Patients were randomly assigned to receive either 4 cycles of docetaxel 100 mg/m² on day 1 every 3 weeks followed by capecitabine 1000 mg/m² by mouth twice daily on days 1 to 14 every 3 weeks for 4 cycles (total 24 weeks) or docetaxel 50 mg/m² on day 1 concomitant with capecitabine 1000 mg/m² by mouth twice daily on days 1 to 7 every 2 weeks for 8 cycles (total 16 weeks). The total planned dose of each agent (400 mg/m²docetaxel and 112,000 mg/m² capecitabine) and number of cycles were equal in each arm. In case of tumor progression in the sequential arm, patients could proceed with capecitabine or discontinue study treatment (investigator's choice). In case of tumor progression in the concomitant arm, study treatment was discontinued and additional therapy was at the discretion of the investigator. Growth factor support was permitted. In case of excess toxicity, dose modification guidelines were followed.

After completion of chemotherapy, patients were required to undergo surgery. The type of surgery was at the discretion of the surgeon. After surgery, further chemotherapy was allowed at the investigator's discretion, pCR being the primary endpoint. Radiation therapy was given to all patients who underwent breast conservation. Postmastectomy radiation therapy was given to patients with large breast tumors and involved lymph nodes at the discretion of the radiation oncologist. All patients with estrogen receptor– and/or progesterone receptor–positive (ER⁺/PR⁺) tumors received tamoxifen or an aromatase inhibitor at the investigator's discretion. Patients were followed as described in the informed consent after completion of study treatment per standard practice guidelines. Treating physicians were contacted to determine disease status at last clinic visit for analysis of DFS.

Assessment of Endpoints

The target lesion and regional lymph nodes were assessed by clinical examination. Hematologic and biochemical parameters were assessed at every cycle. Patients underwent repeated breast imaging (mammography, ultrasonography, and/or breast magnetic resonance imaging) after 4 cycles and at the completion of chemotherapy to assess response.

Pathologic response was assessed at the time of surgical resection. pCR was defined as the absence of invasive breast cancer in the surgical specimen (breast and axilla) at the time of definitive surgery. Presence of in situ cancer alone was considered a pCR. Complete response (CR) was defined as complete disappearance of all measurable malignant disease by clinical and/or radiologic examination. Partial response was defined as a reduction by at least 30% of the sum of the products of the longest perpendicular diameters of all measurable lesions by clinical and/or radiologic examination. Stable disease for bidimensionally measurable disease was defined as < 25% increase in the sum of the products of the longest perpendicular diameters of all measurable lesions. Progressive disease was defined as a 20% or greater increase in a single lesion, reappearance of any lesions that had disappeared, or clear worsening of any evaluable disease. The OS time was calculated as the time from enrollment to either death if the patient died or last follow-up if the patient is alive. The DFS time is calculated as the time from admission to either recurrence if the patient had recurrence or last follow-up if the patient did not have recurrence.

Statistical Analysis

Interim analysis was planned after 22 subjects had been accrued on each arm, completed treatment, and undergone surgery. Using a Simon 2-stage design and historical data from NSABP B-18, which reported a pCR rate of 13% after 4 cycles of doxorubicin and cyclophosphamide, the study was to be terminated early if 2 or fewer pathologic complete responders were seen in a treatment arm. The study was not powered to compare efficacy between the 2 arms, although pCR could be compared between the arms and with historical controls. The descriptive statistics were used mainly for the analyses. Treatment arm association with OS and DFS was performed by log-rank test. The Kaplan-Meier survival curves are produced for both treatment arms.

All analyses were performed using SAS, version 9.2 (SAS Institute Inc, Cary, NC).

Results

Patient Characteristics

A total of 51 patients were enrolled between April 2007 and July 2009: 25 patients were assigned to the sequential D → C group and 26 patients were assigned to the concurrent DC group. Accrual continued while the 44th subject completed therapy before planned interim analysis. Baseline characteristics were well balanced in the 2 groups (Table 1). Twenty-three patients (45%) were accrued at an academic medical center, 17 patients (33%) were accrued at a county hospital, and 11 patients (22%) were accrued at community sites. Eighteen (35%) of patients accrued were white and 29 (57%) were African-American.

The median baseline tumor size—measured by palpation, mammography, and ultrasonography—was similar between the 2 treatment groups (Table 1). Median tumor size for all patients enrolled was 3.8 cm: In the sequential D → C group, median tumor size was 3.5 cm (range, 1-13 cm) and in the combination DC group, median tumor size was 4.0 cm (range, 1-16 cm). Thirty-six of 51 (71%) enrolled patients had histologically confirmed node-positive disease at baseline: In the sequential D → C group, 19 of 25 (76%) patients had node-positive disease and in the combination DC group 17 of 26 (65%) patients had node-positive disease. Twenty-one of 51 (41%) enrolled patients had triple-negative tumors (ER⁻, PR⁻, and HER2⁻). In the sequential D → C group, 11 of 25 (44%) cases were triple negative; in the combination DC group, 10 of 26 (38%) cases were triple negative. The remainder of patients had ER⁺ and/or PR⁺ disease.

