|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Journal of Clinical Oncology, Vol 23, No 9 (March 20), 2005: pp. 1934-1940 © 2005 American Society of Clinical Oncology. DOI: 10.1200/JCO.2005.04.032 Sequencing of Chemotherapy and Radiation Therapy in Early-Stage Breast Cancer: Updated Results of a Prospective Randomized TrialFrom the Departments of Radiation Oncology and Medical Oncology, Brigham and Women's Hospital; Department of Biostatistics, Dana-Farber Cancer Institute; Harvard Medical School; Departments of Radiation Oncology and Medicine, Beth Israel Deaconess Medical Center; and Harvard School of Public Health, Boston, MA Address reprint requests to Jennifer R. Bellon, MD, Department of Radiation Oncology, Dana-Farber Cancer Institute, Brigham and Women's Hospital, 75 Francis St, Boston, MA 02115; e-mail: jbellon{at}lroc.harvard.edu
PURPOSE: The optimal integration of chemotherapy with radiation (RT) for patients with early-stage breast cancer remains uncertain. We present the long-term results of a prospective randomized trial to address this question. PATIENTS AND METHODS: Two hundred forty-four patients were randomly assigned after conservative breast surgery to receive 12 weeks of cyclophosphamide, doxorubicin, methotrexate, fluorouracil, and prednisone (CAMFP) before RT (CT-first) or after RT (RT-first). Median follow-up for surviving patients was 135 months. RESULTS: There were no significant differences between the CT-first and RT-first arms in time to any event, distant metastasis, or death. Sites of first failure were also not significantly different. CONCLUSION: Among breast cancer patients treated with conservative surgery, there is no advantage to giving RT before adjuvant chemotherapy. However, this study does not have enough statistical power to rule out a clinically important survival benefit for either sequence.
The optimal way to integrate chemotherapy and radiation therapy after breast-conserving surgery for patients with early-stage breast cancer remains unknown. Several retrospective reviews have shown increased rates of local-regional recurrence when radiation therapy is delayed after surgery.1–4 This has not been a uniform finding, however, with other retrospective series reporting no increased risk of local recurrence when radiation is delayed in order to administer chemotherapy.5–7 There is also concern that delaying chemotherapy in order to give radiation may increase the risk of distant metastasis and ultimately affect survival.8,9 The validity of the conclusions of these studies, however, is uncertain because of biases in treatment selection and heterogeneity of patient and tumor characteristics. In 1984, we began a prospective trial (designed by S.E.C., I.C.H., R.S.G., and J.R.H.) to determine the influence of the sequencing of radiation and chemotherapy after conservative surgery for women with early-stage breast cancer on doses of chemotherapy delivered, cosmetic outcome, and disease-free survival. The initial results of this study were published in 1996 with a median follow-up time of 58 months.10 At that time, the difference in the pattern of sites of first recurrence approached significance (P = .07). Patients on the radiotherapy-first arm had a 5-year crude rate of local recurrence (with or without simultaneous distant or regional nodal failures) of 5%, compared with 14% for patients on the chemotherapy-first arm. The 5-year crude rates of distant and/or regional recurrence (without simultaneous local failure as the site of first failure) were 32% in the radiotherapy first (RT-first) and 20% in the chemotherapy first (CT-first) arms. Contralateral breast cancers occurred first in 3% on the RT-first arm and 8% of the CT-first patients; nonbreast second malignancies occurred first in 1% of patients in each arm. Although the difference in overall survival rates did not reach statistical significance, the time to distant metastasis was significantly different between the two arms (P = .05), with 5-year actuarial rates of distant metastases of 36% in the RT-first arm and 25% in the CT-first arm. No differences in late complications were seen between the two groups.11 We now report the updated results of this trial with a median follow-up time of 135 months in surviving patients.
