Publications and Presentations
CADRG Annuals Reports
2002 CADRG Annual Report
Introduction
The Division of Cancer Prevention supports intramural and extramural research in several broad areas, including chemoprevention, nutrition, biomarkers, biometry, basic science, and early detection. The focus of the Chemoprevention Agent Development Research Group (CADRG) is to identify specific chemical substances with potential chemopreventive activity in humans, and to support their development through Phase I clinical trials. Specifically, the goals of the CADRG research program are to (a) develop candidate agents sequentially through a series of standardized mechanistic, in vitro and animal efficacy tests and a battery of toxicological tests; (b) develop chemistry, manufacturing, and control dossiers and Investigational New Drug (IND) applications; and (c) conduct clinical Phase I pharmacokinetics and safety testing.
In support of these goals, CADRG funds a contract research program as well as a grant-based research portfolio. These programs allow interaction among extramural scientists and CADRG staff, and facilitate the review and implementation of ideas and research projects pertinent to chemoprevention. For a listing of current requests for proposals/applications, and program announcements, please refer to CAD funding opportunities.
More than 90 principal investigators at more than 60 academic and other research sites have participated in the preclinical and Phase I program. Since 1996 the activities of the program have resulted in more than 100 primary research publications; for a listing of journal articles published by CADRG staff in 2001-2002, please see Journal Articles. There are currently 45 negotiated collaborative research and development agreements with pharmaceutical and food industries. Seven new chemical entity INDs have been opened, and there are 32 active INDs (covering 28 agents). To date, 46 Phase I studies have been completed and the results published; 13 are ongoing. In support of clinical investigations, all the preclinical toxicology and pharmacokinetic studies and many of the preclinical efficacy studies sponsored by CADRG are submitted to FDA. Research undertaken to support agent development also includes the evaluation of drug effect markers and of preclinical models and surrogate endpoints of carcinogenesis. Additionally, the CADRG is implementing web-based initiatives to make technical results of completed studies and bibliographic information available to the research community; this will be linked to the NCI Resource Database.
Chemoprevention Research
Chemoprevention is a novel approach to cancer control that seeks to inhibit neoplastic development prior to or during the preneoplastic period through pharmacologic, nutritional, or endocrinologic intervention. As such, chemoprevention provides a useful complement to therapeutic modalities in current clinical use, and may be particularly useful in the control of cancer in tissues for which therapeutic intervention is relatively ineffective.
Studies conducted in experimental animal models for human cancer have demonstrated that carcinogenesis in a number of tissues is subject to inhibition through the administration of biological or chemical agents. Chemoprevention has been studied most extensively in animal models for cancer of the lung, prostate, skin (non-melanoma), mammary gland, urinary bladder, colon, and pancreas. Although less widely explored, effective chemoprevention has also been demonstrated in animal models for cancers of the oral cavity, skin (melanoma), ovary, and esophagus. Anticarcinogenic activity has been demonstrated for a highly diverse group of biological and chemical agents, including:
- retinoids (natural and synthetic analogs of vitamin A);
- plant phenolics (green and black tea, flavonoids, soy isoflavones, curcumin);
- hormone antagonists (tamoxifen, raloxifene and related antiestrogens);
- inhibitors of hormone synthesis (vorozole, finasteride)
- inhibitors of polyamine biosynthesis (difluoromethylornithine (DFMO));
- Phase 2 enzyme inducers (indole-3-carbinol, diindolylmethane, sulforaphan, garlic sulfides); and
- antiinflammatory drugs (aspirin, sulindac, R-flurbiprofen, COX-2 inhibitors, budesonide).
Back to Top
Chemoprevention Research Models: Molecular Targets
The 1998 DCP Chemoprevention Implementation Group encouraged further development and use of animal model systems that parallel human carcinogenesis for evaluation of preventive agents. In the past five years, the CADRG agent development program has increasingly employed animal and in vitro models that incorporate prominent molecular targets in cancer development. This effort has included the development and use of genetically engineered and other novel animal models for cancer prevention studies. For example, transgenic, knockout, and cross-bred mouse models, with or without carcinogen/viral induction, are being validated for preventive agent efficacy testing. Cell cultures from such animals and from humans with genetic predispositions to cancer are being evaluated by in vitro/in vivo pharmacological comparisons for preventive agent efficacy screening. Genomic and proteomic approaches are being incorporated into preventive agent development to identify agent mechanisms of action. Examples of progress in these areas are provided below.
