Vol. 3, 85-98, January/February 1994 Cancer Epidemiology, Biomarkers & Prevention 85 Review Chemopreventive Drug Development: Perspectives and Progress Gary J. Kelloff,1 Charles W. Boone, James A. Crowell, many cancers. While prevention ofexposune and changes in Vernon E. Steele, Ronald Lubet, and Caroline C. Sigman diet may someday alter incidence, chemical intervention Chemoprevention Investigational Studies Branch, Division of Cancer offers an attractive approach with potential for more imme- Prevention and Control, National Cancer Institute, Bethesda, Maryland diate results. The NCI’s2 chernoprevention drug develop- [G.J.K., C.W.B., ).A.C., yES., R.L.[; and CCS Associates, ment program, which has been described previously (1-3), Palo Alto, California EC.C.S.[ has as a goal the identification of safe and effective chemical agents for the prevention of human cancers. This program is an applied drug development science effort with clinical Abstract trials as the endpoint. It begins with the identification of Chemoprevention drug development has the goal of candidate agents for development and the characterization identifying safe and effedive chemopreventive agents for of these candidates for efficacy using in vitro and animal clinical use. Several distinctive strategies are pursued in screens. Promising agents are then further tested in animal developing chemopreventive agents: (a) identifying models to evaluate the design of regimens for clinical testing and validating predysplastic and early dysplastic lesions and use. Agents judged to have potential as human cherno- that can be used instead of cancers as endpoints for preventives are subjected, as appropriate, to preclinical tox- measuring chemopreventive adivity; (b) identifying and icity and phanmacokinetic studies. The most successful testing candidate agents based on considerations of agents then progress to clinical trials. The purpose of this mechanisms of adion; (c) evaluating combinations of article is to discuss the strategies, perspectives, and progress agents with potential for maximizing efficacy and of chemoprevention drug development, often using ex- minimizing toxicity; and (d) applying a systematic amples from the NCI program. As evidence of progress, the methodology for identifying and ranking candidate status and rationale for development of agents that currently agents at each stage of development to ensure discovery appear to be most promising (Table 1 ) are reviewed. of the best agents and most effedive use of available resources. This article discusses 22 drugs and three drug Strategies for Chemopreventive Drug Development combinations which have reached an advanced stage of Epithelial Lesions that Are Targets for Chemopreventive development as chemopreventive agents. The first Agents. The rational design and development of chernopre- generation of drugs are the most advanced, now being ventive agents requires a clear understanding of the epithe- in Phase II and Phase Ill clinical trials. These drugs hal lesions that are targets for the action ofthese agents. Fig. include several retinoids [vitamin A, 1 3-cis-retinoic acid, i diagrams the early development of epithelial neoplasia. A all-trans-N-(4-hydroxyphenyl)retinamidel, calcium, major target epithelial lesion is that of histologically visible j3-carotene, tamoxifen, and finasteride. The second neoplasia very early in its development, long before it in- generation drugs are those in Phase I clinical trials. From vades across the basement membrane. This “intraepithelial most to least advanced, these drugs are neoplasia,” as it is called at this stage, begins as a mono- 2-difluoromethylornithine, sulindac, piroxicam, oltipraz, clonal focus near the basement membrane and expands up- N-acetyl-I-cysteine, aspirin, ibuprofen, carbenoxolone, wand and laterally. When it finally becomes invasive across 1 83-glycyrrhetinic acid, and the combination of the basement membrane, it is at this point termed “cancer.” 2-difluoromethylornithine with piroxicam. The third Prior to invasion, the morphological changes of intnaepi- generation includes agents with significant evidence of thelial neoplasia are collectively termed “dysplasia.” It is the chemopreventive activity in animal models. These agents consensus of pathologists that carcinoma in situ and severe are now in preclinical toxicity testing. They are dysplasia form an indistinguishable continuum (4). The S-allyl-I-cysteine, phenhexyl isothiocyanate, curcumin, other major target lesion for chernopreventive agents may be ellagic acid, fumaric acid, fluasterone, and the termed “predysplasia,” which is the stage of neoplastic de- combinations of all-trans-N-(4-hydroxyphenyl)retinamide velopment after initiating DNA mutational changes have oc- with