메인 International Journal of Radiation Oncology*Biology*Physics Second nonbreast malignancies after conservative surgery and radiation therapy for early-stage...
문제 보고This book has a different problem? Report it to us
"네" 선택하시는 조건: "네" 선택하시는 조건: "네" 선택하시는 조건: "네" 선택하시는 조건:
파일 열기 성공했습니다
파을 내용은 책 (또는 만화)입니다
책 내용이 적당합니다
파일의 제목, 작성자와 언어가 책 설명과 일치합니다. 다른 필드는 보조이므로 무시하셔도 좋습니다.
"아니요" 선택하시는 조건: "아니요" 선택하시는 조건: "아니요" 선택하시는 조건: "아니요" 선택하시는 조건:
- 잘못된 파일입니다
- 이 파일이 DRM으로 보호돼 있습니다
- 파일은 책이 아닙니다 (예: xls, html, xml)
- 파일은 기사입니다
- 파일은 책에 일부입니다
- 파일은 잡지입니다
- 파일은 시험지 또는 테스트입니다
- 파일은 스팸입니다
책의 내용이 적당하지 않으며 차단되어야 한다고 생각합니다
파일의 제목, 작성자와 언어가 책 설명과 일치하지 않습니다. 다른 필드는 무시하셔도 좋습니다.
Change your answer
Int. J. Radiation Oncology Biol. Phys., Vol. 52, No. 2, pp. 406 – 414, 2002 Copyright © 2002 Elsevier Science Inc. Printed in the USA. All rights reserved 0360-3016/02/$–see front matter PII S0360-3016(01)02661-X CLINICAL INVESTIGATION Breast SECOND NONBREAST MALIGNANCIES AFTER CONSERVATIVE SURGERY AND RADIATION THERAPY FOR EARLY-STAGE BREAST CANCER SHARON GALPER, M.D., M.P.H.,*† REBECCA GELMAN, PH.D.,†‡ ABRAM RECHT, M.D.,†§ BARBARA SILVER, B.A.,*† ANITA KOHLI, B.A.,† JULIA S. WONG, M.D.,*† TERESA VAN BUREN, M.D.,†§ ELIZABETH H. BALDINI, M.D., M.P.H.,*† AND JAY R. HARRIS, M.D.*† *Department of Radiation Oncology, Brigham and Women’s Hospital and Dana-Farber Cancer Institute, Boston, MA; ‡Department of Biostatistics, Dana-Farber Cancer Institute and Harvard School of Public Health, Boston, MA; §Department of Radiation Oncology, Beth Israel Deaconess Medical Center, Boston, MA; †Joint Center for Radiation Therapy, Harvard Medical School, Boston, MA Purpose: Breast cancer patients treated with conservative surgery and radiation therapy are at risk of developing second nonbreast malignancies (SNBMs). The purpose of this study was to determine the incidence of all SNBMs and SNBMs by specific location among long-term survivors and to compare the risk of these events to the age-specific incidence of malignances as first cancers in the Surveillance Epidemiology and End-Results Program (SEER) population. Methods and Materials: We analyzed the likelihood of SNBM development for 1884 patients with clinical Stage I or II breast cancer treated with gross excision and >60 Gy (median 63) to the breast between 1970 and 1987. Fifty-seven percent received supraclavicular/axillary radiation (median dose 45 Gy, range 20 – 60) and 28% received systemic therapy. The median age at diagnosis was 52 years. The median clinical tumor size was 2 cm. Patients were considered at risk of an SNBM until the development of the first of distant metastases or contralateral breast cancer or death or, if alive and disease-free, until the; last follow-up visit. The expected numbers of cancers were obtained from the SEER database, using the age-specific incidence for white women within 5-year age groups and 5-year calendar intervals. The median time at risk for an SNBM was 10.9 years (range 0.2–27.9). Results: By 8 years of follow-up, 432 patients (23%) had developed distant metastases, 295 patients (16%) a local/regional recurrence, and 159 (8%) a contralateral primary. Of the 1884 patients in our cohort, 147 (8%) developed an SNBM compared with the 127.7 expected from SEER. This corresponds to an absolute excess of 1% of the study population and a relative increase of 15% greater than that expected from SEER (p ⴝ 0.05). Within the first 5 years, the observed and expected rates of SNBMs were identical (47 vs. 46.9). After 5 years, 24% more SNBMs were observed than expected (100 vs. 80.8, p ⴝ 0.02). Among patients <50 years old at breast cancer diagnosis, 43% more observed SNBMs occurred than expected (40 vs. 28, p ⴝ 0.02). For patients >50 years, 7% more SNBMs were observed than expected (107 vs. 99.7, p ⴝ 0.25). Lung SNBMs were observed in 33 women, 52% more than the 21.67 predicted by SEER (p ⴝ 0.01). Most of the lung SNBMs occurred >5 years after treatment (n ⴝ 23) and in women who were >50 years at the time of their breast cancer diagnosis (n ⴝ 27). The observed incidence of ovarian cancer was significantly greater than expected among patients <50 years (7 vs. 1.96, p ⴝ 0.004) but was not different than expected for patients >50 years (5 vs. 5.3, p ⴝ 0.61). Among the 7 sarcomas, 3 developed in the radiation field. Conclusions: SNBMs occur in a substantial minority (8%) of patients treated with conservative surgery and radiotherapy. However, the absolute excess risk compared with the general population is very small (1%). This excess risk is only evident after 5 years. In particular, a slightly increased incidence of lung SNBMs and a somewhat larger increase in ovarian cancer among younger patients was found. Our data suggest that preventive strategies to reduce the incidence of certain cancers (e.g., smoking cessation and prophylactic oophorectomy) and/or continued monitoring for SNBMs to increase the likelihood of early detection and treatment may be prudent in this population. © 2002 Elsevier Science Inc. Breast cancer, Second malignancy, Radiation therapy. INTRODUCTION all patients with invasive breast cancer is 40% (1). The highest percentage of long-term survivors is found among patients with small invasive breast tumors who do not have histologic evidence of axillary lymph node involvement. There are many long-term survivors among all subsets of breast cancer patients. The overall survival at 20 years for Presented in part at the Forty-Second Annual Meeting of the American Society of Therapeutic Radiology and Oncology, Boston, MA, October, 2000. Received Jan 3, 2001, and in revised form Aug 23, 2001. Accepted for publication Sep 4, 2001. Reprint requests to: Sharon Galper, M.D., M.P.H; Department of Radiation Oncology, Dana-Farber Cancer Institute, 44 Binney St., Boston, MA 02115. Tel: (617) 732-7906; Fax: (617) 2645242; E-mail: email@example.com Dr. Galper is the recipient of a Health Services Research Fellowship from the Agency for Health Care Policy and Research. 