Through participation in the NCI initiative Mouse Models of Human Cancer Consortium (MMHCC), CADRG has applied a number of relevant animal models in chemopreventive agent assessment. Models that have shown promising results in the agent development program include:
- Min mouse, and Min mouse with p53 lesions (susceptibility to desmoid tumors) or mice with DNA repair deficiencies;
- C3(1)-SV40, a hormonally independent transgenic mouse model for breast and prostate cancer;
- HPV-infected (K14-HPV16 heterozygote) mouse expressing the E6 and E7 proteins of HPV16 which develop skin and anal lesions and cervical squamous carcinomas from progression of cervical intraepithelial neoplasia (CIN)-like lesions following treatment with exogenous estrogen; DFMO is effective;
- Mice bearing a dominant negative mutation in p53 which is highly susceptible to carcinogen induced lung, colon, and liver cancers and uterine sarcomas. These mice are presently being used in smoke-induced carcinogenesis and prevention studies; dexamethasone and piroxicam have activity in this model;
- SKH hairless mice which develop UV-induced skin cancers and which have p53 mutations and dipyrimidine dimers; indomethacin, celecoxib, and DFMO are active;
- Neu-mutated mouse model for hormone-dependent and hormone-independent cancers; perillyl alcohol has activity;
- Strain A/J mouse exposed to side stream tobacco smoke develop lung tumors; indole-3-carbinol and oltipraz are effective in this model;
- Whole body exposure of Swiss albino mice to environmental cigarette smoke induces a appreciable number of lung tumors.
- BRCA-1 conditional knockout mice which appear to model BRCA-1 carriers;
- Colorectal cancer prevention in normal and transgenic mice (Apc1638, MLH-1, MLH-1/Apc1638) on folate deficient diet.
Several in vitro assays have enabled direct study of the potential inhibitory effect of preventive agents on molecular targets of particular importance in cancer development. Examples include: - ras oncogene inhibition as measured by protein farnesylation inhibition (using rat brain farnesyl protein transferase)
- epidermal growth factor receptor (EGFR) inhibition (measured as inhibition of EGFR from A431 human epidermoid carcinoma cells in phosphorylation of angiotensin).
- aromatase inhibition (measured as the inhibition of the enzyme from human placenta).
- HMG-CoA reductase inhibition (measured as inhibition of the enzyme from rat liver microsomal fraction).
- survivin gene expression inhibition in human cancer cell lines.
New mechanism-based in vitro/ in vivo systems for screening cancer preventive agents include: - Cell-based reporter assays to screen and prioritize validated chemopreventive agents against specific orphan receptors;
- Human-derived premalignant lesions (colon, breast, prostate), wherein the biopsy specimens are briefly grown as cell cultures, and are then orthotopically implanted in athymic mice;
- In vivo and in vitro characterization of phenotypically normal cells from autosomal dominant forms of cancer by genomic and proteomic approaches before and after treatment with preventive agents;
- genomic instability, Rb1, p16 and telomerase activity in human cells bearing p53 mutations, and paired sets of normal, preneoplastic and transformed cells;
- genomic instability, molecular changes and telomerase in cells from Apc mutant mice;
- cytokeratin alterations and apoptosis in dysplastic oral mucosal cells.
Back to Top
Intraepithelial neoplasia (IEN) as a Target for Preventive Intervention
IEN is the histologically-defined premalignant lesion and clinical target for prevention that is treated (usually surgically) in cases of oral leukoplakia, colorectal adenomas, cervical intraepithelial neoplasia, actinic keratosis, Barrett's esophagus, atypical hyperplasia and ductal and lobular carcinoma in situ in breast, and superficial bladder tumors. DCP-sponsored clinical studies with tamoxifen in breast and celecoxib in colon demonstrated that cancer preventive agents can be used to treat these premalignant lesions. The discovery and development of new cancer preventive agents to treat IEN are needed as new techniques and strategies in risk identification and modulation make early intervention feasible and practical. The evaluation of potential cancer preventive agents in IEN is a priority of CADRG.
Back to Top
Examples of Current Research Areas
Recent progress in the CADRG research program has included support for the development of specific agents, as well as efforts to address significant issues in chemoprevention research. CADRG-sponsored support of the development of specific cyclooxygenase-2 (COX-2) inhibitors is discussed in detail below. Also addressed is the development of aerosolized glucocorticoids and other approaches directed to chemoprevention of lung cancer in current and former smokers. In addition, CADRG-sponsored research in support of estrogen-receptor negative breast cancer prevention is discussed.
