Paclitaxel With or Without Carboplatin and/or Bevacizumab Followed by Doxorubicin and Cyclophosphamide in Treating Patients With Breast Cancer That Can Be Removed by Surgery

PRIMARY OBJECTIVES:

I. To determine whether adding bevacizumab to neoadjuvant weekly paclitaxel (+/- carboplatin)
and subsequent dose-dense doxorubicin and cyclophosphamide (ddAC) significantly raises the
rate of pathologic complete response (pCR) in the breast in patients with hormone receptor
(HR)-poor/human epidermal growth factor receptor 2 (HER2) (-), resectable breast cancer.

II. To determine whether adding carboplatin every 3 weeks to neoadjuvant weekly paclitaxel
followed by ddAC (+/- bevacizumab) significantly raises the rate of pCR in the breast in
patients with HR-poor/HER2(-), resectable breast cancer.

III. To determine whether adding bevacizumab every 2 weeks to neoadjuvant weekly paclitaxel
(+/- carboplatin) and subsequent ddAC significantly raises the rate of pCR in the breast in
patients with basal-like breast cancers, as defined by gene expression array.

IV. To determine whether adding carboplatin every 3 weeks to neoadjuvant weekly paclitaxel
followed by ddAC (+/- bevacizumab) significantly raises the rate of pCR in the breast in
patients with basal-like breast cancers, as defined by gene expression array.

SECONDARY OBJECTIVES:

I. To determine the pCR rates in the breast and axilla, using American Joint Committee On
Cancer (AJCC) TNM criteria (version 6), to neoadjuvant weekly paclitaxel, with or without
carboplatin, followed by ddAC, with or without bevacizumab, given concurrently with the
weekly paclitaxel and ddAC, in (a) patients with HR-poor/HER2(-), resectable breast cancer
and (b) the subset of patients with basal-like breast cancers, as defined by gene expression
array.

II. To assess whether there is an interaction between the addition of carboplatin and
bevacizumab to neoadjuvant chemotherapy (NAC) with weekly paclitaxel followed by ddAC as
regards the path pCR rates in (a) patients with HR-poor/HER2(-), resectable breast cancer and
(b) the subset of patients with basal-like breast cancers, as defined by gene expression
array.

III. To assess the toxicity of the control regimen (weekly paclitaxel followed by ddAC) and
any incremental toxicities associated with the addition of carboplatin and/or bevacizumab in
this patient population, including the incidence of febrile neutropenia, grade >= 3
thrombocytopenia, grade >= 2 neurotoxicity, grade >= 3 hypertension, and clinically
significant bleeding or thrombotic (including cardiovascular and cerebrovascular) events.

IV. To determine the recurrence-free survival (RFS) measured from definitive surgery to first
event, and time to first failure (TFF) measured from study entry to first event.

V. To determine overall survival (OS), defined as time from registration to death from any
cause.

VI. To assess the impact of NAC with weekly paclitaxel followed by ddAC, with or without
carboplatin and/or bevacizumab, on axillary lymph node involvement at surgery, particularly
in patients with clinically or histologically positive axillary lymph nodes prior to
initiation of NAC.

VII. To assess the impact of the addition of bevacizumab to NAC on the incidence and severity
of post-op complications, especially excessive bleeding, delayed wound healing, and
thrombotic complications.

VIII. To evaluate residual cancer burden (RCB) as a predictor of RFS, TFF and OS.

IX. To determine the correlation between clinical, radiographic, and pathologic response.

TERTIARY OBJECTIVES:

I. To assess whether the impact of the addition of carboplatin and/or bevacizumab to NAC with
weekly paclitaxel followed by ddAC on achievement of pathologic CRs in patients with
HR-poor/HER2(-), resectable breast cancer is influenced by molecular subtype, as defined by
gene expression array.

II. To obtain blood, fresh frozen and fixed tumor tissue to test specific hypotheses for
which biomarker data exist and to evaluate biomarkers in tissue, blood, and serum that may
influence response to and toxicity of weekly paclitaxel, ddAC, carboplatin, and/or
bevacizumab.

III. To obtain blood samples to test specific hypotheses for which biomarker data exist and
to evaluate biomarkers in blood that may influence response to and toxicity of weekly
paclitaxel, ddAC, carboplatin and/or bevacizumab.

IV. To determine the surgical practice patterns for breast conservation and sentinel
lymphadenectomy in patients undergoing neoadjuvant chemotherapy.

V. To examine the practice patterns and use of sentinel lymphadenectomy (pre-chemotherapy or
post-chemotherapy) in patients with T2 or T3 breast cancer.

VI. To examine the proportion of patients who presented with T2 or T3 cancers who undergo
mastectomy despite cytoreduction adequate for breast conservation.

VII. To determine the radiotherapy practice patterns for post-mastectomy and regional nodal
irradiation in patients undergoing neoadjuvant chemotherapy.

OUTLINE: Patients are randomized to 1 of 4 treatment arms.