406 Incidence of second nonbreast malignancies The disease-free survival rates at 20 years for patients with T1N0 tumors range between 70% and 85% (2). Some subsets of node-positive patients may also experience prolonged survival. For patients with T1N1 tumors and 1–3 positive axillary lymph nodes, the disease-free survival rate at 20 years is 57% (2). Even for patients with ⱖ 10 positive lymph nodes, the disease-free survival rates at 10 years range between 15% and 31% (3). All subsets of breastcancer patients are therefore at risk of developing second nonbreast malignancies (SNBMs). The possible causes of SNBMs among long-term survivors of breast cancer are several. Some SNBMs are sporadic and would have occurred in the absence of a breast cancer diagnosis. These SNBMs are considered “expected” cancers. Some SNBMs may be linked to the breast cancer diagnosis. The same environmental or genetic factors that predisposed the patient to breast cancer may have also contributed to the development of the SNBM. Finally, some SNBMs may be induced by breast cancer treatment. The purpose of our study was to determine the incidence of all SNBMs and, specifically, the incidence of SNBMs in specific anatomic sites among long-term survivors treated with conservative surgery and radiotherapy (RT). In addition, we compared the incidence of these events among long-term survivors to the age-specific incidence of malignancies as first cancers in the Surveillance, Epidemiology, and End Results Program (SEER). METHODS AND MATERIALS Between July 1968 and December 1987, 2140 patients without a history of a prior malignancy (except nonmelanoma skin cancer or in situ carcinoma of the cervix) were treated at the Joint Center for Radiation Therapy for Stage I or II invasive breast carcinoma. Patients who presented with synchronous bilateral primary tumors or who had prior cancer in the opposite breast were excluded from this group. To reflect current standards of practice and ensure adequate follow-up, the study population was limited to 1884 patients who met certain eligibility criteria. All patients had clinically node-negative breast cancer treated with complete gross excision of the primary. All patients received tangential radiation to the breast followed by a boost, rendering the total dose to the tumor bed at least 60 Gy. All patients had a minimum potential follow-up time of 8 years (i.e., the patient was treated at least 8 years ago and had a recurrence or died within 8 years of the start of RT or was followed for a minimum of 8 years without recurrence, death from other causes, or being lost to follow-up). The median time at risk of an SNBM was 10.9 years (range 0.2–27.9). All patients in the study population were treated between 1970 and 1987. Within the first 5 years of follow-up, 1884 patients were at risk of an SNBM; between 5 and 10 years, 1406 patients were at risk; and after 10 years, 1036 patients were at risk of an SNBM. The patient characteristics are listed in Table 1. The median patient age at diagnosis was 52 years (range 23– 88). ● S. GALPER et al. 407 The median clinical tumor size was 2 cm (range 0 –5). Smoking history and family history were not routinely available. Surgery for the primary tumor involved gross total excision in all patients, without regard for microscopic resection margins. Eighty-one percent of patients underwent axillary dissection, of whom 80% had ⱖ6 lymph nodes removed. Among those who underwent axillary dissection, the median number of positive nodes was 0 (range 0 –26). Thirtyfive percent had at least 1 positive node. RT to the breast was usually delivered by a 4 or 6-MV linear accelerator. Patients with large separations along the medial-to-lateral plane defined by the posterior border of the tangential fields were occasionally treated on an 8-MV accelerator. Bolus was rarely used on the breast. The dose to the entire breast was usually 45– 46 Gy (median 46, range 25.2– 62.5; for 92%, it was between 45 and 50 Gy), delivered at 1.8 –2.0 Gy per fraction, 5 times weekly. The median total dose to the primary tumor site (including the boost) was 63 Gy (range 60 – 84), with 90% of patients receiving between 60.4 and 70.2 Gy. Nodal irradiation was given at the discretion of the individual radiation oncologist. Thirteen percent of patients underwent RT to an axillary apex/supraclavicular field, and 44% of patients received RT to the full axillary/supraclavicular fossa. Thus, a total of 58% of patients received third-field radiation. The “hanging block” or “corner block” technique was usually used. The details of treatment technique and patient setup have been described previously (4 – 6). The median doses to the supraclavicular fossa and axilla were 45 Gy (range 19.8 – 60.0) and 46 Gy (range 5–72.8), respectively. Deliberate efforts to treat the internal mammary nodes were not routinely recorded. Adjuvant systemic therapy was administered to 