Other requirements were an Eastern Cooperative Oncology Group (ECOG) performance status of 0–1; life expectancy > 6 months; negative metastatic workup (bone scan and CT chest/abdomen/pelvis); adequate bone marrow, liver, and kidney function; and peripheral neuropathy ≤ grade 1. All patients of child-bearing potential were required to consent to dual methods of contraception during treatment and for 3 months afterward. A negative pregnancy test was required for these women before treatment, and any suspicion of pregnancy had to be reported to the treating physician.
Study endpoints
The primary endpoint of the study was the in-breast pCR after four cycles of platinum-based neoadjuvant chemotherapy. Pathologic complete response was defined as complete disappearance of invasive and in situ disease or invasive disease alone. During the course of this trial, it became generally acceptable to include patients with only residual DCIS as equivalent to pCR.45
The secondary endpoints were pCR in the lymph nodes; clinical response rate; tolerability; breast conservation; time to disease progression (local, regional, and distant); and OS. Also recorded was minimal residual disease (MRD), which we arbitrarily defined as ≤ 1 cm invasive carcinoma at resection. The overall treatment plan included postoperative AC to provide a standard-of-care regimen to maximize curative potential.
Statistical analysis
Data were analyzed on an intentto- treat basis. Although pCR rates with doxorubicin plus either cyclophosphamide or docetaxel have been < 15%, the studies by Smith et al26 and Hurley et al39 with in-breast pCR rates of at least 20% served as comparators (albeit imprecise).
Applying the min/max statistical design, the procedure tests the null hypothesis H0: P ≤ 0.15 against the alternative hypothesis H1: P ≥ 0.30. The overall level of significance and power for this design are 5% and 80%, respectively. The sample size needed for the first stage was 23 evaluable patients. If three or fewer pCR responses were observed, then the study would be terminated and the treatment regimen would not be investigated further. Otherwise, an additional 25 evaluable patients would be accrued for a total of 48 study patients. If 11 or fewer responses were observed, then the study would be terminated. Otherwise, this treatment regimen would be recommended to proceed to phase III for further investigation.
Tolerability assessment
At each visit, toxicities were assessed and graded according to the National Cancer Institute Common Toxicity Criteria, version 2.46 Two dose reductions were allowed for all drugs.
Ethical considerations
The investigational nature of this study was fully disclosed to each patient. In accordance with institutional and federal guidelines, the patients were guided through and subsequently signed the informed consent approved by the appropriate site Institutional Review Board.
Literature review
The terms “neoadjuvant” and “breast” were used in a literature search on PubMed, with filters “English” and “clinical trials.” Abstracts for each of the 398 results were reviewed We used phase II or III trials with at least 30 patients, at least four cycles of chemotherapy, and clearly defined pCR for comparison to this study.
Results Patients
Between June 2003 and December 2006, 50 women with a median age of 49 years (range, 28–75 years) were enrolled. One patient was ineligible due to preceding lumpectomy. The 49 eligible patients were treated with ≥ 1 cycle of neoadjuvant chemotherapy between June 27, 2003, and April 12, 2007.
The baseline characteristics of the 49 eligible patients are summarized in Table 3. Thirty-one patients (63%) were premenopausal. Twenty patients (41%) were positive for either ER or PR and were negative for HER2. Eight patients (16%) had HER2- positive tumors, and 23 (46%) had triple-negative tumors. At baseline, 22 patients (45%) had clinical lymphadenopathy, and 1 patient (2%) had inflammatory breast cancer.
The 41 patients (83%) who completed all four cycles of therapy were evaluable for response; 8 (16%) were inevaluable due to noncompliance (1), grade 3 or 4 toxicity (5), or withdrawal of consent (2). The following efficacy assessments apply to the 41 evaluable patients, whereas the toxicity assessments include the 49 patients who received at least one full cycle of chemotherapy.
Clinical response
At study onset, of the 49 eligible patients, 38 (78%) had a palpable inbreast tumor (median size, 5.5 cm); 22 (45%) had enlarged nodes, and 34 (69%) had confirmed nodal involvement (by biopsy or imaging). A clinical complete response (cCR) rate in the breast was seen in 23 of 41 (56%) evaluable patients. Of 22 patients with baseline lymphadenopathy (by imaging or physical examination), 13 had axillary assessment by physical examination throughout treatment, with 12 (92%) exhibiting a cCR in the axilla.