Treatment

Forty-one of 51 patients completed treatment. Eight (15.7%) patients discontinued treatment because of progression of disease (7 in the D → C group and 1 the in DC group). Disease in 5 of the 7 patients in the sequential D → C group progressed while they were receiving single-agent capecitabine. Two patients discontinued treatment because of adverse events. Four of 25 (16%) patients in the sequential arm required a dose reduction of capecitabine. Ten of 26 (38.5%) patients in the combination arm required dose reductions of capecitabine. Dose intensity of capecitabine during treatment (excluding cycles missed because of disease progression) in the sequential arm was 96.5%, and in the combination arm it was 90.1%. Dose intensity for docetaxel was 100% in the sequential arm and 98.2% in the combination arm.

Efficacy

Data from 51 patients were included in analysis of the primary endpoint, the pCR rate (Table 2). One subject from D → C group was excluded from analysis of secondary endpoints of OS and DFS because of the unavailability of long-term follow-up data. Overall, the pCR rate was 10%. pCR was observed in 8% of patients in the sequential D → C group (2/25) and in 12% of patients in the combination DC group (3/26). Analyses of pCR rate according to ER and PR status are shown in Table 2. Among patients with ER⁺/PR⁺disease, the pCR rate was 3% (1/30): 1 of 14 (7%) patients in the sequential D → C arm and 0 of 16 patients in the combination DC arm. Among all patients with triple-negative breast cancer, the pCR rate was 19% (4 of 21): 1 of 11 (9%) of patients in the sequential D → C arm and 3 of 10 (30%) patients in the combination DC arm.

The overall clinical response rate was 75%, with 20 patients (39%) achieving a complete clinical response and 18 patients (35%) having a partial response to therapy. Complete and partial clinical response rates were higher among patients treated in the combination DC arm than among patients treated in the sequential D → C arm (Table 2).

Toxicity

Toxicity data are shown in Table 3. The most common adverse events were fatigue, diarrhea, hand-foot-syndrome, and neuropathy. The grade 3 incidence of diarrhea was 4% for both treatment arms. Grade 3 neuropathy was not seen in the sequential arm; 4% of those treated in the combination arm had grade 3 neuropathy. Grade 3 hand-foot syndrome was observed in 4% of all patients. One patient in the D → C group experienced esophageal obstruction, which was likely unrelated to therapy, and 1 patient in the DC group experienced grade 3 neuropathy and discontinued therapy.

Follow-Up Data on Recurrence and Survival

At a median follow-up of 37 months, 19 patients (76%) in the D → C arm and 21 (81%) patients in the DC arm remained alive and relapse free. Five patients receiving D → C had died and 4 patients receiving DC had died. There were no recurrences among the 5 patients who had a pCR. There were fewer recurrences among patients with ER⁺/PR⁺ breast cancer than among patients with triple-negative breast cancer.

Discussion

The combination of docetaxel and capecitabine resulted in relatively modest pCR rates of 10% in patients with early-stage or locally advanced HER2⁻ breast cancer, regardless of whether patients received the agents concurrently or sequentially. This trial was not powered to compare the concurrent and sequential use of docetaxel and capecitabine and because of the small numbers of patients, no conclusions can be drawn about whether 1 regimen is superior to the other. However the pCR rate in either arm appears consistent with other studies that have used this combination of agents in the neoadjuvant setting. As has been noted in other studies,^{4, 8 and 9} pCR was seen more often in patients with triple-negative breast cancer (19%), compared with hormone receptor-positive breast cancers (3%). The relatively low rate of pCR in this trial may be attributed to the fact that many patients had locally advanced breast cancer and the majority had node-positive disease. The majority of patients had ER⁺/PR⁺ disease, which is known to be associated with a lower rate of pCR. Additionally, the high accrual of African American patients, who have been demonstrated to have worse outcomes from breast cancer,^13 and 14 may have impacted our results. The safety profiles for capecitabine and docetaxel were consistent with previously reported studies.

