The Oncologist, Vol. 10, No. 4, 242-249, April 2005; doi:10.1634/theoncologist.10-4-242 © 2005 AlphaMed Press
Pathological and Clinical Response of a Primary Chemotherapy Regimen Combining Vinorelbine, Epirubicin, and Paclitaxel as Neoadjuvant Treatment in Patients with Operable Breast Cancera Centre Jean Perrin, Clermont-Ferrand, France; b INSERM U484, Clermont-Ferrand, France; c Centre Paul Papin, Angers, France; d Centre Hospitalier Général, Brive-la-Gaillarde, France; e Clinique des Dômes, Clermont-Ferrand, France Correspondence: Catherine Abrial, M.D., Centre Jean Perrin, Bureau de Recherche Clinique, 58 rue Montalembert, BP 392, 63011 Clermont-Ferrand Cedex 1, France. Telephone: 33-4-73-27-80-05; Fax: 33-4-73-27-80-29; e-mail: recherche.clinique{at}cjp.fr
This phase II study investigated the efficacy and tolerability of a primary chemotherapy regimen combining vinorelbine, epirubicin, and paclitaxel (VEP protocol) in women with stage II/III operable breast cancer. Patients (n = 50) were treated with six cycles of VEP according to the following schedule: vinorelbine (Navelbine®; Pierre Fabre, Boulogne, France; http://www.pierre-fabre.com) 20 mg/m2, epirubicin (Farmorubicin®; Pharmacia, New York, NY; http://www.pnu.com) 35 mg/m2 given on days 1 and 8, paclitaxel (Taxol®; Bristol-Myers Squibb, New York, NY; http://www.bmsoncology.com) 175 mg/m2 given on day 9, and G-CSF 5 mg/kg/day given on days 1020 of a 21-day cycle, followed by surgery and radiotherapy. After six cycles of VEP, the pathological response rate (pCR) in breast was confirmed in six patients (12%; 95% confidence interval [CI]: 321)) using Chevalliers classification and in nine patients (18%; 95% CI: 7.428.6) using Sataloffs classification. The clinical response rate was 42% (95% CI: 28.355.7), including 26% complete responses. Breast conservation was achieved in 68% of patients. After a median follow-up of 48 months (range, 3462 months), 16 relapses were observed. The overall and disease-free survivals at 5 years were 54.1% (95% CI: 40.367.9) and 38% (95% CI: 24.151.9), respectively. The principal toxicities of VEP were grade 3/4 neutropenia observed in 30% of patients and grade 3 anemia observed in 12% of patients. There was no case of severe cardiac toxicity, thrombocytopenia, or any other serious adverse events. In conclusion, whereas this regimen was relatively well tolerated, it appears inferior to other regimens and its use is not recommended. Key Words. Breast cancer • Neoadjuvant therapy • Vinorelbine • Epirubicin • Paclitaxel
Neoadjuvant chemotherapy has been used to treat inflammatory and locally advanced breast carcinoma for the past three decades, first to improve patient survival. More recently this strategy has been extended to the management of patients with operable disease who are eligible for mastectomy, mainly in order to increase the rate of breast conservation [1]. It allows regression of the tumor in order to avoid mastectomy (in about 80% of cases) and to treat clinically undetectable micrometastatic disease. In addition, neoadjuvant chemotherapy permits the assessment of response of the primary tumor to a particular chemotherapy regimen and provides an early opportunity to change the agents if the tumor appears clinically resistant. The appropriate selection of drugs for neoadjuvant treatment must reflect the need to use agents likely to achieve rapid control of the underlying disease. The most active groups of cytotoxic compounds in this setting include anthracyclines and taxanes, which have demonstrated high response rates in advanced disease [24]. Previous experience with other anthracycline-based combinations has resulted in responses that subsequently permitted breast conservation in 85% of patients [5]. At Centre Jean Perrin, neoadjuvant chemotherapy in patients with operable breast cancer has been evaluated in prospective clinical trials. The first protocol was AVCF/M (doxorubicin, vincristine, cyclophosphamide, 5-fluorouracil ± methotrexate), which had a 57.5% objective response (OR) rate and a 5.7% pathological complete response (pCR) rate [5]. Because vinorelbine and epirubicin have high response rates (respectively, 50% and 45%) in metastatic breast carcinoma [6], it seemed logical to combine these two compounds with the well-tolerated methotrexate (20% responses in metastatic breast carcinoma). Thus, the NEM protocol (vinorelbine 25 mg/m2, epirubicin 35 mg/m2, methotrexate 20 mg/m2 given on days 1 and 8 of a 28-day