- Email: [email protected]
Sarcoma as a second malignancy after treatment for breast cancer
Int. J. Radiation Oncology Biol. Phys., Vol. 52, No. 5, pp. 1231–1237, 2002 Copyright © 2002 Elsevier Science Inc. Printed in the USA. All rights reserved 0360-3016/02/$–see front matter
PII S0360-3016(01)02799-7
CLINICAL INVESTIGATION
Breast
SARCOMA AS A SECOND MALIGNANCY AFTER TREATMENT FOR BREAST CANCER JOHNNY YAP, M.D.,* PAUL J. CHUBA, M.D., PH.D.,† RON THOMAS, PH.D.,‡ AMR AREF, M.D.,† DAVID LUCAS, M.D.,§ RICHARD K. SEVERSON, PH.D.,㛳 AND MERLIN HAMRE, M.D., M.P.H.‡ Departments of *Radiation Oncology, ‡Pediatrics, §Pathology, and 㛳Family Medicine, Wayne State University School of Medicine, Detroit, MI; †Department of Radiation Oncology, Saint John Hospital and Medical Center, Detroit, MI Background: Second malignant neoplasms may be a consequence of radiotherapy for the treatment of breast cancer. Prior studies evaluating sarcomas as second malignant neoplasms in breast cancer patients have been limited by the numbers of patients and relatively low incidence of sarcoma. Using data from the Surveillance, Epidemiology and End Results registries, we evaluated the influence of radiation therapy on the development of subsequent sarcomas in cases with primary breast cancer. Methods: Cases with primary invasive breast cancer (n ⴝ 274,572) were identified in the Surveillance, Epidemiology and End Results Cancer Incidence Public-Use Database (1973–1997). The database was then queried to determine the cases developing subsequent sarcomas (n ⴝ 263). Eighty-seven of these cases received radiation therapy, and 176 had no radiation therapy. The cumulative incidence of developing secondary sarcoma and the survival post developing secondary sarcoma were determined by the Kaplan–Meier method. Results: The occurrence of sarcoma was low, regardless of whether cases received or did not receive radiation therapy: 3.2 per 1,000 (SE [standard error] ⴝ 0.4) and 2.3 per 1,000 (SE ⴝ 0.2) cumulative incidence at 15 years post diagnosis, respectively (p ⴝ 0.001). Of the sarcomas occurring within the field of radiation, angiosarcoma accounted for 56.8%, compared to only 5.7% of angiosarcomas occurring in cases not receiving radiotherapy. The cumulative incidence of angiosarcoma at 15 years post diagnosis was 0.9 per 1,000 for cases receiving radiation (SE ⴝ 0.2) and 0.1 per 1,000 for cases not receiving radiation (SE < 0.1). Overall survival was poor for cases of sarcoma after breast cancer (27–35% at 5 years), but not significantly different between patients receiving or not receiving radiation therapy for their primary breast cancer. Conclusions: Radiotherapy in the treatment of breast cancer is associated with an increased risk of subsequent sarcoma, but the magnitude of this risk is small. Angiosarcoma is significantly more prevalent in cases treated with radiotherapy, occurring especially in or adjacent to the radiation field. The small difference in risk of subsequent sarcoma for breast cancer patients receiving radiotherapy does not supersede the benefit of radiotherapy. © 2002 Elsevier Science Inc. Radiotherapy, Breast cancer, Sarcoma, SEER, Second malignant neoplasms.
INTRODUCTION
breast cancer. Of major interest were the histologies encountered, the cumulative incidence rates, and the latency periods for sarcomas occurring with or without radiotherapy given as part of breast cancer treatment.
Radiotherapy (RT) is considered a risk factor in breast cancer patients for the development of soft-tissue and bone sarcomas (1–5). Because of the low incidence of sarcomas, the absolute level of risk is not known with certainty. Prior studies examining radiation-associated sarcoma in breast cancer patients have been based on relatively small numbers of cases. Nonetheless, with new indications for RT and increased screening, more and younger patients are having multimodality treatment (including radiotherapy) for breast cancer (6). Using Surveillance, Epidemiology and End Result (SEER) data (7), we attempted to better define risk levels and the impact of radiation therapy on the development of sarcoma in women previously diagnosed as having
METHODS AND MATERIALS Women having a diagnosis of invasive breast cancer were identified (n ⫽ 274,572) in the SEER Cancer Incidence Public-Use Database, 1973–1996 (7). Excluded from the analysis were cases initially identified based on autopsy reports or death certificates, cases for which breast cancer was diagnosed as a subsequent malignant neoplasm, and male breast cancers. Mixed mullerian tumor and carcino-
Reprint requests to: Paul J. Chuba, M.D., Ph.D., Dept. of Radiation Oncology, St. John Hospital and Medical Center, 22101 Moross, Detroit, MI 48236. Tel: (313) 343-3750; Fax: (313) 3435186.
