Prostate cancer chemoprevention: Strategies for designing efficient clinical trials

R. Lieberman, Prostate cancer chemoprevention: Strategies for designing efficient clinical trials, UROLOGY, 57(4A), 2001, pp. 224-229
Citations number
Categorie Soggetti
Urology & Nephrology
Journal title
ISSN journal
0090-4295 → ACNP
Year of publication
224 - 229
SICI code
A chemoprevention (CP) strategy has evolved for conducting efficient clinic al trials for prostate cancer (PCa) prevention. It integrates five key comp onents, including agents, biomarkers, cohorts, designs, and endpoints. The rationale for the CP strategy relates to the natural history of prostate ca ncer. There is a wide array of natural and synthetic agents that hold promi se for inhibiting, reversing, or modulating the transition from normal to p recancer and from precancer to cancer. These agent classes include antiandr ogens, antiestrogens, phytoestrogens, antioxidants, anti-inflammatory (proa poptotic) agents, antiproliferation/antidifferentiation agents, signal tran sduction modulators of receptor tyrosine kinase and ras farnesylation, anti angiogenesis agents, insulinlike growth factor (IGF)-1, peroxisome prolifer ator-activator receptor modulators (-gamma and -delta), and gene-based inte rventions. Biomarkers and endpoints are guided by the level of evidence req uired leg, phase 1, 2, 3). Two candidate surrogate endpoints (SE) based on histology are high-grade prostatic intraepithelial neoplasia (HGPIN) and co mputer-assisted image analysis of dysplastic lesions. Phase 1 trials use st andard endpoints of safety, pharmacokinetics and limited pharmacodynamics. Phase 2 trials use endpoints of modulation of biomarkers and correlation wi th histology. Phase 3 trials use endpoints of clinical benefit, such as can cer incidence reduction and quality of life, Validation of a biomarker as a SE involves correlation of the biomarker with clinical benefit. Cohorts (t arget populations) for phase 2/3 trials include the general population of m en over age 50 with a normal prostate-specific antigen (PSA), subjects with a strong family history of PCa, subjects with elevated PSA/negative biopsy , and subjects with HGPIN/negative biopsy. These at-risk populations reflec t key individual risk factors (age, race, serum PSA [free/total]; serum IGF -1/IGF binding protein (IGFBP)-3; 1, 25(OH)(2) D3; family history of PCa; c arriers of PCa susceptibility genes [ELAC2, CYP3A4, SRD5A2, etc.]; and hist ology such as atypia and HGPIN) that could be combined into a multivariate risk model for PCa. The probability of cancer risk (recurrence) is a key fa ctor that impacts on the clinical trial design (power, sample size, and pri mary endpoint). Multivariate predictive mathematical models for biochemical recurrence after radical prostatectomy by decreasing sample size and time to clinical outcomes maximize trial efficiency and identify the patients mo st likely to benefit from secondary prevention. The two large primary preve ntion trials, Prostate Cancer Prevention Trial/Seleninium and Vitamin E Che moprevention Trial (PCPT/ SELECT), in low- and average-risk subjects have s ample sizes of 18,000 to 32,000, with a treatment duration of 7 years to de tect a 25% reduction in biopsy-proven PCa. Subjects with HGPIN have the hig hest known cancer risk (approximately 50% at 3 years), and thus require a s mall sample size (n = 450) to detect a 33% reduction in cancer incidence. A schema involving three sequential trials for agent registration is describ ed. in summary, a CP strategy that incorporates well-defined agents, clinic al and validated SE, and high-risk cohorts defined by genetic at-id acquire d risk factors in a series of well-designed randomized controlled trials pr ovides an efficient pathway for evaluating and approving new agents for PCa prevention.