The

n e w e ng l a n d j o u r na l

of

m e dic i n e

original article

Adjuvant Docetaxel for High-Risk, NodeNegative Breast Cancer

From the Hospital General Universitario Gregorio Marañón, Madrid (M. Martín); Corporacion Sanitaria Parc Taulí, Sabadell (M.A.S., A.S.); Hospital Universitario Miguel Servet (A. Antón, M.J.R.) and Hospital Clínico Universitario Lozano Blesa (D.I.) — both in Zaragoza; Instituto Valenciano de Oncología (A.R., S.A.), Hospital La Fe (B.M.), and Hospital Clinico de Valencia (A.L.) — all in Valencia; Centro Oncológico de Galicia (M.R.) and Complejo Hospitalario de la Coruña (L.C.) — both in Corunna; Hospital General Universitario de Alicante, Alicante (E.A., I.A., G.P.); Hospital Universitario de Elche, Elche (A.R.-L.); Hospital Germans Trias i Pujol, Badalona (A.B.); Hospital de Basurto, Bilbao (P.M.P.); Hospital Virgen del Rocío, Seville (M.R.B.); Consorci Sanitari de Terrassa, Terrassa (A. Arcusa); Hospital Clinic i Provincial, Barcelona (M.M.); Hospital Marqués de Valdecilla, Santander (J.M.L.V.); Complejo Hospitalario Xeral Calde, Lugo (J.R.M.); Hospital General de Elda, Elda (C.L.); Fundación Hospitalaria de Alcorcón, Alcorcón (C.J.); Hospital Clínico Virgen de la Victoria, Malaga (E.A.); Hospital Comarcal de Barbastro, Huesca (J.F.); and Hospital Universitario Príncipe de Asturias, Alcalá de Henares (J.A.L.G.-A.) — all in Spain; Klinik und Poliklinik Gynäkologie Martin-Luther-Universitäts­ klinik, Halle, Germany (R.G.); Wielkopolskie Centrum Onkologii, Poznan, Poland (J.Z.); and Sanofi-Aventis, Bridgewater, NJ (J.L.). Address reprint requests to Dr. Martín at Servicio de Oncologia Medica, Hospital General Universitario Gregorio Marañón, Universidad Complutense, Dr Esquerdo 46, Madrid 28007, Spain, or at [email protected]. *Deceased. †Additional investigators who participated in the Spanish Breast Cancer Research Group (GEICAM) study are listed in the Supplementary Appendix, available at NEJM.org. This article (10.1056/NEJMoa0910320) was updated on February 23, 2011, at NEJM.org. N Engl J Med 2010;363:2200-10. Copyright © 2010 Massachusetts Medical Society.

2200

Miguel Martín, M.D., Ph.D., Miguel A. Seguí, M.D., Antonio Antón, M.D., Ph.D., Amparo Ruiz, M.D., Manuel Ramos, M.D., Encarna Adrover, M.D., Ignacio Aranda, M.D., Alvaro Rodríguez-Lescure, M.D., Ph.D., Regina Große, M.D., Lourdes Calvo, M.D., Agustí Barnadas, M.D., Ph.D., Dolores Isla, M.D., Ph.D., Purificación Martinez del Prado, M.D., Manuel Ruiz Borrego, M.D., Jerzy Zaluski, M.D.,* Angels Arcusa, M.D., Montserrat Muñoz, M.D., José M. López Vega, M.D., Ph.D., José R. Mel, M.D., Ph.D., Blanca Munarriz, M.D., Ph.D., Cristina Llorca, M.D., Ph.D., Carlos Jara, M.D., Ph.D., Emilio Alba, M.D., Ph.D., Jesús Florián, M.D., Junfang Li, Ph.D., José A. López García-Asenjo, M.D., Amparo Sáez, M.D., María José Rios, M.D., Sergio Almenar, M.D., Gloria Peiró, M.D., and Ana Lluch, M.D., Ph.D., for the GEICAM 9805 Investigators†

A BS T R AC T BACKGROUND

A regimen of docetaxel, doxorubicin, and cyclophosphamide (TAC) is superior to a regimen of fluorouracil, doxorubicin, and cyclophosphamide (FAC) when used as adjuvant therapy in women with node-positive breast cancer. The value of taxanes in the treatment of node-negative disease has not been determined. METHODS

We randomly assigned 1060 women with axillary-node–negative breast cancer and at least one high-risk factor for recurrence (according to the 1998 St. Gallen criteria) to treatment with TAC or FAC every 3 weeks for six cycles after surgery. The primary end point was disease-free survival after at least 5 years of follow-up. Secondary end points included overall survival and toxicity. RESULTS

