Trial Information

Intrinsic Breast Cancer Subtypes and Benefit of Paclitaxel in CALGB 9344 and Dose Dense Therapy in CALGB 9741

OBJECTIVES:

Primary

- To determine whether subtype-specific treatment effects correlate with disease-free
survival (DFS), as determined by a significant interaction between PAM50-based
intrinsic subtypes, and (a) paclitaxel benefit in CLB-9344 and (b) dose density in
CALGB-9741.

- To determine whether subtype-specific treatment effects correlate with DFS for the
HER2-negative subsets in CALGB-9344 and CALGB-9741, as determined by analysis of tissue
microarray (TMA) and slides.

- To determine the relationship between PAM50-defined risk of relapse (ROR) score and DFS
in CALGB-9344 and CALGB-9741.

- To evaluate the relationship between PAM50-defined ROR score and DFS in the
HER2-negative subsets in CALGB-9344 and CALGB-9741, as determined by analysis of TMA
and slides.

- To examine the relationship between PAM50-defined proliferation score and DFS in
CALGB-9344 and CALGB-9741 in multivariate Cox-proportional hazards models including the
following covariates: (a) number of positive lymph nodes, square root transformation;
(b) menopausal status (pre versus peri/post); CALGB-9344 only; c) dose of doxorubicin
hydrochloride (60/75/90 mg/m^2); and CALGB-9741 only; and (d) sequence of treatment.

Secondary

- To evaluate overall survival (OS) in a Cox-proportional hazards-regression model for
testing the interaction between ROR with (a) paclitaxel benefit in CALGB-9344 and (b)
dose density in CALGB-9741.

- To test for a significant interaction between ROR and paclitaxel benefit at 5-year and
10-year DFS.

- To test whether 5-year and 10-year DFS rates can be associated to a significant
interaction between the proliferation score with (a) paclitaxel benefit in CALGB-9344
and (b) dose density in CALGB-9741.

OUTLINE: Tissue blocks from CALGB-9344 and CALGB-9741 are utilized to purify RNA to be
tested in the PAM50 assay (a 50-gene quantitative PCR assay, that provides an intrinsic
breast cancer subtype diagnosis) and generate risk of relapse (ROR) scores.

The assay identifies five subtypes with the following characteristics:

- Luminal A: This subtype expresses estrogen receptor (ER) accompanied by high levels of
ER-associated gene expression. Genes associated with cell cycle activation are not
highly expressed and this tumor type is only very rarely HER2+. This subgroup has the
most favorable prognosis and is enriched for endocrine therapy responsive tumors.

- Luminal B: This subtype expresses ER and ER-associated gene expression but to a lower
extent. Genes associated with cell cycle activation are highly expressed and this tumor
type can be HER2+ (~20%) or HER2- thus, from the clinical perspective, Luminal B tumors
are at least two further subtypes defined by the presence or absence of HER2-gene
amplification. The prognosis is unfavorable (despite ER expression) and endocrine
therapy responsiveness is generally diminished.

- Basal-like: This subtype is ER-, is almost always clinically HER2- and expresses a
suite of "basal" biomarkers. Genes associated with cell cycle activation are highly
expressed.

- HER2-enriched: This subtype is ER- and is HER2+ in the majority of cases. Genes
associated with cell cycle activation are highly expressed and these tumors have a poor
outcome. Tumors within this classification that are clinically HER2- fall into a class
previously described as double-negative non-basal.

- Normal-like: A tumor subtype diagnosis cannot be provided from samples that exhibit a
normal-like profile. Since this profile was trained on samples without cancer,
"normal-like" implies there are too few tumor cells in the sample to make a true tumor
subtype diagnosis.

PROJECTED ACCRUAL: A total of 2,245 tissue blocks from CALGB-9544 and 1,432 tissue blocks
from CALGB-9741 will be accrued for this study.