28% of patients, consisting of chemotherapy alone for 24%, tamoxifen alone for 4%, and both for 0.6%. The chemotherapy regimens varied with regard to drugs used, number of cycles, and sequencing with RT, but cyclophosphamide, methotrexate, and fluorouracil-based or doxorubicin-containing regimens were given to 428 (94%) of the patients who received chemotherapy. An SNBM was defined as any nonbreast malignancy, except basal cell or squamous cell carcinoma of the skin and carcinoma in situ of the cervix. Because patients who develop contralateral breast cancer or distant metastasis may receive additional therapies that could contribute to the development of SNBMs, patients who experienced one of these events were subsequently censored from the analysis. Patients were thus considered at risk of an SNBM until the time of developing contralateral breast cancer, distant metastasis, or death. Patients who developed local failure alone were not censored from our analysis because these patients were likely to undergo continued monitoring for new sites of disease. In addition, such patients were most often treated with surgery alone and without additional therapies that might have an impact on their risk of developing an SNBM. Three hundred fifty-one patients had local failure as the first 408 I. J. Radiation Oncology ● Biology ● Physics Volume 52, Number 2, 2002 Table 1. Patient characteristics Descriptive statistics Age (yr) Median Range ⬍35 35–44 45–54 55–64 65–74 75–88 Clinical size (cm) Median Range Unknown T stage 1 2 N stage 0 1 Clinical stage 1 2 Estrogen receptor status Positive Negative Unknown Nodes positive if ⱖ6 sampled Median Range 0 1–3 ⱖ4 Unknown Total dose to primary (Gy) Median Range 5th percentile 95th percentile Axillary radiation dose (Gy) ⬎0 Median if ⬎0 Range if ⬎0 5th percentile, ⬎0 95th percentile, ⬎0 Supraclavicular radiation dose (Gy) ⬎0 Median if ⬎0 Range if ⬎0 5th percentile, ⬎0 95th percentile, ⬎0 Adjuvant chemotherapy only Adjuvant hormonal therapy only Adjuvant chemotherapy and hormonal therapy Infiltrating ductal histology Lymphatic vessel invasion Absent Indeterminate Present Unknown Margins Negative Close Positive Unknown % 52 23–88 7 24 26 22 15 6 2 0–5 13 61 39 89 11 57 43 35 20 45 0 0–26 65 25 10 0.2 63.0 60.0–84.0 60.4 70.2 44 46.0 5.0–72.8 44.0 54.0 57 45.0 19.8–60.0 43.2 46.0 24 4 0.6 82 45 11 22 22 15 5 10 70 Incidence of second nonbreast malignancies ● S. GALPER et al. 409 Table 2. Observed vs. expected SNBMs Overall ⬍5 yr ⱖ5 yr Age (y) ⬍50 ⱖ50 Patients (n) Median follow-up (yr) Observed (n) 95% Confidence interval* Expected Observed/expected p 1884 1884 1400 10.9 5.0 12.4 147 47 100 124–173 35–62 81–122 127.7 46.9 80.8 1.15 1.00 1.24 0.05 0.51 0.02 828 1056 11.5 10.3 40 107 29–54 88–129 28.0 99.7 1.43 1.07 0.02 0.25 * 95% confidence interval of observed cases of SNBM. SNBM ⫽ second nonbreast malignancy. site of failure in this cohort, of whom 95 (27%) received systemic therapy at the time of relapse, consisting of chemotherapy alone in 59 (17%), tamoxifen alone in 77 (22%), and both in 13 patients (4%). In addition, 4 (1%) of the 351 patients who experienced local failure as the first site of failure underwent wide excision followed by RT. Data regarding the use of postmastectomy RT in the remaining patients are not available. All SNBMs were confirmed pathologically. Lung lesions were classified as either lung metastasis from the original breast primary or an SNBM of the lung according to the following criteria. Lesions with squamous, small cell, or large cell histologic features were classified as an SNBM of the lung. For adenocarcinomas, the clinical and pathologic features of the new lesion were compared with those of the initial breast primary to help distinguish between breast cancer metastatic to the lung and a new lung primary. In cases in which this distinction could not be made, we systematically classified such indeterminate lesions as metastatic breast cancer. Thus, the true incidence of SNBMs of the lung may be somewhat greater than our estimate of it. Estimates of expected numbers of cancers were obtained from the SEER databases (SEER 9 Registries Public Use, August 1998) for 1982–1997 using age-specific incidences of first cancers for white women within 5-year age groups and within 5-year calendar intervals. The SEER data from 1982 to 1986 were used for the first 5 years after RT; the SEER data from 1987 to 1991 for 5–10 years after RT, and the SEER data from 1992 to 1996 for ⱖ10 years after RT. These periods for SEER were chosen because 5-year intervals give more accurate estimates than 1-year intervals (especially for rare tumors) and because 79% of patients in our study had their first full year of follow-up in the period between 1982 and 1986. Because registry data are notoriously bad at distinguishing recurrences from second primaries, we chose to base the comparisons of observed and expected SNBMs on only the first SNBM observed in each patient of our cohort. There was not an easy way to get summary rates from SEER based only on patients reported to have a single primary or the first of several cancers reported in a single patient; therefore, the SEER incidence rates (and the expected numbers of cancers for each group) were based on the overall incidence reported by SEER. However, the incidence rate of second or higher number malignancies in the SEER data set was only ⱕ1% of the total incidence for lung cancer (⬍1% for patients ⬍50 or ⱖ75 years; 1% for patients aged 50 –74), and similar for several other cancers we checked; thus, the expected number of first malignancies of a particular type was only slightly overestimated by this method. In addition, we chose to compare the observed number of SNBMs to the number expected among an age-matched population of white females, because the large majority of patients in our cohort were white. Estimates of expected SNBMs by specific anatomic location were also obtained from the SEER database using age-matched estimates for white women during the follow-up period of our study. For the purposes of this study, SNBMs of the thorax were defined as any SNBM of the lung, bronchus, mediastinum, trachea, esophagus, or pleura. However, SEER does not have a separate category for sarcomas (some sarcomas are categorized by anatomic site and others are categorized as “soft tissue”). In addition, SEER does not divide melanomas and lymphomas by anatomic site. Thus, sarcomas, lymphomas, and melanomas were included in the overall calculation of SNBMs but excluded from the analysis of specific sites such as the thoracic region. Comparisons of observed and expected SNBMs and individual cancer sites were performed using the Exact Poisson Test. All p values were one-sided. Ninety-five percent exact Poisson confidence intervals for the number of second tumors were two-sided and thus do not correspond to the one-sided test. RESULTS Of the 1884 patients in our study population, 432 patients (23%) developed distant metastasis, 295 patients (16%) a locoregional recurrence, and 159 (8%) a contralateral primary; a total of 147 (8%) developed an SNBM. On the basis of the age-matched estimates from the SEER database calculated for the follow-up period of our study, 128 patients would have been expected to develop an SNBM, yielding an absolute excess of 19 cases or 1% of the study population. Table 2 compares the number of observed SNBMs with the number of expected cases based on age-matched and calendar year-matched data from the SEER database. Over- 410 I. J. Radiation Oncology ● Biology ● Physics Volume 52, Number 2, 2002 virtually identical to the 46.9 predicted by SEER ( p ⫽ 0.51). More than 5 years after treatment, 100 patients developed an SNBM, 24% more than the 80.8 expected from SEER ( p ⫽ 0.02). Forty patients who were ⬍50 years at the time of breast cancer treatment developed SNBMs, 43% more than the 28 predicted by SEER ( p ⫽ 0.02). Among patients ⱖ50 years at the time of breast radiation, 107 developed SNBMs, 7% more than the 99.7 predicted by SEER ( p ⫽ 0.25). One hundred fifty-two SNBMs were observed in 147 patients (Table 3): 33 lung cancers, 2 mesotheliomas, 1 thyroid cancer in 1 patient who had undergone RT for a benign condition as a child, 23 colorectal cancers, 2 cancers of the gallbladder, 4 cancers of the pancreas, 1 hepatoma, 1 carcinoid of the colon, 1 esophageal cancer, 2 gastric cancers, 12 ovarian cancers, 9 endometrial cancers, 4 cervical cancers, 1 vaginal cancer, 1 vulvar cancer, 1 borderline malignant ovarian tumor, 7 sarcomas, 14 melanomas, 8 lymphomas, 6 meningiomas, 3 leukemias, 2 multiple myeloma, 1 polycythemia vera, 2 glioblastoma multiforme, 1 astrocytoma, 2 head-and-neck cancers (1 adenoid cystic of the parotid gland and 1 maxillary sinus tumor), 1 neuroendocrine tumor (involving the liver), 1 neuroectodermal tumor (involving the liver), 2 cancers of the bladder, 2 renal cell carcinomas, 1 squamous cell carcinoma of the nipple, and 1 unknown primary. Four patients developed a third nonbreast malignancy after the initial SNBM diagnosis. One patient developed endometrial cancer 16 months after lung cancer, 1 patient developed acute myelogenous leukemia (AML) 72 months after ovarian cancer, 1 patient developed gastric cancer a little more than 1 year after lymphoma, and 1 patient developed diffuse large cell lymphoma 40 months after chronic lymphocytic leukemia (CLL). A fifth patient was found to have 2 synchronous SNBMs involving the ovary and lung. Table 3. Specific sites of SNBM Second cancer Neck and thoracic Lung Mesothelioma Thyroid Gastrointestinal Colorectal Gallbladder Pancreas Hepatoma Carcinoid Esophagus Gastric Gynecologic Ovarian Endometrial Cervical Vaginal Vulvar Borderline malignant ovarian Other Melanoma Sarcoma Lymphoma Meningioma Leukemia Multiple myeloma Polycythemia vera Glioblastoma multiforme Astrocytoma Head and neck Bladder Renal cell Squamous cell of skin Unknown primary Neuroendocrine Neuroectodermal Synchronous second Third cancer cancer 32 2 1 1 23 2 4 1 1 1 1 1 11 8 4 1 1 1 1 1 14 7 7 6 2 2 1 2 1 2 2 2 1 1 1 1 1 1 SNBM ⫽ second nonbreast malignancy. Thoracic SNBMs Thoracic SNBMs were observed in 35 patients (2%). Two patients developed mesotheliomas and 33 developed lung cancer as their first malignancy after breast cancer, 52% more than the 21.67 expected ( p ⫽ 0.01) (Table 4). Among patients ⬍50 years at breast cancer diagnosis, 6 cases of lung cancer were observed compared with the 3.29 expected ( p ⫽ 0.12). Among patients ⱖ50 years at breast all, 147 patients developed SNBMs, 15% more than the 127.7 expected from SEER ( p ⫽ 0.05). Because most effects of radiation-induced carcinogenesis are unlikely to be manifest until ⱖ5 years after treatment, we divided our analysis into events that occurred in the first 5 years and events that occurred later. Within the first 5 years after treatment, 47 patients were observed to develop