Presented in part at the American Society of Clinical Oncology meeting in Atlanta, Georgia, May, 1999. Received Jul 11, 2001, and in revised form Nov 20, 2001. Accepted for publication Nov 21, 2001. 1231
1232
I. J. Radiation Oncology
● Biology ● Physics
Volume 52, Number 5, 2002
Table 1. Characteristics of primary breast cancer cases No RT
Age at diagnosis ⬍20 20–39 40–59 60–79 ⱖ80 Race Caucasian African-American Other Year of diagnosis 1973–1977 1978–1982 1983–1987 1988–1992 1993–1997 All cases
RT
No.
(%)
No.
(%)
22 12,826 67,962 85,172 26,294
(0.01) (6.67) (35.35) (44.30) (13.68)
1 6,875 36,419 34,890 4,111
(0.00) (8.35) (44.25) (42.40) (5.00)
163,756 13,808 14,712
(85.17) (7.18) (7.65)
68,655 6,432 7,209
(83.42) (7.82) (8.76)
29,201 35,540 42,589 44,992 39,954 192,276
(15.19) (18.48) (22.15) (23.40) (20.78)
9,663 9,183 14,098 19,542 29,810 82,296
(11.74) (11.16) (17.13) (23.75) (36.22)
Univariate p value
Multivariate p value
⬍0.0001
⬍0.0001
⬍0.0001
0.08
⬍0.0001
⬍0.0001
⬍0.0001*
Notes: Primary breast cancers ascertained in the SEER database are tabulated by age at diagnosis, race, and year of diagnosis. The (column) percentages indicate the proportion of cases for any subcategory represented among the total cases in that category (age, race, and year of diagnosis). * Whole model.
sarcoma were excluded from our classification as secondary sarcomas. The database was then queried to identify cases subsequently diagnosed with a sarcoma at 1 or more years after diagnosis of primary breast cancer. Cases diagnosed at less than 1 year from diagnosis were excluded to eliminate the possibility of concurrent diagnosis. The cases developing sarcoma were grouped as receiving or not receiving radiotherapy. For cases receiving radiotherapy, sarcomas were defined as occurring within the field of radiation if the primary site of the subsequent sarcoma was within, or in the proximity of, the chest or breast (See below). The most prevalent subsequent sarcomas were grouped by ICD histology codes (8) as fibrosarcoma (8810), malignant fibrous histiocytoma (8802, 8830), dermatofibrosarcoma (8832), liposarcoma (8850-5), leiomyosarcoma (8890-1), angiosarcoma, including lymphangiosarcoma and hemangiosarcoma (9120, 9170), osteosarcoma (9180-1), chondrosarcoma (9220, 9231, 9240), endometrial stromal sarcoma (8930-1), and not otherwise specified (8800, 8801). Remaining histologies occurred with low frequency and were classified as “other” (codes 8900-1, 8940-1, 8951, 9040-2, 9044, 9140, 9250). Cumulative incidence was determined by the Kaplan– Meier method (9), with differences between groups assessed by the log–rank method (9 –11). Multivariate analysis was performed as previously described (10, 11). Contingency tables were evaluated by chi-squared test (10, 11). RESULTS The characteristics of primary breast cancer cases studied are shown in Table 1. Of the 274,572 cases of invasive
breast cancer identified in the SEER database 1973–1997, 82,296 (30%) received radiation therapy, and 192,276 (70%) did not receive radiation therapy. In 5,232 cases, it was unknown whether patients received radiation therapy; these were excluded from analysis. Few breast cancers were diagnosed in patients under the age of 20 (n ⫽ 23), and only about 7% of patients were aged 20 –39 years. Approximately 85% of patients were Caucasian, 7% African-American, and 8% other races. Breast conservation and the use of radiation therapy were more prevalent during the later time periods (Table 1; cf. Ref. 6). In cases diagnosed between 1993 and 1997, 29,810 out of 69,764 (42.7%) received radiation therapy, compared with 19,542 out of 64,534 (30.3%) between 1988 and 1992 and 14,098 out of 56,687 (24.9%) between 1983 and 1987 (Table 1). Overall, radiation therapy was used less in the elderly. Receiving radiation were 36,419 out of 104,381 (34.8%) women aged 40 –59, 34,890 out of 120,062 (29.0%) women aged 60 –79, and 4,111 out of 30,405 (13.5%) women aged 80⫹ (Table 1). Two hundred sixty-three cases were identified as having a subsequent sarcoma, regardless of site, 1 year or more after the diagnosis of breast cancer. Of these cases, 176 (67%) had not received radiation, and 87 (33%) had received radiation. The median latent period for developing second malignant sarcoma was 6 years (range: 1–21 years). Table 2 shows the histology of subsequent sarcomas occurring in the breast cancer cases (with ICD-0 codes collapsed to 10 individual sarcoma subtypes). The most prevalent sarcoma category occurring after breast cancer was leiomyosarcoma (22.1%), followed by malignant fibrous histiocytoma (15.2%), angiosarcoma (13.7%), and liposarcoma
Sarcoma after breast cancer
● J. YAP et al.