Two hundred forty-four women were enrolled between June 1984 and December 1992. The protocol was approved by the institutional review board of each participating hospital. Informed consent was obtained from all patients. After lumpectomy and axillary nodal dissection, patients were randomly assigned to receive chemotherapy followed by radiation therapy (designated the CT-first arm; n = 122) or radiation therapy followed by the same chemotherapy regimen (designated as the RT-first arm; n = 122). Adjuvant treatment was mandated to begin within 6 weeks of the last breast surgery, but 44 patients (14 in the CT-first and 30 in the RT-first arm) had a longer lag time (48 to 85 days). Eligibility requirements included stage I or II breast cancer. Initially, only node-positive patients were included and no patient was to receive tamoxifen. Randomization was stratified by number of histologically positive axillary nodes (one to three, four to nine, or 10 or more) and menopausal status (premenopausal v postmenopausal). The accrual goal of 200 patients was chosen to have 80% power when comparing the two arms for three end points: percentage of patients receiving at least 85% of planned chemotherapy dose, percentage of patients with cosmesis graded poor or fair at 4 years after randomization, and disease-free survival. In September 1988, the protocol was amended to include node-negative patients with either estrogen receptor (ER)-negative tumors or lymphatic vessel invasion in the primary tumor and to give 5 years of tamoxifen after completion of all chemotherapy and radiation to postmenopausal women with ER-positive tumors. Because both changes were expected to increase median disease-free survival, the overall accrual goal was increased to 230 patients to have 80% power to detect a true average hazard ratio (HR) of 0.61. Surgery included excision of all gross disease from the breast and a level I/II axillary dissection. Surgical margins were classified as positive, close, negative, or unknown based on the original pathology reports. Central pathology review was not performed. Systemic therapy consisted of the following six-drug regimen repeated every 21 days for four cycles: cyclophosphamide (500 mg/m2 administered intravenously [IV], day 1); doxorubicin (45 mg/m2 IV, day 3); methotrexate (200 mg/m2 IV, days 1 and 15); fluorouracil (500 mg/m2 IV, day 1); prednisone (40 mg/m2 administered orally, day 1 through 5), and leucovorin (10 mg/m2 administered orally every 6 hours, days 2 to 4 and 16 to 18). A total of 11 patients in the CT-first group and seven patients in the RT-first group received tamoxifen. Protocol guidelines dictated radiation treatment of the entire breast to 45 Gy in 1.8 Gy fractions, giving five fractions weekly, followed by an additional 16 to 18 Gy to the lumpectomy site using either electrons or an interstitial implant. The regional lymphatics were included at the discretion of the radiation oncologist and were treated in 88 patients. Local recurrences were defined as being within the parenchyma or skin of the ipsilateral breast, with or without simultaneous distant or regional nodal recurrences. Regional recurrences included disease in the ipsilateral axilla, infraclavicular, or internal mammary lymph nodes without local recurrence, with or without simultaneous distant recurrence. Distant recurrences included ipsilateral or contralateral supraclavicular nodes, contralateral internal mammary nodes, and all breast cancer recurrences at sites other than the breast or regional lymphatics. Other failures were defined as contralateral breast cancer, primary cancers in a site other than the breast, or death from other causes. All statistical tests used were two-sided. Kaplan-Meier12 curves were calculated for times from randomization to various events. Contralateral breast cancer was not counted as an event in calculating the time to distant failure, nor did it censor the patient for distant failure. Primary cancers in a site other than the breast were not counted as events in calculating the time to distant failure but did censor the patient for follow-up for time to distant failure. Log-rank analysis was used to compare curves,13 and power calculations assumed that times had an exponential distribution. The proportional hazards model was used to examine the effect of covariates listed in Table 1 on the time to distant failure (whether or not local failure occurred first). Patterns of first failure were compared using the Fisher's exact test.14 Interactive models were constructed to examine whether the relationship between treatment sequence and site of first failure was affected by other variables.