an SNBM, Table 4. Lung cancer: Observed vs. expected Overall ⬍5 yr ⱖ5 yr Age (y) ⬍50 ⱖ50 Patients (n) Observed (n) 95% Confidence interval* Expected Observed/expected p 1,884 1,884 1,400 33 10 23 23–46 5–18 15–35 21.67 7.31 14.36 1.52 1.37 1.60 0.01 0.20 0.02 828 1,056 6 27 2–13 18–39 3.29 18.38 1.82 1.47 0.12 0.03 * 95% confidence interval of lung cancers observed. Incidence of second nonbreast malignancies cancer diagnosis, 27 cases of lung cancer were observed compared with the 18.38 expected ( p ⫽ 0.03). When the analysis was divided into events that occurred during the first 5 years after treatment of breast cancer and those that occurred later, 10 lung cancers were observed during the first 5 years after breast cancer treatment, 37% more than the 7.3 expected ( p ⫽ 0.20). However, ⬎5 years after treatment, 23 patients developed lung cancer, 60% more than the 14.36 expected ( p ⫽ 0.02). Of note, 94 patients (5%) developed lung metastasis as a component of their metastatic disease. Isolated lung metastasis developed in 51 patients (3%). The records of the 34 patients who developed a thoracic SNBM as the first site of SNBM were reviewed to determine whether the SNBM occurred on the ipsilateral or contralateral side of the initial breast cancer treatment. Among the 33 patients who developed lung cancer, 12 were located in the ipsilateral lung, 10 in the contralateral lung, and 1 was bilateral. The location could not be retrospectively determined from the available records in 10 patients. Of the 2 patients who developed mesothelioma, one occurred in the ipsilateral pleura and the other in the contralateral pleura. Of note, a third radiation field directed to the supraclavicular fossa/axilla was included as part of the initial breast cancer treatment in 25 (76%) of the 33 patients who developed an SNBM of the lung. In addition, 5 of these 33 patients had received chemotherapy. The frequency of chemotherapy administration among patients with lung SNBMs was not dissimilar from that of our cohort overall. However, a somewhat larger percentage of patients with lung SNBMs had received third-field radiation than those who did not develop a lung SNBM (76% vs. 58%, p ⫽ 0.03). Ovarian cancer Because patients with breast cancer may have a genetic predisposition to ovarian cancer, we compared the number of observed cases of ovarian cancer with the number expected. Overall, ovarian cancer was observed in 12 patients, 65% more than the 7.26 expected from SEER ( p ⫽ 0.07). For patients ⬍50 years old at the time of RT, 7 cases of ovarian cancer were observed compared with the 1.96 expected from the SEER database ( p ⫽ 0.04). Among patients ⱖ50 years, the numbers of observed and expected cases of ovarian cancer were virtually identical (5 vs. 5.30, p ⫽ 0.61). The median age of diagnosis of ovarian cancer was 50 years (range 34 –77), and the median time to the development of ovarian cancer was 65 months (range 7–157). Colorectal cancers Because patients with breast cancer may have a genetic predisposition to colorectal cancer, we compared the number of patients observed to develop colorectal cancer with the number expected. Overall, the number of observed and expected cases of colorectal cancer were similar (22 vs. 23.44, p ⫽ 0.64). For patients ⱖ50 years at the time of their ● S. GALPER et al. 411 breast cancer diagnosis, 18 cases of colorectal cancer were observed compared with the 20.47 expected ( p ⫽ 0.74). Sarcomas Because SEER does not record sarcomas as a separate site, we were unable to compare the number of sarcomas observed with the number expected. Among the 7 sarcomas observed, 3 were found after close examination of the treatment records to have arisen in the radiation field. One was an angiosarcoma that developed in the breast 85 months after tangential radiation alone. This patient did not have breast edema. The second patient developed an angiosarcoma of the pleura 211 months after three-field irradiation. The exact location of this sarcoma in relation to the matchline could not be determined. The third patient developed a malignant fibrous histiocytoma of the right clavicular head above the matchline 107 months after three-field irradiation. Of the 4 patients who developed sarcomas outside the radiation field, 1 was located on the hand, 1 on the back, and 2 on the thigh. Melanomas Fourteen melanomas were observed compared with the 5.6 expected from the SEER database ( p ⫽ 0.0002). However, it is possible that in the earlier years of the SEER database, some melanomas were counted with cancers by anatomic site rather than as melanomas. In addition, some may have been misdiagnosed as nonmelanomatous skin cancer. These potential errors in misclassification would lead to an underestimate of the expected cases of melanoma. Among the 14 patients who developed melanoma, the location was determined in 10 and included 1 contralateral forearm, 1 contralateral axilla, 1 contralateral supraclavicular fossa, 1 ipsilateral wrist, 1 ipsilateral arm, 3 legs, 1 scalp, and 1 tonsil. Lymphomas Eight patients developed lymphoma as the first site of second malignancy, compared to the 6.47 expected ( p ⫽ 0.32). Hematologic malignancies Three patients developed leukemia: 1 CLL, 1 acute nonlymphocytic lymphoma (ANLL), and 1 AML. Only the patient who developed CLL had received chemotherapy at the time of the breast cancer diagnosis. Neither of the 2 patients who developed multiple myeloma had received