1233
Table 2. Secondary sarcomas developing after primary breast cancer* Secondary sarcoma†
ICD-0 codes
Fibrosarcoma MFH DFSP Liposarcoma Leiomyosarcoma Angiosarcoma Osteosarcoma Chondrosarcoma ESS Not specified
Other Totals
8810 8802, 8830 8832 8850-5 8890-1 9120,9170 9180-1 9220, 9231, 9240 8930-1 8800, 8801, 8900–8901, 8940-1, 8951, 9044, 9140 9040–2,9250
No RT (no. [%])
RT out-of-field (no. [%])
RT in-field (no. [%])
All cases (no. [%])
Cases of primary sarcoma‡
6 (3.4) 25 (14.2) 13 (7.4) 20 (11.4) 48 (27.3) 10 (5.7) 6 (3.4) 6 (3.4) 12 (6.8) 20 (11.4)
0 (0.0) 8 (18.6) 1 (2.3) 3 (7.0) 8 (18.6) 1 (2.3) 0 (0.0) 4 (9.3) 3 (7.0) 7 (16.3)
3 (6.8) 7 (15.9) 0 (0.0) 0 (0.0) 2 (4.5) 25 (56.8) 1 (2.3) 3 (6.8) 0 (0.0) 2 (4.5)
9 (3.4) 40 (15.2) 14 (5.3) 23 (8.7) 58 (22.1) 36 (13.7) 7 (2.7) 13 (4.9) 15 (5.7) 29 (11.0)
466 (2.9) 1567 (9.8) 1115 (6.9) 1173 (7.3) 3588 (22.4) 344 (2.1) 357 (2.2) 560 (3.5) 792 (4.9) 1182 (7.4)
8 (18.6) 43
1 (2.3) 44
10 (5.7) 176
19 (7.2) 263
925 (5.8) 16046
* Absolute number and percent of sarcomas occurring after primary breast cancer are tabulated. The (column) percentages indicate the proportion of the total number of secondary sarcomas for any category (no RT, in-field, out-of-field, all cases). † Secondary sarcomas were scored as occurring within (or adjacent to) or outside the radiation field, as described under “Materials and Methods.” ‡ For comparison purposes, the percentage incidence of each histology was also determined using all cases of primary sarcoma (n ⫽ 16,046) in the SEER database. Abbreviations: MFH ⫽ malignant fibrous histiocytoma; DFSP ⫽ dermatofibrosarcoma protruberans; ESS ⫽ endometrial stromal sarcoma.
(8.7%). Dermatofibrosarcoma protruberans (DFSP), a relatively low-grade neoplasm, made up 5.3% of the secondary sarcomas. Osteosarcoma and chondrosarcoma, taken together, made up 7.6% of cases. These proportions were closely comparable to those observed for primary sarcomas (n ⫽ 16,046) in the SEER database (Table 2), with the
significant exception of angiosarcoma (2.1% overall and 13.7% after breast cancer). For further analysis, the secondary sarcomas were grouped according to the anatomic locations of occurrence (Table 3). Thoracic and breast sarcoma locations were classified as “in-field” (i.e., within or adjacent to the radiation
Table 3. Anatomic locations of secondary sarcoma* Anatomic location †
Head and neck Thoracic‡ Breast§ Abdominal㛳 Upper extremity¶ Lower extremity** Pelvis†† Other Total
No RT (no. [%])
RT out-of-field (no. [%])
RT in-field (no. [%])
All RT cases (no. [%])
Total
11 (6.3) 25 (14.2) 5 (2.8) 37 (21.0) 17 (9.7) 35 (19.9) 42 (23.9) 4 (2.3) 176
4 (9.3) 0 (0.0) 0 (0.0) 8 (18.6) 5 (11.6) 11 (25.6) 15 (34.9) 0 (0.0) 43
0 (0.0) 28 (63.6) 11 (25.0) 0 (0.0) 5 (11.4) 0 (0.0) 0 (0.0) 0 (0.0) 44
4 (4.6) 28 (32.2) 11 (12.6) 8 (9.2) 10 (11.5) 11 (12.6) 15 (17.2) 0 (0.0) 87
15 (5.7) 53 (20.2) 16 (6.1) 45 (17.1) 27 (10.3) 46 (17.5) 57 (21.7) 4 (1.5) 263
* Absolute number and percent of sarcomas are tabulated according to anatomic location. The (column) percentages indicate the proportion of the total number of sarcomas for any category (no RT, out-of-field, in-field, all cases). Secondary sarcomas were scored as occurring within (or adjacent to) or outside the radiation field, as described under “Materials and Methods.” † Category includes cases categorized as parotid gland, mandible, larynx, skin of the head, skull, and soft tissue of the head, face, or neck. ‡ Includes cases categorized as lung, soft tissue of the thorax or trunk, pleural, rib, sternum, clavicle, and skin of the thorax or trunk. Breast sarcoma is excluded from this category. § Includes all breast sarcomas only. 㛳 Includes cases categorized as colon, gallbladder, stomach intestinal tract, kidney, peritoneum, retroperitoneum, small intestine, soft tissue of the abdomen. ¶ Includes cases categorized as skin of the upper extremity, soft tissue of the upper extremity, and long bones of the upper extremity. ** Includes skin of the lower extremity, soft tissue of the lower extremity, and long bones of the lower extremity. †† Includes cases categorized as uterus, uterine cervix, uterine corpus, endometrium, female genital tract, myometrium ovary, soft tissue of the pelvis, and vagina.