Patient, tumor, and treatment characteristics are listed in Table 1. The two treatment sequences were fairly well balanced with respect to most patient and treatment characteristics, with the exception of more patients with tumor size between 1 and 2 cm and more patients with an extensive intraductal component (EIC) in the RT-first arm. In addition, fewer RT-first patients had a boost dose of 16 Gy or higher. As previously reported,10 the median dose of chemotherapy delivered was significantly lower in the RT-first patients (81% v 88% of cumulative planned dose; P = .01). The median times to complete chemotherapy, however, was the same in both groups (84 days). The median dose of radiation delivered to the lumpectomy site was the same in both groups (61 Gy) as was the median lumpectomy site boost dose (16 Gy) and the median duration of radiation (49 days). The percentage of patients receiving regional nodal radiation was also the same in both arms (35%). The median length of follow-up among surviving patients was 135 months (range, 17 to 196 months). Nine patients (four patients in the RT-first arm and five patients in the CT-first arm) who were without recurrence or second cancer were lost to follow-up within the first 10 years after randomization (range, 1.4 to 9.7 years). Fifty-nine percent of the 115 patients who at last follow-up were alive without breast cancer recurrence, contralateral breast cancer, or second malignancy had at least 10 years of follow-up after randomization. There was no statistically significant treatment difference in the rates of freedom from any event, including breast cancer recurrence, contralateral breast cancer, second malignancy, or death (P = .88; Fig 1, average ratio of CT-first hazard to RT-first hazard of 0.97; 95% CI, 0.69 to 1.37), freedom from distant metastases (P = .70; Fig 2; average HR, 0.92; 95% CI, 0.61 to 1.39), or overall survival (P = .41; Fig 3; average HR, 0.83; 95% CI, 0.53 to 1.30). The 10-year rate of any event was 46% for patients in the CT-first arm compared with 51% in the RT-first arm. The 10-year rates of distant metastasis were 35% and 36% in the two arms, respectively; and the 10-year rates of death were 28% and 33%, respectively.
The rates of freedom from distant metastases had been significantly different between the two arms in our prior report of this study. As seen in Fig 2, the curves still separate until approximately 4 years after randomization, but thereafter they start to come together. The average ratio of distant failure hazards (CT-first divided by RT-first) is 0.71 for the first 5 years after randomization and 2.27 after 5 years. Proportional hazards models were used to determine factors influencing freedom from distant metastases. Treatment sequence was not significant (HR, 0.86; P = .48). T stage I versus stage II (HR, 0.47; P < .0001), negative axillary nodes versus positive axillary nodes (if at least six nodes were examined; HR, 0.43; P = .05), and pre- or perimenopausal status versus postmenopausal status (HR, 0.60; P = .03) were associated with decreased rates of distant failure. The distributions of the sites of first failure were compared between the two treatment arms (Table 2). The overall difference was not significant (P = .41). During the entire follow-up period, 11% of patients in the CT-first arm experienced local recurrence (with or without simultaneous regional recurrence) as the first site of recurrence, and 4% had simultaneous local and distant failure, as compared with 10% and 2% in the RT-first arm, respectively. Twenty-five percent of patients in the CT-first arm experienced a distant or regional metastasis as their first site of recurrence, compared with 33% of patients in the RT-first arm. Twelve percent of patients in the CT-first arm and 8% of the patients in the RT-first arm experienced a contralateral breast cancer, second malignancy, or death as their first failure.