chemotherapy. The patient who developed polycythemia vera had received chemotherapy. DISCUSSION Patients with breast cancer who undergo breast-conserving surgery and RT are at increased risk of SNBMs. The possible causes of SNBM in such patients are several. Some may be sporadic and thus manifest in a population of patients without a prior cancer diagnosis. Some may be due 412 I. J. Radiation Oncology ● Biology ● Physics to underlying genetic or environmental factors that predisposed the patient to breast cancer. Some may be due to treatment of the initial primary cancer. This study determined the long-term incidence of all SNBMs and specifically those in the lung, among patients with early-stage breast cancer treated with conservative surgery and irradiation and compared these incidences to the expected rates of cancers from the SEER database. We found that treated patients were at a slightly increased absolute risk of SNBMs (1%) compared with an age-matched population. This risk was only evident after the first 5 years. In addition, this risk was greater for patients diagnosed with breast cancer at a younger age compared with older patients. In particular, patients were at slightly increased risk of lung cancer. An increased incidence of ovarian cancer was also observed, especially in younger patients. Some of this increased risk may be related to underlying genetic factors found in patients with breast cancer. Overall, 147 patients (8%) developed SNBMs compared with 128 expected during this follow-up period, yielding an excess of 19 SNBMs observed or 1% of the total population. This suggests that the great majority of cancers observed in our cohort were sporadic and most likely would have occurred independent of the breast cancer diagnosis or treatment. Only a small percentage of SNBMs in our cohort could be attributed to either underlying genetic factors that predisposed the patient to her initial breast cancer diagnosis or to treatment. We found an increased risk of SNBM of the thorax compared with the expected incidences from the SEER database. Most of this increased risk of thoracic SNBM was due to SNBM of the lung. Distinguishing between a new lung primary and a lung metastasis from breast cancer can often be difficult. In general, when our pathologists were unable to classify a lung lesion, such cases were coded as metastasis from breast cancer and not new primaries. Our estimate of lung cancers observed in this cohort was thus quite conservative. There are several possible causes of lung SNBMs. Some may be sporadic, related to environmental factors such as smoking and asbestos exposure, and others may be related to breast cancer treatments such as RT. To what extent is the excess risk of lung cancer observed in our cohort attributable to RT for breast cancer? To answer this question, we attempted to determine the precise location of new lung primaries. However, this information was available for ⬍75% of patients in our cohort, and simulation films from breast cancer treatment were not readily available for all patients. We were thus unable to determine the precise relationship between the lung SNBM and the radiation field. One can speculate that in a patient who developed a lung SNBM in the ipsilateral lung, the involved lung may have received high doses of radiation. However, even in patients who developed an SNBM in the contralateral lung, the involved lung may have received scatter doses of radiation. We observed 2 cases of cancer in sites that received relatively low scatter doses of radiation: 1 in the thyroid in a patient who had undergone RT for a benign condition as a child and 1 in the esophagus. In general, structures that receive scatter doses of radiation appear to be at minimal risk of Volume 52, Number 2, 2002 SNBM, although we do not have enough detailed information regarding the location of lung cancers to comment more precisely on the dose-related risk of contralateral lung cancer. Our data are consistent with those of other studies. Among 41,109 women who developed breast cancer in Connecticut between 1935 and 1982, the overall risk of developing a subsequent neoplasm, excluding breast cancer as a second cancer, was only slightly elevated (relative risk ⫽ 1.15, 95% confidence interval 1.10 –1.20) with an average follow-up of 6.6 years (7). In a subsequent casecontrol study from the Connecticut Tumor Registry, Neugut et al. (8) found that 10 years after the initial breast cancer diagnosis, the risk of lung cancer overall was increased in women who underwent RT compared with those who did not, with a relative risk of 2.0 (95% confidence interval 1.0 – 4.3). In addition, after breast RT, the risk of lung cancer in the ipsilateral lung was 10% greater than the risk in the contralateral lung, but this was not statistically significant. In general, patients in the series of Neugut et al. were treated with orthovoltage or 60Co and were likely to undergo more extensive nodal irradiation than the patients in our series, increasing the dose to the underlying lung and potentially increasing the risk of carcinogenesis. Specific information on RT doses and treatment plans and cigarette smoking were not available in that series. In the current series, we did not observe an increased incidence of lung cancer in the ipsilateral lung compared with the contralateral lung. However, the location of the lung primary could not be determined in more than one quarter of patients who