1234
I. J. Radiation Oncology
● Biology ● Physics
Volume 52, Number 5, 2002
Fig. 1. (a) Cumulative incidence of sarcoma in patients after primary breast cancer. The incidence of second malignant sarcoma per 1,000 patients having radiotherapy (hash-marked line) or no radiotherapy for breast cancer treatment (solid line) is plotted against time. (b) Cumulative incidence of angiosarcoma after primary breast cancer. The incidence of second malignant angiosarcoma per 1,000 patients having radiotherapy (hash-marked line) or no radiotherapy (solid line) for breast cancer treatment.
treatment portal); upper extremity lesions were classified as in-field or out-of-field based on whether they were ipsilateral to the primary breast cancer. Other locations were classified as out-of-field. Cases were approximately equally distributed between in-field and out-of-field classifications. Of the secondary sarcoma patients not receiving radiotherapy, the most prevalent histology was leiomyosarcoma (27.3%), followed by malignant fibrous histiocytoma (14.2%). DFSP totaled 7.4% of cases, whereas 6.8% were osteosarcoma/chondrosarcoma, and 5.7% were angiosarcoma. This contrasted sharply with relative numbers of histologies in patients having in-field radiotherapy, of which 56.8% were angiosarcoma (Table 2). Out of 36 angiosarcomas, 25 occurred in-field. The increase in angiosarcoma in
patients having in-field radiotherapy was compensated for by the relative decreases in liposarcoma (0%), leiomyosarcoma (4.5%), and DFSP (0%). Leiomyosarcoma, as expected because of anatomic location, was more common outside the field of radiation (Table 2). Fifteen out of 58 leiomyosarcomas were uterine (data not shown). The cumulative incidence of sarcoma in the breast cancer cases is shown in Fig. 1a. The cumulative incidence of sarcoma at 15 years for cases receiving radiation therapy, regardless of site, was 3.2 per 1,000 (SE [standard error] ⫽ 0.4) compared to 2.3 per 1,000 (SE ⫽ 0.2, p ⫽ 0.001) without radiotherapy. The cumulative incidence of angiosarcoma was determined separately for breast cancer cases receiving radiation therapy, as shown in Fig. 1b. The cu-
Sarcoma after breast cancer
● J. YAP et al.
1235
Fig. 2. Overall survival after diagnosis of sarcoma in breast cancer patients treated with or without radiotherapy. Kaplan–Meier survival curves are shown for sarcoma treated with in-field radiation (solid line) and out-of-field radiation (hash-marked line) vs. without radiation (interlaced line). p ⫽ 0.6.
mulative incidence of angiosarcoma at 15 years was 0.9 per 1,000 for cases receiving radiation (SE ⫽ 0.2) and 0.1 per 1,000 for cases not receiving radiation (SE ⬍ 0.1). All but two of the angiosarcomas occurred within 3 to 7 years from diagnosis of the primary breast cancer. The remaining two occurred at 1.9 and 12.5 years from diagnosis. The overall survival of primary breast cancer cases after the development of a secondary sarcoma is shown in Fig. 2. Survival rates from diagnosis of second sarcoma were similar for cases that received radiotherapy (in-field or out-offield secondary sarcomas) or did not receive radiotherapy (p ⫽ 0.6). Five-year survival rates were 27.7% for sarcomas occurring within the radiation field (n ⫽ 46), 30.1% for outside the radiation field (n ⫽ 76), and 35% (n ⫽ 272) for nonradiotherapy cases. The differences observed were not statistically significant (p ⫽ 0.60). The median survival after diagnosis of sarcoma was 2.3 years for all patients. DISCUSSION Ionizing radiation is known to be a potent carcinogen (12); malignancy induced by radiation may result from natural sources or a radiation accident, or it may be a side effect of cancer therapy (13). Solid cancers associated with exposure to ionizing radiation may have latency periods of up to 40 years (13). Approximately 6.7% of new cancers represent second primary cancers in cancer survivors (14), and breast cancer survivors constitute up to 25% of all cancer survivors (15, 16). In addition to cancer recurrence and second malignant neoplasms, breast cancer survivors remain at long-term risk for other late effects of therapy, such as heart disease, premature menopause, and economic and psychosocial problems (14, 15, 16). A number of investigators have pointed out an increased
incidence of soft-tissue and bone sarcoma after treatment for breast cancer (1–5, 17–24). This may occur in the breast after breast conservation therapy or in the chest wall after mastectomy, with or without postoperative adjuvant radiotherapy. Large single-institution retrospective studies have estimated cumulative incidence rates for breast cancer patients receiving radiation at about 2 per 1,000 at 10 years. The Gustave Roussy Institute study included 6,919 patients followed for more than 1 year, with 11 radiation-induced sarcomas identified (19). Cumulative incidences of 0.2% at 10 years, 0.43% at 20 years, and 0.78% at 30 years were reported. The latency period ranged from 4 to 24 years (mean: 9.5 years). Remarkably similar rates of radiationassociated sarcoma have been reported in different institutional studies. Pierce et al. (2) found three in-field sarcomas in 1,624 patients at the Joint Center for Radiation Therapy and calculated a crude incidence of 0.18%; Hatfield and Schulz (3) estimated a rate of 0.22% based on 5 radiationinduced sarcomas in 2,250 patients, and Phillips and Sheline (4) found a similar rate of 0.22% with 1 sarcoma out of 445 patients. Doherty et al. (5) described 4 in-field bone sarcomas for 3,199 patients with 18 –28-year follow-up (0.26% percentage risk). Zucali et al. (17) reported 3 cases of soft-tissue sarcoma among 3,295 patients (2 out of 3 were angiosarcoma). Kurtz et al. (18) found 2 out of 2,850 patients with in-field soft-tissue sarcoma among patients treated with breast conservation therapy in Marseilles. Previous population studies indicate that breast cancer patients who do not receive radiotherapy are also at increased risk of second malignant sarcoma. Both the data from Denmark (Ref. 21, n ⫽ 54,964, 1943–1980) and from the Connecticut tumor registry (Ref. 22, n ⫽ 41,109, 1935– 1982) would suggest a relative risk of about 2.1 to 2.3 for second malignant sarcoma in all breast cancer survivors. For