Potential interactions between treatment sequence and various tumor and treatment characteristics were also examined in site of first failure models. There were significant interactions of treatment and three covariates: negative margins (local interaction risk ratio, 0.09; P = .01; patients with negative margins had fewer than expected local failures in the CT-first group), ER-positive disease (local interaction risk ratio, 10.27; P = .01, meaning that patients who were ER-positive had more than expected local failures in the CT-first group), and time from last breast surgery to randomization  30 days (local interaction risk ratio, 0.08; P = .02). We previously reported on the interaction of treatment sequence with margin and nodal status on sites of first failure. Although margin status was significant in the model, axillary lymph node involvement did not preferentially predict for local or distant failure in either group. For the 123 patients with negative margins, the crude local recurrence rates for CT-first and RT-first patients were 6% and 13%, respectively. Corresponding rates of distant and regional recurrences were 18% and 26%. Among women with close margins (n = 47), crude local recurrence rates were 32% and 4%, respectively; distant/regional recurrences were 37% and 43%. In the group with positive margins (n = 51), local recurrences occurred in 23% of CT-first and 20% of RT-first patients; corresponding distant/regional recurrences occurred in 29% of CT-first patients and 35% of RT-first patients.
The current analysis of this trial, with a median follow-up time exceeding 10 years, showed no statistically significant difference in the rates of freedom from any event, freedom from distant metastasis, or overall survival between the CT-first and RT-first arms. Although there was a higher risk of local recurrence in women treated with chemotherapy first and a higher rate of regional/distant metastasis in woman receiving radiation first, the overall pattern of first failure was not significantly different between the two groups. We would like to acknowledge several limitations of this study. Small patient numbers limit our ability to detect small but clinically important differences in times to various end points. This is particularly true when examining patient subsets. At this time, the study has adequate numbers of failures to have at least 80% power to detect a true hazard ratio of 0.63 for freedom from any event, a true ratio of 0.56 for freedom from distant metastasis, and a true ratio of 0.53 for overall survival. With more than 10 years of follow-up, the 95% CIs for the hazard ratio of chemotherapy-first to radiation therapy-first are 0.69 to 1.37 for time to any event, 0.61 to 1.39 for distant metastasis-free survival, and 0.53 to 1.30 for overall survival. In the future, questions of this nature might be better addressed in a cooperative group setting to include a larger number of patients. In addition, patients with positive and unknown margins were included. This does not reflect the current surgical standard for breast conservation15–18 and may have obscured the overall treatment effects. Also, tamoxifen was used in only a minority of patients with ER-positive tumors (22% in the CT-first group, and 13% in the RT-first group). Given the reported effect of tamoxifen on local tumor control,19,20 it is possible that the patterns of first recurrence may have changed if tamoxifen had been more widely used. Since our trial started, the administration of adjuvant chemotherapy before radiation has become common practice in women with early-stage breast cancer who undergo conservative surgery. The current update provides reassurance that delaying radiation to give 12 weeks of chemotherapy does not compromise outcome. Whether this remains true for longer chemotherapy regimens is still unknown. These updated results also suggest that caution should be taken in patients with close margins to be treated with initial chemotherapy. Although subset analysis can at best be considered hypothesis generating, we found a high rate of local recurrence in patients with close surgical margins who were treated with initial chemotherapy. If confirmed by other studies, this would indicate that such patients should be treated with re-excision before chemotherapy. Patients with positive margins seem to have a high rate of recurrence even when radiation is given first and should undergo additional breast surgery. Finally, it is prudent to initiate adjuvant treatment promptly after surgery, particularly when the start of radiation therapy will be further delayed as a result of the use of adjuvant chemotherapy.
The authors indicated no potential conflicts of interest.
Presented in part at the Annual Meeting of the American Society for Therapeutic Radiology and Oncology, San Francisco, CA, November 5, 2001. Authors' disclosures of potential conflicts of interest are found at the end of this article.