developed a lung primary, limiting our ability to draw conclusions regarding the impact of the radiation dose on the risk of lung cancer. In a subsequent analysis, Neugut et al. (9) assessed the impact of cigarette smoking and RT for breast cancer on the risk of a subsequent lung cancer. They found a submultiplicative effect of smoking and breast irradiation on the risk of lung cancer that was observed only for the ipsilateral lung. In contrast, Obedian et al. (10) did not find an increased risk of new lung primaries among patients with breast cancer treated with conservative surgery and RT compared with a matched cohort treated with mastectomy. However, this was a retrospective series of 1029 patients based on data obtained from a hospital tumor registry and thus may have either ascertainment or misclassification bias. Several reports have demonstrated an excess risk of lung cancer among patients treated with mantle irradiation for Hodgkin’s disease (11–13). This risk is related to the radiation dose received by the affected area of the lung and is magnified among smokers, especially those who continue to smoke during mantle irradiation. What are the policy implications of these studies in aggregate? Although these studies suggest that patients who undergo RT are at increased risk of lung cancer, the absolute excess risk is small. Harvey et al. (7) estimated that adjuvant RT as practiced in past decades might be expected to cause an extra 7– 8 cases of lung cancer annually among 10,000 irradiated women who survive 10 years. Risks from localized megavoltage RT almost certainly are lower, re- Incidence of second nonbreast malignancies gardless of smoking status. Physicians must encourage smokers who embark on a course of breast cancer therapy to stop smoking. There is increasing interest in the use of low-dose helical CT scanning for lung cancer screening among smokers without a prior history of cancer (14). If data from studies testing the efficacy of low-dose helical CT scanning show a benefit in early identification of lung cancer among smokers without a history of cancer, smokers with a history of RT for early-stage breast cancer may also be candidates for screening programs. The increased risk of ovarian cancer observed in our cohort, particularly among younger patients, is consistent with other series (7, 10). These findings suggest that such patients may have an inherited predisposition to breast or ovarian cancer. The treatment of patients with an inherited susceptibility to breast cancer is controversial given the paucity of data regarding the long-term efficacy of interventions designed to reduce the risk of ovarian cancer or contralateral breast cancer. Two decision analyses have shown substantial improvements in life expectancy in BRCA-1/2 patients without a prior history of cancer who underwent prophylactic oophorectomy (15, 16). Hartmann et al. (17) showed a 90% reduction in the risk of an initial breast cancer diagnosis among patients who carried the BRCA-1/2 mutation and underwent bilateral prophylactic mastectomy. More recently, Schrag et al. (18) developed a model in which prophylactic oophorectomy was shown to increase the life expectancy among women with a germline BRCA-1/2 mutation and good prognosis breast cancer. These gains were attenuated for older patients and for patients with poor prognosis of their primary breast cancer. No study has prospectively evaluated the use of prophylactic oophorectomy or prophylactic mastectomy among patients who are BRCA-1/2 positive with a prior diagnosis of breast cancer. Until data regarding the efficacy of prophylactic oophorectomy or prophylactic mastectomy become available, recommendations in such patients should be made judiciously. It is striking that in the current series only 3 sarcomas were found in the radiation field. These finding are consistent with those of other series (19). In an earlier report from our institution that included a somewhat different population, Pierce et al. (20) found 3 sarcomas in the radiation field. All of those patients received three-field radiation, and 2 sarcomas were found along the matchline. Those 2 patients were treated with an older matchline technique that was associated with matchline fibrosis. Thus, sarcomas in the radiation field are rare, especially with more careful techniques for matching the supraclavicular/axillary fields with the tangential fields. We would like to acknowledge some of the limitations of our study. First, this was a retrospective series and may have been influenced by selection bias. That is, patients who choose conservative surgery may be at lower risk of other cancers than ● S. GALPER et al. 413 patients who choose other breast cancer therapy. We did not compare the incidence of SNBMs in our study population with a cohort of patients treated with mastectomy alone, because a similarly well-documented reference cohort does not exist at our institution. Consequently, the impact of radiation on the risk of SNBMs could not be fully determined. (Of note, Obedian et al. (10) reviewed the records of a hospital tumor registry to compare the incidence of SNBMs among patients treated with breast-conserving surgery and RT with that among patients treated with mastectomy and found no significant difference.) We chose instead to estimate the incidence of expected cancers from the SEER database, which is less subject to underreporting than a hospital tumor registry