1236
I. J. Radiation Oncology
● Biology ● Physics
patients receiving radiotherapy, the relative risk of sarcoma was approximately doubled, although in-field and out-offield occurrences could not be distinguished. Karlsson et al. reported 19 sarcomas vs. 8.7 expected for the Swedish Cancer Registry data (1960 –1980, n ⫽ 13,490) (24). A total of 263 subsequent sarcomas were identified within 274,572 breast cancer patients identified from the SEER data in the current study. For patients receiving RT, the cumulative incidence of second malignant sarcoma was 3.2 per 1,000 at 15 years compared to 2.3 per 1,000 in patients not receiving radiotherapy. Risk factors for second malignancy in unirradiated patients may be related to underlying genetic susceptibility, environmental exposures, chemotherapeutic agents, or to surgically related treatmentrelated sequelae, such as lymphedema or stasis. Lymphedema has been implicated in the Stewart-Treves syndrome (25) of lymphangiosarcoma occurring postmastectomy. Whether lymphangiosarcoma formation is the result of lymph stasis or postmastectomy radiation (or some combination of the two) is still unknown. Importantly, angiosarcoma may also occur after breast conservation therapy. The subject of angiosarcoma in the intact breast was recently reviewed in detail by Marchal et al. (20), who surveyed cases at 11 French national cancer centers. In the world’s literature, these authors identified 52 cases of angiosarcoma in irradiated intact breasts reported before December 1997. Of 18,115 breast carcinomas treated using a conservation approach, 9 post-therapy breast angiosarcomas were reported by the French centers. In the current study, angiosarcoma (including hemangiosarcoma, angiosarcoma, and lymphangiosarcoma) constituted 5.7% of all histologies encountered for unirradiated patients compared with 56.8% of all histologies occurring within a previously radiated field for breast cancer (p ⬍ 0.0001 [Table 1]). This level of risk is supported by a literature review by Pendlebury et al. (1), who found 46% of in-field soft-tissue sarcomas to be angiosarcomas. In the Swedish Cancer Registry data (23, 24), 6 out of 18 secondary sarcomas were angiosarcomas. Five out of 6 angiosarcomas had lymphedema, and 6 out of 6 had received radiotherapy. These findings favor a causative or a synergistic role for radiation. A number of genetic mechanisms for the formation of and/or predisposition to sarcoma formation may exist. Exposure to radiation and/or chemical carcinogens may be superimposed on pre-existing genetic susceptibility. In Li-
Volume 52, Number 5, 2002
Fraumeni syndrome (26, 27), a condition with associated germline mutation in the tumor suppressor gene p53, patients become highly susceptible at a young age to the development of sarcomas and other cancers, including leukemia, brain tumors, and breast cancer (10 –13). Similarly, the retinoblastoma tumor suppressor gene Rb is related to predisposition to osteogenic sarcoma among patients affected by bilateral retinoblastoma (28). Interestingly, the osteosarcomas are more common in the field of radiation, but also occur with an increased frequency in nonirradiated sites in cases with a germline mutation. Although not specifically associated with sarcomas, the genes BRCA1 and BRCA2 may confer increased sensitivity to the carcinogenic effects of radiation with an elevated risk of second primary breast cancers (29). BRCA1-deficient stem cells are hypersensitive to ionizing radiation and seem to be important in the rad50-mediated response to DNA repair (30). It has been proposed that radiation-induced sarcomas are biologically more aggressive than their spontaneously occurring counterparts. We failed to identify any trend toward poorer survival among breast cancer patients with sarcomas that were within the radiation field 5 years from diagnosis. Five-year survival for sarcoma subsequent to breast cancer was generally poor (27–35%). In summary, our observations indicate that breast cancer patients receiving radiation therapy have a statistically increased risk of developing subsequent sarcoma. However, the magnitude of this difference is very small. Because the incidence rates of radiation-induced cancer are low, an accurate estimate of these rates requires a large number of cases. Limitations of the SEER data include the unavailability of information related to the type of chemotherapy received, radiation energy or dose, and information regarding the exact radiation portals. Ascertainment and follow-up rates are relatively high, and survivals can be accurately estimated. Long-term risk associated with cancer therapies takes on increased importance when disease control and survival rates are high, as with early breast cancer. With effective screening, an increasing number of patients will undergo conservative and multimodality approaches to breast cancer treatment and be at risk for radiation-induced sarcoma. It is important to recognize that radiotherapy is an effective treatment modality, and the small difference in these incidence rates does not supersede the benefit of radiotherapy.