cycle) began in 1991 (OR = 90% and pCR = 14%) [7]. Next, a more intensive regimen was proposed: the TNCF protocol (theprubicin, vinorelbine, cyclophosphamide, 5-fluorouracil; OR = 79.7% and pCR = 33.3%) [8]. Later, with the arrival of the taxanes, which present a high response rate (OR = 40%60%) in metastatic breast carcinoma [9] that is equal to or better than anthracyclines, the Taxobel protocol, with docetaxel alone was started (OR = 70.8% and pCR = 19.8%) [10]. The VEP protocol, based on NEM, was started a year later, with methotrexate replaced by paclitaxel. Thus, the VEP protocol has combined the three cytotoxic drug families considered at present as most efficacious in breast carcinoma treatment: vinorelbine, anthracyclines, and taxanes. Because the pCRs obtained in NEM and Taxobel protocols were relatively high (respectively, 14% and 19.8%), the pCR rate in the VEP protocol was evaluated as the primary end point. pCR is now recognized as a prognostic factor in patient disease-free and overall survival, and a minimum value of 20% of class 1/2 according to the Chevallier classification was expected [11, 12]. The present study evaluated the clinical response rate, breast conservation rate, survival, and toxicity as secondary end points.
Patient Population In this phase II trial, neoadjuvant chemotherapy was administered to 53 patients with operable stage II/III breast carcinoma. The eligibility criteria for this study were: 1870 years of age, pathological proof by true cut biopsy of invasive adenocarcinoma, no metastatic spread, tumor 3 cm or <3 cm but located in the central area, no prior specific treatment, no prior history of heart disease (confirmed by an electrocardiography), adequate biological functions, and written informed consent. The study was conducted in accordance with the Hong Kong Amendment of the Declaration of Helsinki and approved by the Ethics Committee of Auvergne before commencement. The baseline work-up included a complete history and clinical examination, bilateral mammography, bilateral breast and axillary ultrasound, percutaneous cytology of lymph node, and pathological diagnosis by core biopsy. When an invasive adenocarcinoma was demonstrated, the tumor was evaluated with Scarff-Bloom-Richardson (SBR) grading and hormone receptors were assayed by radioimmunology. Laboratory assessment consisted of complete blood cell count, blood chemistry analysis, and tumor markers. The absence of distant metastases was confirmed by chest x-ray, bone scan, and liver ultrasound.
Treatment Modalities
After induction chemotherapy, patients underwent appropriate surgery according to the size of their residual tumor. A conservative surgery was performed if the residual tumor size was <3 cm, or a modified radical mastectomy if the tumor size was
Assessment of Response As eight patients did not receive the entire treatment because of allergy or toxicity, the clinical, mammographic, and ultrasound response rates have been calculated for the intent-to-treat ([ITT] 50 patients) and per-protocol ([PP] 42 evaluable patients) populations. The pathological response was independently evaluated after surgical resection of the remaining tumor and nodes. Pathologic response was classified as follows, according to the Chevallier et al. [11] classification: class 1, disappearance of all tumor either on macroscopic or microscopic assessment; class 2, presence of in situ carcinoma in the breast, no invasive tumor, and no tumor found in the lymph nodes; class 3, presence of invasive carcinoma with stromal alteration, such as sclerosis or fibrosis; or class 4, few modifications of the tumoral appearance. Pathologic response was also classified according to the Sataloff et al. [14] classification: primary site response: T-A) total or near total therapeutic effect; T-B) subjectively >50% therapeutic effect but less than total or near total; T-C) >50% therapeutic effect, but effect evident; T-D) no therapeutic effect; and axillary lymph node response: N-A) evidence of therapeutic effect, no metastatic disease; N-B) no nodal metastasis or therapeutic effect; N-C) evidence of therapeutic effect but nodal metastasis still present; or N-D) viable metastasis disease, no therapeutic effect. Among the eight patients for whom treatment was discontinued for allergy and toxicity, four of them (8%) received other chemotherapy before surgery. Thus, the pCRs have been calculated for the ITT (50 patients) and PP (46 evaluable patients) population.