Back to Top
Development of NSAIDs and Selective COX-2 Inhibitors
Epidemiological observations suggesting that NSAIDs might be effective in reducing colorectal cancer were made almost 20 years ago. Cancer preventive efficacy testing of NSAIDs in rodent models of colon, bladder, and skin carcinogenesis sponsored by the CADRG showed that: (a) NSAIDs including piroxicam, sulindac, ibuprofen, and indomethacin are highly effective in blocking colon carcinogenesis in rats and mice and (b) NSAIDs can be introduced later in process of carcinogenesis and still be highly effective in preventing cancer development.
In addition to their high efficacy, most traditional NSAIDs have risks that must be weighed against their benefits; the most prominent risks are gastrointestinal bleeding and renal dysfunction, attributed to the lack of protective prostaglandins which results from COX inhibition. Several strategies have been suggested for improving the benefit/risk ratio of NSAIDs. For example, one approach evaluated by CADRG is to lower the dose and combine the NSAID with another agent that is also effective but by an independent mechanism of action. In this regard, NSAIDs combined with DFMO (an ornithine decarboxylase inhibitor) were found to have synergistic efficacy. The discovery about 10 years ago of two COX forms (constitutive COX-1 and inducible COX-2) provided another opportunity for improving the benefit/risk ratio of NSAIDs. Higher levels of COX-2 occur in inflammation and are widely associated with cancer progression and tumor angiogenesis. For example, COX-2 levels are increased in bladder, skin (basal cell, squamous cell, and melanoma), endometrial, breast, non-small cell lung, pancreatic, and esophageal cancers. This suggested that COX-2 specific inhibitors might be effective cancer preventive agents and that the toxicity associated with COX-1/2 inhibitors might be reduced.
The most common genetic mutation found in sporadic human colon cancer is in the adenomatous polyposis coli (APC) gene. This gene is also mutated in the germline of individuals with the autosomal dominant cancer syndrome familial adenomatous polyposis (FAP). These individuals typically develop hundreds of intestinal polyps and colon cancer at an early age. The Min (multiple intestinal neoplasia) mouse is one animal model of this syndrome. The Min mouse carries a germline mutation in the APC gene and develops multiple intestinal adenomatous polyps. Exposure of Min mice to celecoxib prior to the development of polyps prevents adenoma formation while treatment later in carcinogenesis, in mice bearing polyps, reduces polyp number. The results of the late intervention in Min mice helped to justify a recent, successful DCP-sponsored trial in patients with FAP who have pre-existing polyps.
Interestingly, NSAIDs and COX-2 inhibitors had very little efficacy in transgenic mice which model lesions other than hereditary forms of colon cancer. For example, COX inhibitors are inactive to prevent (a) desmoid lesions, a pathology often observed in FAP patients, in Min mice and (b) tumors in mice with defects in MLH1/MSH2 DNA repair genes, which parallel cancers in patients with hereditary nonpolyposis colorectal cancer (HNPCC) syndrome.
In CADRG-sponsored studies celecoxib was highly effective in preventing AOM-induced rodent colorectal cancers when administered either throughout the experiment or later in carcinogenesis. The latter experimental design has particular relevance to patients who have had an adenoma(s) and are at increased risk for additional adenomas and colorectal cancer. Based on these preclinical data and on the clinical data in patients with FAP, a Phase II clinical trial is underway to study the effects of COX-2 inhibitors in individuals with prior sporadic colon adenomas.
COX-2 inhibitors are presently being evaluated in a number of other rodent tumor models, including tongue, esophagus, breast, lung, prostate, bladder and skin. Celecoxib has shown efficacy in many of these tissues; for example, the agent reduced by 90% the incidence of both OH-BBN induced bladder cancer and UV-induced skin cancers in rodents when administered at the time of initiation or later during cancer progression. Positive findings such as these support the continued DCP-sponsored Phase II clinical chemopreventive efficacy testing of celecoxib and other specific COX-2 inhibitors in target organs besides colon. Phase II clinical efficacy trials of celecoxib are currently underway in patients with previous superficial bladder cancers, and in patients with actinic keratosis and with basal cell nevus syndrome. Other clinical Phase II efficacy studies of the drug are ongoing in premalignant lesions of the breast, prostate, esophagus, and oral cavity.