REFERENCES 1. Pendlebury SC, Bilous M, Langlands AO. Sarcomas following radiation therapy for breast cancer: A report of three cases and a review of the literature. Int J Radiat Oncol Biol Phys 1995;31:405– 410. 2. Pierce SM, Recht A, Lingos TI, et al. Long term radiation complications following conservative surgery and radiation in patients with early stage breast cancer. Int J Radiat Oncol Biol Phys 1992;23:915–923. 3. Hatfield PM, Schulz MD. Post-irradiation sarcoma including 5
cases after X ray therapy of breast carcinoma. Radiology 1970;96:593– 602. 4. Phillips TL, Sheline GE. Bone sarcoma following radiotherapy. Radiology 1993;81:992–996. 5. Doherty MA, Rodger A, Langlands AO. Sarcoma of the bone following therapeutic irradiation for breast carcinoma. Int J Radiat Oncol Biol Phys 1985;12:103–106. 6. Chuba PJ, Simon MS. Trends in primary surgical and radiation therapy for localized breast cancer in the Detroit Metro-
Sarcoma after breast cancer
7.
8. 9. 10. 11. 12.
13. 14.
15. 16. 17. 18.
19.
politan area 1973–1992;. Int J Radiat Oncol Biol Phys 1997: 38:103–107. Surveillance, Epidemiology, and End Results (SEER) Program Public-Use CD-ROM (1973–1996), National Cancer Institute, DCCPS, Cancer Surveillance Research Program, Cancer Statistics Branch, April, 1999. Percy C, Can Holten V, Muir C, editors. International classification of diseases for oncology. 2nd edition. Geneva (Switzerland): World Health Organization; 1990. Kaplan EL, Meier P. Non-parametric estimation from incomplete observations. J Am Stat Assoc 1958;53:457– 481. Lee ET. Statistical methods for survival analysis. Belmont, CA: Lifetime Learning Publications; 1980. Snedecor GW, Cochran WG. Statistical methods. 7th edition. The Iowa State University Press. 1980. Chakborty R, Little MP, Sancaranarayanan K. Cancer predisposition, radiosensitivity, and the risk of radiation induced cancers. IV: Prediction of risk in relatives of cancer-predisposed individuals. Radiat Res 1998;149:493–507. Health effects of exposure to low levels of ionizing radiation (BEIR V). Washington, DC: National Research Council; 1990. Meadows AT, Varrichio C, Crosson K, et al. Research issues in cancer survivorship: Report of a workshop sponsored by the Office of Cancer Survivorship, National Cancer Institute. Cancer Epidemiol Biomarkers Prev 1998;7:1145–1151. Ganz PA, Greendale GA, Kahn B, et al. Are older breast cancer survivors willing to have hormone replacement therapy? Cancer 1999;86:814 – 820. Landis SH, Murray T, Bolden S, Wingo PA. Cancer statistics 1999. CA Cancer J Clin 1999;49:8 –31. Zucali R, Merson M, Placucci M, et al. Soft tissue sarcoma of the breast after conservative surgery and irradiation for early mammary cancer. Radiother Oncol 1994;30:271–273. Kurtz JM, Almaric R, Brandone H, et al. Contralateral breast cancer and other second malignancies in patients treated by breast-conserving therapy with radiation. Int J Radiat Oncol Biol Phys 1988;15:277–284. Taghian A, De Vathaire F, Terrier P, et al. Long-term risk of
● J. YAP et al.
20.
21.
22.
23.
24.
25.
26.
27. 28.
29.
30.