Patients enrolled in the sequential arm received standard doses of docetaxel at 100 mg/m² and capecitabine at 1000 mg/m² twice daily for 14 days followed by 7 days of rest. The dosing in the DC combination arm was designed so that the total doses of doxorubicin and capecitabine across 8 cycles were equivalent to those received in the sequential arm. Although the total doses and number of cycles were the same, the duration of therapy differed between the 2 groups: subjects on the DC arm received 16 weeks of neoadjuvant therapy compared with 24 weeks of therapy in the D → C arm. One of the limitations of the study is that the combination of docetaxel 50 mg/m² on day 1 with capecitabine 1000 mg/m² by mouth twice daily on days 1 to 7 every 2 weeks is different from the standard dosing of DC in the metastatic setting. More patients experienced disease progression in the sequential arm primarily resulting from progression while receiving single-agent capecitabine. Although the study numbers are too small to draw definitive conclusions regarding the role of single-agent capecitabine in the neoadjuvant setting, both the pCR rate and the rate of breast-conserving surgery were higher in the combination arm. Future trials might consider combining capecitabine with other agents to improve the breast-conservation rate.

Several other neoadjuvant clinical trials have investigated the combination of DC in patients with HER2⁻breast cancer, with pCR rates ranging from 10% to 15%.^{15, 16 and 17} Although comparison of pCR rates across studies is difficult because of differences in dosing and patient populations, these rates are consistent with what we found in our study. Analysis of the GeparQuattro study and the National Surgical Adjuvant Breast and Bowel Project trial B-40 showed that the addition of capecitabine to anthracycline- and taxane-based neoadjuvant chemotherapy did not result in significantly improved pCR rates.^18 and 19Several large clinical trials have investigated the addition of capecitabine to an anthracycline- and taxane-containing regimen in the adjuvant setting. The FinXX trial did show a significant improvement in recurrence-free survival (hazard rate, 0.66; 95% confidence interval [CI], 0.47-0.94; P = .02) when capecitabine was added to docetaxel followed by epirubicin and cyclophosphamide.²⁰ O'Shaughnessy et al recently presented results of a large randomized trial comparing 4 cycles of doxorubicin and cyclophosphamide followed by 4 cycles of docetaxel with or without capecitabine. There was a nonsignificant improvement in 5-year DFS with the addition of capecitabine.²¹

This was the first trial run through the Georgia CORE, a statewide clinical trial network developed in 2003 to allow patients throughout Georgia access to investigator-initiated trials. The inclusion of community sites through the Georgia CORE network as well as a high accrual from a downtown county hospital resulted in the majority of patients enrolled being of African American ethnicity, a group that is underrepresented in oncologic clinical trials. This high accrual of African American patients resulted in more than 40% of the patients enrolled having triple-negative breast cancer. Follow-up of patients after completion of neoadjuvant therapy and surgery was not standardized, limiting the interpretation of long-term outcomes. Interestingly, although the overall pCR rate in this trial was low, with a median follow-up of 37 months, more than three quarters of patients remain alive and disease free. DFS was better among patients with ER⁺/PR⁺ breast cancer compared with patients with triple-negative cancers despite a higher pCR rate in the latter population (Figure 1). Similar to other analyses, we found that pCR was a more robust prognostic factor in triple-negative breast cancers than in ER⁺ cancer; however patients who had a pCR did better regardless of subtype (Figure 1).^{4, 8 and 9}

A significant limitation of this trial is the lack of available baseline tissue to perform molecular studies to correlate with response or nonresponse. It was difficult to obtain sufficient baseline tissue from community sites that enrolled patients in this trial for analysis. Additionally, many of the patients were referred to the academic center after their diagnostic biopsy was performed elsewhere. Another significant issue is the ability to obtain baseline fresh tissue. An ongoing protocol consists of consenting patients who have undergone diagnostic breast biopsies to undergo an additional biopsy to collect fresh tissue. Additionally, residual tissue after preoperative chemotherapy is collected.

Conclusion

Results from this and other clinical trials have not clearly established a role for capecitabine in addition to standard anthracycline- and taxane-based chemotherapy for patients with early-stage or locally advanced breast cancers. Future neoadjuvant trials are focusing on specific breast cancer subtypes, given the difference and prognostic importance of pCR. The acquisition of baseline and fresh tissue is essential to allow tailoring of therapies based on breast cancer phenotype. Models using residual disease after neoadjuvant chemotherapy are critical so that novel therapeutic approaches to overcome resistance can be developed.

Clinical Practice Points

•

Neoadjuvant chemotherapy is the standard of care for patients with inoperable, locally advanced, or inflammatory breast cancer and can be considered a reasonable treatment option for patients with earlier stage breast cancer for whom adjuvant chemotherapy is recommended.

•

pCR to neoadjuvant therapy at the time of surgery is a strong prognostic factor for long-term outcome in certain subtypes of breast cancer (ER⁻ breast cancers).

•

The addition of capecitabine either concurrent with or sequentially after docetaxel is feasible and appears to be of equal efficacy.

•

The pCR rates were higher among patients with hormone receptor–negative than among those with hormone receptor–positive tumors.

Disclosure

Dr O'Regan has received a grant for research from Genentech. The other authors have stated that they have no conflicts of interest.

References

Copyright © 2013 Elsevier Inc. All rights reserved.