Back to Top
Development of Aerosolized Glucocorticoids as Topical Cancer Preventives for Smoking-Related Lung Cancer
Biomarkers of lung, skin and bone marrow carcinogenesis are evaluated in SKH-1 hairless mice exposed to smoke and/or UV light (where sulindac exerted a variety of protective effects). The strain A/J mouse, when exposed to carcinogens, develops lung adenomas which can progress to adenocarcinomas. This validated model has also been used to evaluate prevention of tobacco smoke-induced cancers. In studies supported by the CADRG and in those reported in the literature, a number of agents were found to inhibit adenomas in strain A/J mice on systemic administration. In order to reduce the systemic toxicity of some of these agents, the use of local aerosol delivery of chemopreventives to the lung is being explored.
Aerosol formulations of the anti-asthmatic glucocorticoid budesonide and the phase 2 metabolic enzyme modulators oltipraz and phenethyl isothiocyanate were found to prevent lung cancers in the A/J model. Budesonide was the most effective of the three agents and showed a dose-response related inhibition of tumor multiplicity up to 90%. A second glucocorticoid, beclamethasone, was also effective. These experiments demonstrated that topical glucocorticoids could be effective against lung cancer with minimal systemic toxicity.
Additional studies in the A/J lung cancer model have demonstrated that: (a) treatment with budesonide started during the progression stage of carcinogenesis is effective in inhibiting growth of adenomas and carcinomas, (b) the activity is independent of the carcinogen used to induce tumors, and (c) glucocorticoids are effective in neoplasias which have p53 mutations. p53 mutations are often evident early in the development of smoking-related cancers of lung, head and neck, and esophagus. Mutant p53+/- A/J mice are more sensitive than p53 wild-type mice to smoke-related lung carcinogenesis.
Aerosolized budesonide is under evaluation in a DCP-funded Phase II clinical trial of lung cancer prevention in current and former smokers. This approach illustrates the strategy of topical delivery to increase regional dose intensity and minimize systemic toxicity and first-pass metabolism. Topical formulations are being developed for other cancer preventive target organs such as skin, colon, and head and neck.
Farnesyl transferase inhibitors (FTIs) are also effective in preventing growth of lung adenomas in A/J mice when administered either early or late during carcinogenesis. These data are being considered by various investigators to support a possible clinical cancer prevention trial with FTIs.
Back to Top
Non-Hormonal Approaches to Prevention of Breast Cancer
Several classes of agents treat and potentially prevent hormonally responsive (ER+) breast cancer by altering estrogen activity. Perhaps the most important class of these agents is the antiestrogens/estrogens such as tamoxifen, toremifene, and raloxifene, which were synthesized specifically to interact with the ER receptor and may be the first and one of the more successful examples of molecular target-mediated treatment in cancer. Additionally, this class of agents has proven to be highly effective in preventing the development of ER+ breast tumors. Antiestrogens and agents affecting tissue and systemic estrogen levels (e.g., aromatase inhibitors, which have been extensively studied in the program) are likely to remain the primary approach to preventing hormonally responsive tumors.
Although the majority of breast cancers are ER+, a significant subset of cancers, particularly in younger women, are not responsive to hormonal manipulation and may represent a significantly different disease. Thus, achieving a prevention strategy for all human breast cancer will require the development of agents or agent combinations which might work preferentially against hormonally nonresponsive tumors or that can work against both hormonally responsive and nonresponsive tumors. Classes of agents that appear to be effective via mechanisms independent of the estrogen pathway and that might prove to be effective independent of hormonal responsiveness of tumors include RXR agonists, FTIs, EGFR inhibitors, antiproliferatives and certain NSAIDs.
Anticipating the need to screen for agents which might be effective against hormonally nonresponsive mammary tumors, CADRG has sponsored research to characterize and perform initial cancer prevention studies in two models which appear to be hormonally non-responsive. In the first, sporadic ductal mammary tumors are induced in rats by introduction of a mutated neu retrovirus. If the recipient rats are ovariectomized, 100 of the rat tumors are hormonally independent. If the recipient animals are not ovariectomized then 50% of the mammary tumors are hormonally independent. The efficacy of RXR agonists has been observed in this hormonally independent. model. In the second model, SV40 T-antigen-driven transgenic mice, all breast tumors are hormonally non-responsive. In this model RXR agonists were moderately effective against breast carcinogenesis. Further studies in these models are likely to identify strong candidate agents for preventing hormonally nonresponsive tumors, including agents directed against neu or EGFR. In addition, a transgenic conditional BRCA-1 knockout mouse is being evaluated as a model for BRCA-1 mediated hereditary breast cancer in women. This model should be useful for identifying agents with potential cancer preventive activity in this high risk cohort, in which most tumors are hormonally unresponsive.