At a median follow-up of 77 months, the proportion of patients alive and diseasefree was higher among the 539 women in the TAC group (87.8%) than among the 521 women in the FAC group (81.8%), representing a 32% reduction in the risk of recurrence with TAC (hazard ratio, 0.68; 95% confidence interval [CI], 0.49 to 0.93; P = 0.01 by the log-rank test). This benefit was consistent, regardless of hormonereceptor status, menopausal status, or number of high-risk factors. The difference in survival rates (TAC, 95.2%; FAC, 93.5%) was not significant (hazard ratio, 0.76; 95% CI, 0.45 to 1.26); however, the number of events was small (TAC, 26; FAC, 34). Rates of grade 3 or 4 adverse events were 28.2% with TAC and 17.0% with FAC (P5 cm). and had negative axillary lymph nodes (≥10 nodes examined). Patients were also required to meet Assessments one or more of the high-risk criteria described in Baseline assessments (bilateral mammography, the 1998 St. Gallen consensus recommendations chest radiography, electrocardiography, abdominal for patients with operable node-negative breast ultrasonography or computed tomography, bone cancer (tumor size >2 cm, negative results on tests scanning, and bone radiography in the case of for expression of estrogen receptor and proges- suspicious lesions on the bone scan) were perterone receptor, tumor histologic grade 2 or 3, or formed a maximum of 12 weeks before enrollage 5 cm Unknown

0

1 (0.2)

Tumor grade — no. (%) Grade 1

38 (7.1)

34 (6.5)

Grade 2

216 (40.1)

230 (44.1)

Grade 3

259 (48.1)

231 (44.3)

26 (4.8)

26 (5.0)

Premenopausal

285 (52.9)

272 (52.2)

Postmenopausal

254 (47.1)

249 (47.8)

Positive

344 (63.8)

349 (67.0)

Negative

192 (35.6)

170 (32.6)

Unknown

3 (0.6)

2 (0.4)

287 (53.2)

247 (47.4)

24 (4.5)

24 (4.6)

22 (4.1)

20 (3.8)

206 (38.2)

230 (44.1)

Unknown Menopausal status — no. (%)‡

Hormone-receptor status — no. (%)§

Breast-conserving surgery — no. (%) With radiation therapy Without radiation therapy Mastectomy — no. (%) With radiation therapy Without radiation therapy

* FAC denotes fluorouracil, doxorubicin, and cyclophosphamide; and TAC docetaxel, doxorubicin, and cyclophosphamide. † The upper age limit was not meant to be exclusionary but was based on the lack of safety data on the TAC regimen for women older than 70 years of age. Four patients in the FAC group and 1 patient in the TAC group were more than 70 years old. ‡ Patients were classified as premenopausal if they had not had a hysterectomy, were menstruating within 6 months before randomization, or had had a hysterectomy and were younger than 40 years of age or were 40 years of age or older with premenopausal levels of luteinizing hormone and follicle-stimulating hormone. Patients were classified as postmenopausal or perimenopausal if they had not had a hysterectomy, had not menstruated within 6 months before randomization, or had had a hysterectomy and were either older than 55 years of age or 55 years old or younger with postmenopausal levels of luteinizing hormone and follicle-stimulating hormone. § The hormone-receptor status of the primary tumor was assessed at local laboratories before randomization. Positive hormone-receptor status refers to breast cancer that was positive for expression of the estrogen receptor, the progesterone receptor, or both.

n engl j med 363;23  nejm.org  december 2, 2010

The New England Journal of Medicine Downloaded from nejm.org on January 28, 2017. For personal use only. No other uses without permission. Copyright © 2010 Massachusetts Medical Society. All rights reserved.

2203

The

n e w e ng l a n d j o u r na l

Table 2. First Events According to the Intention-to-Treat Analysis.* Event

TAC (N = 539)

FAC (N = 521)

no. of patients (%) Any event

66 (12.2)

95 (18.2)

Local

4 (0.7)

17 (3.3)

Regional

0

4 (0.8)

Local and regional

3 (0.6)

2 (0.4)

Distant and local

4 (0.7)

2 (0.4)

Relapse of breast cancer

Distant and regional Distant only Unknown

2 (0.4)

3 (0.6)

34 (6.3)

38 (7.3)

1 (0.2)

0

Second primary cancer Contralateral breast

4 (0.7)

10 (1.9)

Endometrium

1 (0.2)

4 (0.8)

Ovary

2 (0.4)

1 (0.2)

Other

6 (1.1)

10 (1.9)

Unknown

0

1 (0.2)

0

0

From breast cancer†

1 (0.2)

0

From other causes

4 (0.7)

3 (0.6)

Death, without evidence of relapse From toxic effects of treatment

* FAC denotes fluorouracil, doxorubicin, and cyclophosphamide; and TAC docetaxel, doxorubicin, and cyclophosphamide. † Site investigators reported deaths from breast cancer, but there was no diagnostic confirmation of relapse.