oltipraz and all-trans-N-(4-hydroxyphenyl) cunned but before the onset ofdysplasia, when the tissues still retinamide with tamoxifen. Introduction As the understanding ofthe process ofcancer increases, pre- 2 The abbreviations used are: NCI, National Cancer Institute; AFB,, aflatoxin ventive intervention is becoming scientifically practical for B,; B(a)P, benzo(a)-pyrene; DFMO, 2-difluoromethylomnithine; 1)HEA, de- hydroepiandrosterone; DMBA, 7, 1 2-dimethylhenz)a)anthracene; DMH, 1 ,2-dimethylhydrazine; G6PDH, glucose-6-phosphate dehydmogenase; GSH, glutathione; 4-HPR, all-trans-N-)4-hydroxyphenyl)retinamkle; MN U, N-methyl-N’-nitrosourea; MTD, maximum tolerated dose; NNK, 4-)methyl- Received 4/5/93; revised 7/6/93; accepted 7/i 6/93. nitrosamino)-i -)3-pyridyl)-1 -butanone; NOEL, no observed eftect level; I To whom requests for reprints should be addressed, at Chemoprevention NSAID, nonsteroidal anti inflammatory drug; ODC, omnithine dcc amboxylasc’; Investigational Drug Unit, National Cancer Institute, Executive Plaza North, PAH, polycyclic aromatic hydrocarbon; TPA, 1 2-O-tetradecanoylphorbol- Suite 201, 9000 Rockville Pike, Bethesda, Maryland 20892. 1 3-acetate. Downloaded from cebp.aacrjournals.org on September 29, 2021. © 1994 American Association for Cancer Research. 86 Review: Chemopreventive Drug Development Table 1 Cancer chemopr eventive age nts unde r development have (e.g., the netinoids) confounds the discovery ofthe most important mechanisms. Nevertheless, the known phanma- Preclinical Clinical trials cological properties of the agents being evaluated and the toxicology Phase I Phase II Phase III experimental testing data on various classes of agents pro- vide very useful insights into mechanism that may lead to the First generation Retinoids development of more effective chemopreventive drugs. In Vitamin A” b (3)c + (2) some cases, the pharmacological activity of a compound 13-cis-metinoic acid” + + (2) + (2) suggests very specifically the target tissues and cancers 4-HPR” + + + + against which a chemopreventive drug may be active. For + + (3) + (2) Calcium example, an antiestrogen such as tamoxifen would be de- n-Carotene + + (6) + veloped for use against estrogen-sensitive cancers such as Tamoxifen” + + + Finastenide” + + those of the breast. In other cases, the testing of classes of compounds known to have a very general chemopreventive Second generation” activity may lead to the discovery oftissue specificities, such DFMO + + +(2) as those of antiinflammatonies in colon and bladder. Sulindac” + + (2) Table 2 presents a working classification of the che- Pinoxicam + + mopreventive agents listed in Table 1 arranged according to Oltipnaz + + structure or pharmacological effects associated with che- N-acetyl-l-cysteine + + Aspirin + + mopreventive activity. Note that many ofthe agents fall into Ibuprofen + + monethan oneclass. Moreover, the Iistofclasses is undoubt- Carbenoxolone + + edly incomplete. 18f3-Glycymrhetinic Acid + + In Table 2, the specific chemopreventive activities and DFMO + Pinoxicam + + structures are grouped into three general classes. The firstis inhibitors of cellular proliferation (antiprolifematives), mani- Third generation” S-Allyl-l-cysteine + + fested in such specific mechanisms as ODC inhibition, pro- Phenhexyl Isothiocyanate + tein kinase C inhibition, and antiestrogenic activity. Anti- Curcumin + proliferatives i nclude netinoids, polyphenols, anti hormones, Ellagic acid + calcium, DFMO, and the DHEA analogue, fluastenone. A Fumanic acid + second general class is carcinogen blocking agents. Block- Fluasterone + ing is often produced by enhancing the carcinogen- 4-HPR + Oltipraz + detoxifying enzymes, especially the Phase II metabolic en- 4-HPR + Tamoxifen + zymes, including GSH S-tnansferases, which perform a Previous development by the pharmaceutical industry. conjugation and other reactions. Wattenbeng and Talalay b Testing is completed or in progress. (6-8) have emphasized the desirability of selecting chemo- ‘ For Phase II and III studies, numbers in parentheses, number of trials. (I Second- and third-generation agents are listed in order from most to least preventive agents which induce mostly Phase II metabolic advanced in development. enzymes as opposed to compounds which induce both Phase I mixed function oxidases and Phase II enzymes. In- duction of mixed function oxidases carries the potential of activating procancinogens. appear rnorphologically normal. The natural history of in- A third general class of inhibitors is antioxidants, tnaepithelial neoplasia in the major human epithelia, with such as S-allyl-/-cysteine, curcumin, N-acetyl-i-cysteine, implications for chernopreventive strategy, has been me- NSAIDs, and polyphenols. These agents trap electrophilic