Back to Top
Interactions with the External Scientific Community
CADRG has initiated several formal mechanisms for soliciting input and collaboration with the academic and industrial research communities. The Rapid Access to Preventive Intervention Development Program (RAPID) supports and facilitates new opportunities in molecularly- targeted preventive agent discovery from academic scientists. CADRG also plays a key role in establishing and maintaining interactions with industrial collaborators. In addition, CADRG has also sought to enhance the awareness and access of the research community to recent results in prevention research.
Back to Top
RAPID
RAPID facilitates the future discovery of preventive agents by making the resources of the CADRG agent development program available to academically based investigators. While CADRG has historically provided investigational agents, toxicological, and regulatory support to grantee investigators and institutions on an ad hoc basis, the RAPID program supports the formal expansion of this effort to compound synthesis and formulation under Good Manufacturing Practices regulations, preclinical efficacy and toxicity testing, and regulatory support through Phase I clinical testing. The RAPID program provides support to academic investigators for these aspects of agent development. In FY2000, resources were committed to seven projects; support for an additional five and six projects was initiated in FY2001 and FY2002, respectively (see http://www.cancer.gov/prevention/
programs-resources/programs/rapid/projects). These projects have entailed development of novel lipoxygenase inhibitors, farnesyl transferase inhibitors, developmental peptides, retinoids, deltanoids and other promising agents/classes. RAPID will continue to support the development of these new cancer preventive agents introduced by the academic community through Phase I clinical studies.
Back to Top
Industrial Interactions
CADRG and the DCP has also continued to solicit and cultivate the interests of the pharmaceutical, food, nutritional supplement, biotechnology industries and advocacy groups in the promise of cancer prevention. Forty-five (45) Clinical Trials, Material Transfer, and other agreements with these industrial collaborators are in place or pending. Several successful prevention studies have been completed in collaboration with the DCP. The pharmaceutical industry is increasingly recognizing the feasibility and economics of cancer prevention; in many cases where noncancer indications have been developed, preclinical efficacy testing is all that is needed to support Phase II clinical testing in cancer prevention setting. For promising nonpatentable chemical entities without industrial support, the CADRG program has synthesized and/or purified and agents such as indole-3-carbinol, phenethyl isothiocyanate (PEITC), curcumin, and selenomethionine and is making these products available to the research community. CADRG has also developed and maintained INDs for such agents. In collaboration with the food industry, CADRG has developed complex mixtures derived from natural products, such as soy-derived isoflavone and protease inhibitors and tea-derived mixtures and formulations of polyphenolic catechins, under pharmaceutical standards. The American public is spending up to 10 billion dollars annually on complementary and alternative medicines, such as dietary supplement and nutriceutical products. CADRG staff serves on the advisory board for the NCI's effort in complementary and alternative medicine development, and the CADRG program is also advancing the evaluation of such products in randomized, blinded clinical trials.
The CADRG program continues to coordinate meetings between academic investigators, industrial representatives, and NCI staff to address specific agent development issues related to the efficient clinical evaluation of promising agents.
Back to Top
Cancer Prevention Informatics
To increase the awareness and access of the research community to recent results in prevention research, CADRG has recently undertaken a project to make its existing databases of nonproprietary bibliographic citations and monographs on potential cancer preventive agents, as well as current products and all developmental test results, available to the research community via the internet. The existing information is being incorporated into a relational database system that will support the preclinical and clinical information needs of the DCP and that will be compatible with the informatics systems of the NCI. This project is intended to provide a resource for the cancer research community of independent investigators, the Community Clinical Oncology Program (CCOP) network, cancer centers, the SPOREs, and the Early Detection Research Network (EDRN). Advocacy groups have also expressed interest in cancer prevention research and are being included in agent evaluation strategies to provide internet links to disseminate information and support participation in clinical trials.
Back to Top
