Breast cancer prevention--clinical trials strategies involving aromatase inhibitors

Estrogens and their metabolites have been implicated in both the initiation and the prevention of breast cancer. The reduction in breast cancer incidence seen in the tamoxifen arms of the four prospective trials to date has established the proof of principle that antagonizing estrogen is a potential means of reducing breast cancer risk. However, the areas to improve on these results include: (a) enhanced efficacy, (b) reduction in the incidence of receptor-negative tumors, (c) improved overall and endocrinological side effects, and (d) improved function on end-organs other than the breast. The aromatase inhibitors offer the potential to achieve these goals in part in the following ways: (a) greater reduction in risk of disease as evidenced by superior efficacy in advanced breast cancer and by inhibition of both initiation and promotion of breast cancer, (b) reduction in receptor-negative tumors by synergy with COX-2 inhibitors resulting in growth factor inhibition, anti-angiogenesis and inhibition of tumor-associated aromatase expression, (c) fewer vasomotor and urogenital abnormalities, and (d) reduced thromboembolism and cardiovascular complications and satisfactory effects on bone metabolism. Important differences may exist between non-steroidal reversible inhibitors and steroidal irreversible inactivators in particular related to the androgenic/anabolic effects of the steroidal inactivators. Pilot studies of aromatase inhibitors described elsewhere in this session have begun in healthy women with dense mammography, or a high-risk genetic and/or histocytopathologic profile, to determine potential efficacy, as well as effects on end-organ function. A number of phase three trials with aromatase inhibitors are also underway or in planning. Among these are the BRCA 1 and 2 study of exemestane versus placebo in unaffected postmenopausal carriers, the International Breast Intervention Study 2 (IBIS 2) of anastrozole versus placebo in women with a high-risk profile, and the National Cancer Institute of Canada’s Clinical Trial Group (NCIC CTG) study of exemestane with or without celecoxib versus placebo in women at risk of the disease. For premenopausal women, combination strategies of gonadotrophin agonists and aromatase inhibitors are being investigated. The potential of using low doses of aromatase inhibitors to lower “in breast” estrogen levels without unduly perturbing plasma concentrations is also being explored. The potential of the aromatase gene functioning as an oncogene within the breast may be tied to breast density which in turn may represent both a selection tool for elevated risk and an intermediate marker of prevention. The strong link between postmenopausal estrogen levels and breast cancer risk suggests the possibility that plasma estrogen levels may be a useful intermediate marker of prevention. The aromatase inhibitors offer us the first ever tool to render women virtually free of estrogen and are potentially an exciting tool for the prevention of breast cancer.     

A response-adaptive randomization procedure for multi-armed clinical trials with normally distributed outcomes

We propose a novel response-adaptive randomization procedure for multi-armed trials with continuous outcomes that are assumed to be normally distributed. Our proposed rule is non-myopic, and oriented toward a patient benefit objective, yet maintains computational feasibility. We derive our response-adaptive algorithm based on the Gittins index for the multi-armed bandit problem, as a modification of the method first introduced in Villar et al. (Biometrics, 71, pp. 969-978). The resulting procedure can be implemented under the assumption of both known or unknown variance. We illustrate the proposed procedure by simulations in the context of phase II cancer trials. Our results show that, in a multi-armed setting, there are efficiency and patient benefit gains of using a response-adaptive allocation procedure with a continuous endpoint instead of a binary one. These gains persist even if an anticipated low rate of missing data due to deaths, dropouts, or complete responses is imputed online through a procedure first introduced in this paper. Additionally, we discuss how there are response-adaptive designs that outperform the traditional equal randomized design both in terms of efficiency and patient benefit measures in the multi-armed trial context.     

Application of Partially Balanced Block Designs to Factorial Experiments

In this paper we present partially balanced block designs with F_p and C_p association schemes. These designs play a great role in the experimental theory since they can be applied to factorial experiments. A particular stress is put on the formulation of association schemes _p and C_p, presentation of exemplary constructions of these designs and their analysis of variance. The presented considerations are abundantly illustrated with examples.     

Analysis of Crossover Designs with a Categorical Response

In the analysis of two-period crossover designs, one frequently must consider the case of a categorical response with binary as a special case. These circumstances are considered in this paper and the proposed method—an extension of GART'S tests for the binary case—is similar to that of PIKE, CASAGRANDE and SMITH for the analysis of pair-matched case-control studies. The results are illustrated with a numerical example.     

On Construction of Some Augmented Block Designs

Augmented block designs are frequently used in practice. In this paper procedures of construction of some types of augmented block designs are discussed using the concept of efficiency-balance and partially efficiency-balance.     

Augmented p-rep designs

Early generation variety trials are very important in plant and tree breeding programs. Typically many entries are tested, often with very little resources available. Unreplicated trials using control plots are popular and it is common to repeat the trials at a number of locations. An alternative is to use p-rep designs, where a proportion of the test entries are replicated at each location; this can obviate the need for control plots. α-Designs are commonly used for replicated variety trials and we show how these can be adapted to produce efficient p-rep designs.     

On the analysis of phylogenetically paired designs

As phylogenetically controlled experimental designs become increasingly common in ecology, the need arises for a standardized statistical treatment of these datasets. Phylogenetically paired designs circumvent the need for resolved phylogenies and have been used to compare species groups, particularly in the areas of invasion biology and adaptation. Despite the widespread use of this approach, the statistical analysis of paired designs has not been critically evaluated. We propose a mixed model approach that includes random effects for pair and species. These random effects introduce a “two-layer” compound symmetry variance structure that captures both the correlations between observations on related species within a pair as well as the correlations between the repeated measurements within species. We conducted a simulation study to assess the effect of model misspecification on Type I and II error rates. We also provide an illustrative example with data containing taxonomically similar species and several outcome variables of interest. We found that a mixed model with species and pair as random effects performed better in these phylogenetically explicit simulations than two commonly used reference models (no or single random effect) by optimizing Type I error rates and power. The proposed mixed model produces acceptable Type I and II error rates despite the absence of a phylogenetic tree. This design can be generalized to a variety of datasets to analyze repeated measurements in clusters of related subjects/species.