1237
sarcoma following radiation treatment for breast cancer. Int J Radiat Oncol Biol Phys 1991;21:361–397. Marchal C, Weber B, De LaFontan B, et al. Nine breast angiosarcomas after conservative treatment for breast carcinoma: A survey from French comprehensive cancer centers. Int J Radiat Oncol Biol Phys 1999;44:113–119. Ewertz M, Mouridsen H. Second cancer following cancer of the female breast in Denmark 1943–1980. National Cancer Inst Monograph 1985;68:325–329. Harvey E, Brinton L. Second cancer following cancer of the breast in Connecticut 1935– 82. National Cancer Inst Monograph 1981;68:99 –112. Karlsson P, Holmberg E, Samuelsson A, et al. Soft tissue sarcoma after treatment for breast cancer—a Swedish population-based study. Eur J Cancer 1998;34:2068 –2075. Karlsson P, Holmberg E, Johansson KA, et al. Soft tissue sarcoma after treatment for breast cancer. Radiother Oncol 1996;38:25–31. Stewart FW, Treves N. Classics in oncology: Lymphangiosarcoma in postmastectomy lymphedema. A report of six cases in elephantiasis chirurgica. CA Cancer J Clin 1981;31: 284 –299. Li FP, Fraumeni JF Jr, Mulvihill JJ, et al. A cancer family syndrome in twenty-four kindreds. Cancer Res 1988;48:5358 – 5362. Li FP, Fraumeni JF Jr. Familial breast cancer, soft-tissue sarcomas, and other neoplasms. Ann Intern Med 1975;83:833. Chuba PJ, Hamre M, Blackburn R, et al. Second malignant neoplasms in retinoblastoma (Abstr). Eur J Cancer 1999; 35(Suppl. 4):S112. Turner BC, Harrold E, Matloff E, et al. BRCA1/BRCA2 germline mutations in locally recurrent breast cancer patients after lumpectomy and radiation therapy: Implications for breast-conserving management in patients with BRCA1/ BRCA2 mutations. J Clin Oncol 1999;17:3017–3024. Zhong Q, Chen C, Li S, et al. Association of BRCA1 with the hRad50-hMre11-p95 complex and the DNA damage response. Science 1999;285:747–750.
PII S0360-3016(01)02799-7
CLINICAL INVESTIGATION
Breast
SARCOMA AS A SECOND MALIGNANCY AFTER TREATMENT FOR BREAST CANCER JOHNNY YAP, M.D.,* PAUL J. CHUBA, M.D., PH.D.,† RON THOMAS, PH.D.,‡ AMR AREF, M.D.,† DAVID LUCAS, M.D.,§ RICHARD K. SEVERSON, PH.D.,㛳 AND MERLIN HAMRE, M.D., M.P.H.‡ Departments of *Radiation Oncology, ‡Pediatrics, §Pathology, and 㛳Family Medicine, Wayne State University School of Medicine, Detroit, MI; †Department of Radiation Oncology, Saint John Hospital and Medical Center, Detroit, MI Background: Second malignant neoplasms may be a consequence of radiotherapy for the treatment of breast cancer. Prior studies evaluating sarcomas as second malignant neoplasms in breast cancer patients have been limited by the numbers of patients and relatively low incidence of sarcoma. Using data from the Surveillance, Epidemiology and End Results registries, we evaluated the influence of radiation therapy on the development of subsequent sarcomas in cases with primary breast cancer. Methods: Cases with primary invasive breast cancer (n ⴝ 274,572) were identified in the Surveillance, Epidemiology and End Results Cancer Incidence Public-Use Database (1973–1997). The database was then queried to determine the cases developing subsequent sarcomas (n ⴝ 263). Eighty-seven of these cases received radiation therapy, and 176 had no radiation therapy. The cumulative incidence of developing secondary sarcoma and the survival post developing secondary sarcoma were determined by the Kaplan–Meier method. Results: The occurrence of sarcoma was low, regardless of whether cases received or did not receive radiation therapy: 3.2 per 1,000 (SE [standard error] ⴝ 0.4) and 2.3 per 1,000 (SE ⴝ 0.2) cumulative incidence at 15 years post diagnosis, respectively (p ⴝ 0.001). Of the sarcomas occurring within the field of radiation, angiosarcoma accounted for 56.8%, compared to only 5.7% of angiosarcomas occurring in cases not receiving radiotherapy. The cumulative incidence of angiosarcoma at 15 years post diagnosis was 0.9 per 1,000 for cases receiving radiation (SE ⴝ 0.2) and 0.1 per 1,000 for cases not receiving radiation (SE < 0.1). Overall survival was poor for cases of sarcoma after breast cancer (27–35% at 5 years), but not significantly different between patients receiving or not receiving radiation therapy for their primary breast cancer. Conclusions: Radiotherapy in the treatment of breast cancer is associated with an increased risk of subsequent sarcoma, but the magnitude of this risk is small. Angiosarcoma is significantly more prevalent in cases treated with radiotherapy, occurring especially in or adjacent to the radiation field. The small difference in risk of subsequent sarcoma for breast cancer patients receiving radiotherapy does not supersede the benefit of radiotherapy. © 2002 Elsevier Science Inc. Radiotherapy, Breast cancer, Sarcoma, SEER, Second malignant neoplasms.
INTRODUCTION
breast cancer. Of major interest were the histologies encountered, the cumulative incidence rates, and the latency periods for sarcomas occurring with or without radiotherapy given as part of breast cancer treatment.