ARM I: Patients receive paclitaxel intravenously (IV) over 60 minutes once weekly in weeks
1-12. Patients then receive dose-dense doxorubicin hydrochloride IV over 3-10 minutes and
cyclophosphamide IV over 5-60 minutes (ddAC) once in weeks 13, 15, 17, and 19.

ARM II: Patients receive paclitaxel and ddAC as in Arm I. Patients also receive bevacizumab
IV over 30-90 minutes once in weeks 1, 3, 5, 7, 9, 11, 13, 15, and 17.

ARM III: Patients receive paclitaxel and ddAC as in Arm I. Patients also receive carboplatin
IV over 30 minutes once in weeks 1, 4, 7, and 10.

ARM IV: Patients receive paclitaxel and ddAC as in Arm I, bevacizumab as in Arm II, and
carboplatin as in Arm III.

Patients in all arms undergo definitive surgery (i.e., modified radical mastectomy or
breast-conserving surgery with appropriate management of the axilla) between 4-8 weeks after
completion of neoadjuvant therapy.

After completion of study treatment, patients are followed up periodically for up to 10
years.

Inclusion Criteria:

- Invasive breast cancer, diagnosed by core needle or incisional biopsy (excisional
biopsy not permitted)

- The invasive tumor must be hormone receptor-poor, defined as both estrogen receptor
(ER) and progesterone receptor (PgR) negative or staining present in =< 10% of
invasive cancer cells by immunohistochemistry (IHC)

- The invasive tumor must be HER2-negative, defined as IHC 0-1+ or with a fluorescent in
situ hybridization (FISH) ratio (HER2 gene copy/chromosome 17) of < 2.0 if IHC 2+

- Clinical stage II-III invasive breast cancer with intent to perform surgical resection
after neoadjuvant therapy; patients with inflammatory breast cancer are not eligible;
staging to rule out metastatic disease is recommended for clinical stage III patients

- Patients with multicentric or bilateral disease are eligible if the target lesion
meets eligibility criteria

- Patient agrees to undergo pretreatment research biopsies

- No prior chemotherapy, hormone therapy, or radiation therapy with therapeutic intent
for this cancer

- The target lesion in the breast must be >= 1 cm, clinically or radiographically;
palpable or radiographically measurable axillary adenopathy will be recorded but will
not serve as measurable disease for the primary endpoint; patients with axillary
disease only (no identifiable tumor in the breast that is >= 1 cm on physical exam or
radiographic study) are not eligible to participate

- Patients with a history of significant bleeding episodes (e.g., hemoptysis, upper or
lower gastrointestinal [GI] bleeding) within 6 months of registration are not eligible

- No serious or non-healing wound, skin ulcers or bone fracture; no abdominal fistula,
gastrointestinal perforation or intra-abdominal abscess within the past 6 months; no
major surgical procedure within 28 days prior to randomization or anticipation of need
for major surgery during the course of study

- The following are not considered to be major surgical procedures that would be
prohibited in the 28 days prior to, or following study randomization: obtaining the
required research needle biopsies; placement of a radiopaque clip to localize a tumor
or tumors for subsequent surgical resection; placement of a port for central venous
access; fine needle aspiration of a prominent or suspicious axillary lymph node;
needle biopsy of a clinically or radiographically detected lesion to rule out
metastatic disease; or pretreatment sentinel lymph node sampling

- No baseline neuropathy grade >= 2

- Zubrod performance status 0-1

- Pregnant or nursing women are not eligible; all women of reproductive potential must
have a negative pregnancy test at baseline and agree to use an effective, non-hormonal
method of contraception during the entire period of treatment on the study

- Patients with congestive heart failure are not eligible, nor are patients with
myocardial infarction, unstable angina pectoris, an arterial thrombotic event, stroke
or transient ischemia attack (TIA) within the past 12 months, uncontrolled
hypertension (systolic blood pressure [SBP] > 160 or diastolic blood pressure [DBP] >
90), uncontrolled or symptomatic arrhythmia, or grade II or greater peripheral
vascular disease

- Patients must have a pretreatment multi gated acquisition (MUGA) scan or
echocardiogram with a left ventricular ejection fraction (LVEF) above the
institutional lower limit of normal

- Granulocytes > 1,000/mcl

- Platelets > 100,000/mcl

- Total bilirubin =< 1.5 x upper limits of normal

- Calculated or measured > 30 ml/min

- Urine protein =< 1+ or urine protein to creatinine (UPC) ratio < 1

- Patients discovered to have >= 2+ proteinuria at baseline must undergo a 24-hour
urine collection that must demonstrate < 1 g of protein/24 hr, or UPC ratio < 1
to allow participation in the study

- Serum alanine aminotransferase (ALT) =< 2.5 x upper limits of normal

- Serum beta human chorionic gonadotropin (HCG) negative (for women of child bearing
potential)

- Prothrombin time (PT)/international normalized ratio (INR) =< 1.5 x upper limit of
normal (ULN)

- Unless patient is on therapeutic doses of warfarin; if so, the patient must have
an INR =< 3 on a stable dose of warfarin, must have not active bleeding or
pathologic condition that is associated with a high risk of bleeding