Statistical Analysis
Patient Characteristics The median age was 49 years (range, 3168 years) and 21 (42%) of these women were postmenopausal. The distribution of clinical involvement showed that all the patients had tumors >30 mm. The median clinical diameter of the primary tumor was 47 mm (range, 3090 mm). The pathological and laboratory characteristics of the tumors are shown in Table 1
Treatment Management Forty-one (82%) patients received all the planned induction therapy, while in nine patients, treatment was discontinued after one, two, three, or five cycles for allergy (five patients), toxicity (three patients), and progression (one patient). A median number of six cycles of treatment (range, 16 cycles) were administered. The median dose (per cycle) of vinorelbine was 38.8 mg/m2 (range, 8.647.1 mg/m2), the median dose of epirubicin was 68.8 mg/m2 (range, 14.189.1 mg/m2), and the median dose of paclitaxel was 174 mg/m2 (range, 34.6180.6 mg/m2). The median dose-intensity was 12.7 (range, 8.4014.80) for vinorelbine, 22.7 (range, 16.9028.10) for epirubicin, and 57.3 (range, 32.2065.80) for paclitaxel.
Tolerability of Neoadjuvant Chemotherapy
Nonhematological toxicity consisted of grade 3 alopecia, which occurred in 52% of patients (40% of cycles). Two cases of grade 3 digestive toxicity (4% of patients) and one case of grade 3 neurologic toxicity (2% of patients) were also reported. No cardiac toxicity and no toxic death were reported.
Clinical Response
As shown in Table 4
Histopathological Evaluation For the ITT population (Table 5
For the PP population (Table 6
The median histological diameter of the residual tumor was 13 mm (range, 080 mm). Of the 50 patients, 10 underwent a modified radical mastectomy, 33 patients had conservative surgery with axillary dissection, one patient had a mastectomy (without axillary dissection), and one patient received conservative surgery only. These last two patients underwent primary axillary dissection. For one patient, the data concerning the surgery were not available and the remaining four patients were not evaluable because they received other chemotherapy before surgery. The surgical margins were involved for four patients and led to a secondary mastectomy.
At surgery, the median number of nodes collected was 12 (range, 328) and the median number of involved nodes was 2 (range, 114). Among 43 patients who underwent axillary dissection, 19 (44%) were N0, 17 (40%) were N13, 4 (9%) were N4-9, and 3 (7%) were N Among the 43 patients who underwent axillary dissection, pathological evidence of node involvement was found in 8 of 16 patients without clinical node disease (50% of N0), and in 16 of 27 patients with palpable nodes (59.3% of N12). Moreover, pathological evidence of node involvement was found in 5 of 13 patients for whom percutaneous cytology was positive (38.5%) and in seven of nine patients for whom percutaneous cytology was negative (77.8%). After surgery, 46 (92%) patients received radiotherapy, two patients did not receive radiotherapy, and data were missing for the last two last patients. Subsequently, three patients (6%) received adjuvant chemotherapy, 45 patients did not receive adjuvant treatment, and the data were missing for the last two patients. Thirty (60%) postmenopausal patients received hormone therapy with tamoxifen for 5 years.
Follow-up
The aim of the present study was to assess as primary end point the pCR rate and as secondary end points the clinical, mammographic, and echographic response rate; the breast conservation rate; disease-free survival; overall survival; and toxicity. The overall pCR rates (in the ITT population) were 12% and 18% according to the Chevallier and Sataloff classifications, respectively. The pCR, according to Chevallier, was comparable with the rate obtained with the NEM protocol (14%) [7] and was lower than the expected 20%, but the difference between these two values was not significant (p = 0.35). However, these results are lower than those obtained with the Taxobel protocol [10]: 19.8% and 35.5% according to the Chevallier and Sataloff classifications, respectively. Moreover, our results according to the Chevallier classification agree with the prospective phase III randomized trial of Green et al. [16]. This trial compared the pCR rates of two groups of patients who received either weekly paclitaxel followed by FAC (5-fluorouracil, doxorubicin, cyclophosphamide) or paclitaxel every 3 weeks followed by FAC, as primary systemic chemotherapy of operable breast carcinoma. The analysis of the results showed that 28.8% of patients who received weekly paclitaxel had a pCR versus 13.6% of patients treated with paclitaxel every 3 weeks.