1. Hartsell WF, Recine DC, Griem KL, et al: Delaying the initiation of intact breast irradiation for patients with lymph node positive breast cancer increases the risk of local recurrence. Cancer 76:2497-2503, 1995[CrossRef][Medline] 2. Buchholz TA, Austin-Seymour MM, Moe RE, et al: Effect of delay in radiation in the combined modality treatment of breast cancer. Int J Radiat Oncol Biol Phys 26:23-25, 1993[Medline] 3. Recht A, Come SE, Gelman RS, et al: Integration of conservative surgery, radiotherapy, and chemotherapy for the treatment of early-stage, node-positive breast cancer: Sequencing, timing, and outcome. J Clin Oncol 9:1662-1667, 1991[Abstract]
4. Huong J, Barbera L, Brouwers M, et al: Does delay in starting treatment affect the outcome of radiotherapy? A systematic review. J Clin Oncol 21:555-563, 2003 5. Vujovic O, Perera F, Dar AR, et al: Does delay in breast irradiation following conservative breast surgery in node-negative breast cancer patients have an impact on risk of recurrence? Int J Radiat Oncol Biol Phys 40:869-874, 1998[CrossRef][Medline]
6. Meek AG, Park TL, Weiss TA, et al: Effect of delayed radiation therapy on local control in breast conservation therapy. Radiology 200:615-619, 1996 7. Leonard CE, Wood ME, Zhen B, et al: Does administration of chemotherapy before radiotherapy in breast cancer patients treated with conservative surgery negatively impact local control? J Clin Oncol 13:2906-2915, 1995[Abstract] 8. Buzdar AU, Kau SW, Smith TL, et al: The order of administration of chemotherapy and radiation and its effect on the local control of operable breast cancer. Cancer 71:3680-3684, 1993[CrossRef][Medline] 9. Brufman G, Sulkes A, Biran S: Adjuvant chemotherapy with and without radiotherapy in stage II breast cancer. Biomed Pharmacother 42:351-355, 1988[Medline]
10. Recht A, Come SE, Henderson IC, et al: The sequencing of chemotherapy and radiation therapy after conservative surgery for early-stage breast cancer. N Engl J Med 334:1356-1361, 1996 11. Hardenbergh PH, Recht A, Gollamudi S, et al: Treatment-related toxicity from a randomized trial of the sequencing of doxorubicin and radiation therapy in patients treated for early stage breast cancer. Int J Radiat Oncol Biol Phys 45:69-72, 1999[CrossRef][Medline] 12. Kaplan EL, Meier P: Nonparametric estimation from incomplete observations. J Am Stat Assoc 53:457-481, 1958[CrossRef] 13. Peto R, Peto J: Asymptotically efficient rank invariant test procedures. J R Stat Soc B 153:185-206, 1972 14. Fisher RA: The logic of inductive inference. J R Stat Soc B 98:39-84, 1935[CrossRef]
15. Park CC, Mitsumori M, Nixon A, et al: Outcome at 8 years after breast-conserving surgery and radiation therapy for invasive breast cancer: Influence of margin status and systemic therapy on local recurrence. J Clin Oncol 18:1668-1675, 2000 16. Obedian E, Haffty BG: Negative margin status improves local control in conservatively managed breast cancer patients. Cancer J Sci Am 6:28-33, 2000[Medline] 17. Peterson ME, Schultz DJ, Reynolds, et al: Outcomes in breast cancer patients relative to margin status after treatment with breast-conserving surgery and radiation therapy: The University of Pennsylvania experience. Int J Radiat Oncol Biol Phys 43:1029-1035, 1999[CrossRef][Medline] 18. Horiguchi J, Iino Y, Takei H, et al: Surgical margin and breast recurrence after breast-conserving therapy. Oncol Rep 6:135-138, 1999[Medline]
19. Buchholz TA, Tucker SL, Erwin J, et al: Impact of systemic treatment on local control for patients with lymph node-negative breast cancer treated with breast-conservation therapy. J Clin Oncol 19:2240-2246, 2001
20. Fisher B, Bryant J, Dignam JJ, et al: Tamoxifen, radiation therapy, or both for prevention of ipsilateral breast tumor recurrence after lumpectomy in women with invasive breast cancers of one centimeter or less. J Clin Oncol 20:4141-4149, 2002 Submitted April 3, 2003; accepted December 20, 2004. This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2005 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