and thus provides a more comprehensive estimate of first cancers. Choosing the SEER population to estimate the expected incidence of second cancers provides a rigorous estimate of second cancers attributable to sporadic causes, but unfortunately does not account for cancers that may be linked to the original breast cancer diagnosis. In addition, because SEER incidence rates are based on 10% of the United States population, confidence limits around the rates for each specific type of cancer for each 5-year diagnosis period are exceedingly small. Thus, comparisons to the very large SEER database have far more power to detect differences than comparisons to a few thousand patients treated with mastectomy. Second, it is possible that some SNBMs were misclassified as metastatic breast cancer, so that our data may have actually underestimated the risk of SNBM in our cohort. However, the records of all patients who developed a thoracic event, either an SNBM of the thorax or an isolated lung metastasis, were reviewed again and no errors in the initial classification were found. Although patients in our series had a minimal potential follow-up of 8 years and the median time at risk of an SNBM was 10.9 years, longer follow-up is needed to validate these results. Furthermore, we were unable to assess fully the impact of environmental and genetic factors on the risk of SNBMs because of incomplete smoking and family histories. Finally, many of our patients were treated with older treatment approaches, such as a more limited use of systemic therapy and an older matchline technique, which may limit the applicability of our data to current patients. We conclude that SNBMs occur in a substantial minority of patients undergoing RT for early-stage breast cancer. However, the absolute excess risk compared with the general population is very small. Such patients should undergo continued monitoring for SNBMs for the remainder of their lives. In addition, patients who smoke should receive counseling regarding smoking cessation and close monitoring. Finally, genetic counseling and possibly testing should be considered in selected patients. REFERENCES 1. Fox MS. On the diagnosis and treatment of breast cancer. JAMA 1979;241:489 – 494. 2. Rosen PR, Groshen S, Saigo PE, et al. A long-term follow-up study of survival in stage I (T1N0M0) and stage II (T1N1M0) breast carcinoma. J Clin Oncol 1989;7:355–366. 3. Jones VE, Raghavan D. Quantum leaps in treatment of high- 414 4. 5. 6. 7. 8. 9. 10. 11. 12. I. J. Radiation Oncology ● Biology ● Physics risk breast cancer? Prove it! Eur J Cancer 1993;10:1488 – 1493. Buck BA, Siddon RL, Svensson GK. A beam alignment device for matching fields. Int J Radiat Oncol Biol Phys 1985;11:1039 –1043. Svensson GK, Bjarngard BE, Larsen RD, et al. A modified three-field technique for breast treatment. Int J Radiat Oncol Biol Phys 1980;6:689 – 694. Siddon RL, Buck BA, Harris JR, et al. Three-field technique for breast irradiation using tangential field corner blocks. Int J Radiat Oncol Biol Phys 1983;9:583–588. Harvey EB, Brinton LA. Second cancer following cancer of the breast in Connecticut, 1935– 82. Natl Cancer Inst Monogr 1985;68:99 –112. Neugut AI, Robinson E, Lee WC, et al. Lung cancer after radiation therapy for breast cancer. Cancer 1993;71:3054 – 3057. Neugut AI, Murray T, Santos J, et al. Increased risk of lung cancer after breast cancer radiation therapy in cigarette smokers. Cancer 1994;73:1615–1620. Obedian E, Fischer DB, Haffty BG. Second malignancies after treatment of early-stage breast cancer: Lumpectomy and radiation therapy versus mastectomy. J Clin Oncol 2000;18: 2406 –2412. van Leeuwen FE, Klokman WJ, Stovall M, et al. Roles of radiotherapy and smoking in lung cancer following Hodgkin’s disease. J Natl Cancer Inst 1995;87:1530 –1537. Mauch PM, Kalish LA, Marcus KC, et al. Second malignancies after treatment for laparotomy staged IA-IIIB Hodgkin’s Volume 52, Number 2, 2002 13. 14. 15. 16. 17. 18. 19. 20. disease: Long-term analysis of risk factors and outcome. Blood 1996;87:3625–3632. Tucker MA, Coleman CN, Cox RS, et al. Risk of second cancers after treatment for Hodgkin’s disease. N Engl J Med 1988;318:76 – 81. Henschke CI, McCauley DI, Yankelevitz DF, et al. Early Lung Cancer Action Project: Overall design and findings from baseline screening. Lancet 1999;354:99 –105. Schrag D, Kuntz KM, Garber JE, et al. Decision analysis— Effects of prophylactic mastectomy and oophorectomy on life expectancy among women with BRCA1 or BRCA2 mutations. N Engl J Med 1997;336:1465–1471. Grann VR, Panageas KS, Whang W, et al. Decision analysis of prophylactic mastectomy and oophorectomy in BRCA1positive or BRCA2-positive patients. J Clin Oncol 1998;16: 979 –985. Hartmann LC, Schaid DJ, Woods JE, et al. Efficacy of bilateral prophylactic mastectomy in women with a family history of breast cancer. N Engl J Med 1999;340:77– 84. Schrag D, Kuntz KM, Garber JE, et al. Life expectancy gains from cancer prevention strategies for women with breast cancer and BRCA1 or BRCA2 mutations. JAMA 2000; 283:617– 624. Taghian A, de Vathaire F, Terrier P, et al. Long-term risk of sarcoma following radiation treatment for breast cancer. Int J Radiat Oncol Biol Phys 1991;21:361–367. Pierce SM, Recht A, Lingos TI, et al. Long-term radiation complications following conservative surgery (CS) and radiation therapy (RT) in patients with early stage breast cancer. Int J Radiat Oncol Biol Phys 1992;23:915–923.