When we consider the secondary end points, the results of the study reported an overall clinical response rate of 42%, with 26% complete responses (in ITT population). The clinical CR (26%) was no better than and sometimes inferior to the results obtained with TNCF (51%) [8], Taxobel (36%) [10], AVCF (27%) [17], and NEM (28%) [7] protocols. Besides, there were 48% fewer clinical CRs for invasive lobular carcinoma than for invasive canalar carcinoma. Furthermore, as it is mentioned in the National Surgical Adjuvant Breast and Bowel Project trial B-18 [18], there is a correlation between tumor size and clinical complete response rate; 64.7% of T2 tumors achieved a clinical CR versus of 50% of T3 tumors. No T4 tumor achieved a clinical CR. As mentioned in the literature [1], estrogen receptor (ER)-negative disease and high nuclear grade are associated with a better CR rate. Seventy-eight percent of ER-negative disease achieved a clinical complete response versus 38% of ER-positive disease, and 70% of SBR grade III tumors achieved a clinical complete response versus 43% of SBR grade I/II tumors. The breast conservation rate was 68%, which was comparable with the rates obtained with Taxobel (72%) [10], AVCF (69%) [17], and TNCF (71%) [8] protocols, but it was lower than in the NEM protocol (87%) [7]. Moreover, as it is reported by Newman et al [19], patients with lobular carcinomas are less likely to become candidates for breast-conserving therapy after neoadjuvant chemotherapy. Indeed, there was 3.8% less conservative surgery for invasive lobular carcinomas than for invasive canalar carcinomas. For the 11 patients who underwent a mastectomy, the size of the residual tumor was still high ( In March 2004, the median follow-up was 48 months (range, 3462 months), thus it is still early to draw definitive conclusions about the outcome of these patients. Nevertheless, we have calculated the overall and disease-free survivals at 5 years, which were of 54.1% and 38%, respectively. There is presently a tendency for a better disease-free survival in patients with pCR at surgery (Chevalliers grades 1 and 2, Sataloffs grade A), however still insignificant due to the total number of patients and limited follow-up [12]. Of the 16 patients who had relapsed, only one (6%) was classified as pCR according to Chevalliers classification. Hematological and nonhematological toxicity was assessed. Grade 3/4 neutropenia was observed in 30% of patients, which is the same as in the NEM protocol [7]. However, this is much lower than in Taxobel [10], which reported grade 3/4 neutropenia in 70.5% of patients. Eight patients (16%) developed febrile aplasia, and grade 3 anemia was also recorded in six patients (12%). Nonhematological toxicity mainly consisted of grade 3 alopecia in 52% of patients. Treatment was stopped for eight patients for allergy and toxicity (dyspnea, neurotoxicity, and myalgia) after one, two, or three cycles. In conclusion, the VEP protocol showed a relatively low clinical response rate (42%), which is globally inferior to the results in the literature and to the results of the epirubicin/docetaxel (ED) arm (84%) in the randomized phase II trial of GIREC S01 [20]. Concerning toxicity results, the VEP protocol was well tolerated and apparently better than the ED arm in GIREC S01. Grade 3/4 neutropenia was observed in 14% of cycles in VEP versus 71% in GIREC S01 study. The breast conservation rate was comparable with rates mentioned in the literature and to that reported in GIREC S01. It is still early to compare the disease-free survival and the overall survival with other neoadjuvant treatments. Moreover, the overall pCR rate (12%), which was lower than the expected 20%, was no better than in the NEM protocol (14%) and was lower than in GIREC S01 (24%). This small pCR rate and the low clinical response rate could be explained by the sequential instead of simultaneous infusion of epirubicin and paclitaxel (in the ED arm of the GIREC S01, epirubicin and docetaxel infusions were simultaneous), but also by predictive factors of chemotherapy response (small size of the tumor, invasive canalar carcinoma, SBR III grade, and ER-negative disease). Indeed, the median clinical diameter of the primary tumor was relatively high (47 mm; range, 3090 mm), there was a non-negligible proportion of invasive lobular carcinoma (24%), and SBR III grade and ER-negative tumors were not a majority (41% and 36%, respectively). Thus, while this protocol was well tolerated, a phase III study with this combination has not been initiated.
The authors indicated no potential conflicts of interest.
This research was supported in part by Bristol-Myers Squibb and Pierre Fabre Pharmaceuticals. We wish to thank Dr. C. Bonnel (CHU Saint André, Bordeaux), F. Kwiatkowski, Dr. G. Le Bouëdec (Pr. J. Dauplat), Dr. V. Feillel (Dr. S. Lemery), and Dr. J.L. Achard (Pr. P. Verrelle), all from Centre Jean Perrin, Clermont-Ferrand.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||