ment (8 in the TAC group and 1 in the FAC group). Four patients did not receive the treatment to which they had been randomly assigned (3 in the TAC group and 1 in the FAC group). A total of 1051 patients could be evaluated for safety (532 in the TAC group and 519 in the FAC group). Baseline characteristics were well balanced between the treatment groups (Table 1). In each group, more than 96% of patients with hormonereceptor–positive tumors received tamoxifen or another hormonal therapy. The majority of treated patients received the 6 cycles of chemotherapy specified in the protocol (94.5% of patients in the TAC group and 97.7% in the FAC group). Rates of treatment compliance have been published previously.15 Efficacy Analysis

By the time of the cutoff date for the efficacy analysis (March 4, 2009), 161 events (defined as relapse, new primary cancer, or death with no 2204

of

m e dic i n e

evidence of relapse) had been reported (66 in the TAC group and 95 in the FAC group) (Table 2). At a median follow-up of 77 months (95% confidence interval [CI], 74 to 79), 473 patients in the TAC group (87.8%) and 426 patients in the FAC group (81.8%) had no events and remained free of disease. This corresponds to a 32% reduction in the risk of having an event for the TAC group (hazard ratio, 0.68; 95% CI, 0.49 to 0.93; P = 0.01) (Fig. 1A). The TAC regimen was associated with fewer local and distant relapses — a trend that cannot be explained by an imbalance between the groups in the proportion of patients who received radiation therapy (Table 1 in the Supplementary Appendix). The estimated disease-free survival rate at 5 years was 90.1% in the TAC group as compared with 85.3% in the FAC group, with a hazard ratio of 0.67 (95% CI, 0.47 to 0.96; P = 0.03). At the cutoff date, 60 patients (5.7%) had died (26 in the TAC group and 34 in the FAC group); 95.2% of patients in the TAC group and 93.5% of those in the FAC group were alive. This corresponds to a 24% reduction in the risk of death among patients who received TAC (hazard ratio, 0.76; 95% CI, 0.45 to 1.26; P = 0.29 on the basis of an unstratified log-rank test) (Fig. 1B). Subgroup Analyses

A planned analysis of disease-free survival in subgroups defined by hormone-receptor status, menopausal status, and number of high-risk factors (with high-risk factors defined as age, tumor size, and histologic grade) suggests that the benefit of TAC as compared with FAC in each subgroup was consistent with the benefit in the overall study population (Fig. 2). Analysis of the consistency of disease-free survival across all predefined subgroups with the use of the Cox model did not reveal an interaction between treatment effect and any subgroup characteristic (P>0.10), including use or nonuse of radiation therapy (P = 0.11) and type of surgery (P = 0.14). Tumors in 608 patients (319 in the TAC group and 289 in the FAC group) were centrally evaluated for both hormone-receptor status and HER2 status. These patients had characteristics and outcomes that were similar to those of the total population (data not shown). Of these 608 patients, 13.6% had HER2-positive tumors and 66.4% had hormone-receptor–positive tumors. In the centrally evaluated subpopulation of patients with HER2-negative tumors (without regard to hormone-receptor status; 280 in the TAC group

n engl j med 363;23  nejm.org  december 2, 2010

The New England Journal of Medicine Downloaded from nejm.org on January 28, 2017. For personal use only. No other uses without permission. Copyright © 2010 Massachusetts Medical Society. All rights reserved.

Adjuvant Docetaxel for Node-Negative Breast Cancer

A 100

TAC (N=539)

90

Disease-free Survival (%)

and 245 in the FAC group), there were 33 events in the TAC group (11.8%) and 53 in the FAC group (21.6%), suggesting an advantage of TAC over FAC (Fig. 3). However, there was no significant treatment interaction with either HER2 status (P = 0.39) or hormone-receptor status (P = 0.28). Disease-free survival in the two HER2-positive subgroups was similar for patients receiving TAC and those receiving FAC, although the numbers of patients in these subgroups were very small.

80

FAC (N=521)

70 60 50 40 30 20 P=0.01

10

Safety

0

Discussion The GEICAM 9805 trial showed that in women with operable, high-risk, node-negative, early-stage breast cancer, adjuvant TAC significantly improved disease-free survival (the primary end point) as compared with FAC, with a 32% reduction in the risk of recurrence (P = 0.01) at a median follow-up of 77 months. The benefit of TAC was consistent, regardless of hormone-receptor status, menopausal status, or the number of highrisk factors. Because of the limited number of deaths, survival data are still immature. The pro-

0

6

12

18

24

30

36

42

48

54

60

66

72

78

84

Months since Randomization No. at Risk FAC TAC

521 517 508 498 488 480 474 466 453 445 429 361 284 215 163 539 529 522 509 501 494 486 480 475 472 460 360 299 242 175

B

Overall Survival (%)

Safety was assessed in 1051 treated patients (532 in the TAC group and 519 in the FAC group). The numbers of grade 2, 3, and 4 toxic effects have been reported previously.15 Rates of grade 3 and 4 adverse events that occurred during the study treatment were 28.2% for patients treated with TAC and 17.0% for those treated with FAC (P