTS AS TUMOUR INHIBITORS

The tumour-inhibitory principles isolated in screening tests are usually new natural products, spanning a wide range of structural types. Examples of these, subdivided into phytochemical groups, are listed in Table 27.1, and structures of the more promising chemicals which were subsequently evaluated in clinical tests are shown in Fig. 27.1. However, a number of the compounds isolated were, in fact, previously known natural products. These were presumably compounds which had not been subjected to rigorous testing for biological, particularly antitumour, activity. Amongst these are usnic acid, ellagic acid, the anthraquinone aloe-emodin, the quinones juglone and lapachol, pyrrolizidine alkaloids retronecine and monocrotaline, the nitrophenanthrene aristolochic acid (see ‘Serpentary’), the bufadienolide hellebrigenin acetate, the Colchicum alkaloids colchicine, demecolcine and 3-demethylcolchicine (see ‘Colchicum’). Other known compounds that proved active but were rejected on terms of toxicity, low therapeutic index, etc. were the alkaloids of Senecio, indicine N-oxide from Heliotropium indicum and the cucurbitacins. Further examples established as tumour inhibitors at much later dates than the above were the alkaloids acronycine, ellipticine, emetine and nitidine. Acronycine, from Acronychia baueri, failed in phase 1 clinical trials. The pyridocarbazole alkaloids ellipticine and 9-methoxyellipticine from Ochrosia elliptica and other related plants, together with a number of synthetic analogues appeared useful but exhibited unacceptable side-effects. However, the quaternization of 9-hydroxyellipticine to give the water-soluble 9-hydroxy-2-N-methylellipticinium acetate (elliptinium acetate) (Fig. 27.2) has produced a highly active material, of value in some forms of breast cancer, and perhaps also in renal cell cancer. A variety of such quaternized derivatives is being tested, and some water-soluble N-glycosides also show high activity. Cephaelis ipecacuanha has been used for many years as an emetic and expectorant and the principal alkaloid, emetine, was shown to have antitumour properties. Clinical usefulness was marginal, however, and some toxic effects were noted. Nitidine is a benzophenanthridine alkaloid isolated from Zanthoxylum nitidum, and has more recently been obtained from screens of Fagara species (Rutaceae). Nitidine was selected for development based on its exceptional antileukaemic activity, but was dropped owing to erratic toxicity. The closely related alkaloid fagaronine is less toxic than nitidine. Similar benzophenanthridine derivatives are present in Chelidonium majus (Papaveraceae), a plant with substantial folklore history of use in the treatment of cancers. More recently, NK 109 (Fig. 27.2), a synthetic isomer of fagaridine found in Fagara xanthoxyloides, has been found to have greater antitumour activity than any of the natural benzophenanthridine structures, coupled with excellent stability, and was entered for clinical trials in Japan.

Table 27.1Some antitumour compounds from plants.