Radiotherapy (RT) is considered a risk factor in breast cancer patients for the development of soft-tissue and bone sarcomas (1–5). Because of the low incidence of sarcomas, the absolute level of risk is not known with certainty. Prior studies examining radiation-associated sarcoma in breast cancer patients have been based on relatively small numbers of cases. Nonetheless, with new indications for RT and increased screening, more and younger patients are having multimodality treatment (including radiotherapy) for breast cancer (6). Using Surveillance, Epidemiology and End Result (SEER) data (7), we attempted to better define risk levels and the impact of radiation therapy on the development of sarcoma in women previously diagnosed as having
METHODS AND MATERIALS Women having a diagnosis of invasive breast cancer were identified (n ⫽ 274,572) in the SEER Cancer Incidence Public-Use Database, 1973–1996 (7). Excluded from the analysis were cases initially identified based on autopsy reports or death certificates, cases for which breast cancer was diagnosed as a subsequent malignant neoplasm, and male breast cancers. Mixed mullerian tumor and carcino-
Reprint requests to: Paul J. Chuba, M.D., Ph.D., Dept. of Radiation Oncology, St. John Hospital and Medical Center, 22101 Moross, Detroit, MI 48236. Tel: (313) 343-3750; Fax: (313) 3435186.
Presented in part at the American Society of Clinical Oncology meeting in Atlanta, Georgia, May, 1999. Received Jul 11, 2001, and in revised form Nov 20, 2001. Accepted for publication Nov 21, 2001. 1231
1232
I. J. Radiation Oncology
● Biology ● Physics
Volume 52, Number 5, 2002
Table 1. Characteristics of primary breast cancer cases No RT
Age at diagnosis ⬍20 20–39 40–59 60–79 ⱖ80 Race Caucasian African-American Other Year of diagnosis 1973–1977 1978–1982 1983–1987 1988–1992 1993–1997 All cases
RT
No.
(%)
No.
(%)
22 12,826 67,962 85,172 26,294
(0.01) (6.67) (35.35) (44.30) (13.68)
1 6,875 36,419 34,890 4,111
(0.00) (8.35) (44.25) (42.40) (5.00)
163,756 13,808 14,712
(85.17) (7.18) (7.65)
68,655 6,432 7,209
(83.42) (7.82) (8.76)
29,201 35,540 42,589 44,992 39,954 192,276
(15.19) (18.48) (22.15) (23.40) (20.78)
9,663 9,183 14,098 19,542 29,810 82,296
(11.74) (11.16) (17.13) (23.75) (36.22)
Univariate p value
Multivariate p value
⬍0.0001
⬍0.0001
⬍0.0001
0.08
⬍0.0001
⬍0.0001
⬍0.0001*
Notes: Primary breast cancers ascertained in the SEER database are tabulated by age at diagnosis, race, and year of diagnosis. The (column) percentages indicate the proportion of cases for any subcategory represented among the total cases in that category (age, race, and year of diagnosis). * Whole model.
sarcoma were excluded from our classification as secondary sarcomas. The database was then queried to identify cases subsequently diagnosed with a sarcoma at 1 or more years after diagnosis of primary breast cancer. Cases diagnosed at less than 1 year from diagnosis were excluded to eliminate the possibility of concurrent diagnosis. The cases developing sarcoma were grouped as receiving or not receiving radiotherapy. For cases receiving radiotherapy, sarcomas were defined as occurring within the field of radiation if the primary site of the subsequent sarcoma was within, or in the proximity of, the chest or breast (See below). The most prevalent subsequent sarcomas were grouped by ICD histology codes (8) as fibrosarcoma (8810), malignant fibrous histiocytoma (8802, 8830), dermatofibrosarcoma (8832), liposarcoma (8850-5), leiomyosarcoma (8890-1), angiosarcoma, including lymphangiosarcoma and hemangiosarcoma (9120, 9170), osteosarcoma (9180-1), chondrosarcoma (9220, 9231, 9240), endometrial stromal sarcoma (8930-1), and not otherwise specified (8800, 8801). Remaining histologies occurred with low frequency and were classified as “other” (codes 8900-1, 8940-1, 8951, 9040-2, 9044, 9140, 9250). Cumulative incidence was determined by the Kaplan– Meier method (9), with differences between groups assessed by the log–rank method (9 –11). Multivariate analysis was performed as previously described (10, 11). Contingency tables were evaluated by chi-squared test (10, 11). RESULTS The characteristics of primary breast cancer cases studied are shown in Table 1. Of the 274,572 cases of invasive
breast cancer identified in the SEER database 1973–1997, 82,296 (30%) received radiation therapy, and 192,276 (70%) did not receive radiation therapy. In 5,232 cases, it was unknown whether patients received radiation therapy; these were excluded from analysis. Few breast cancers were diagnosed in patients under the age of 20 (n ⫽ 23), and only about 7% of patients were aged 20 –39 years. Approximately 85% of patients were Caucasian, 7% African-American, and 8% other races. Breast conservation and the use of radiation therapy were more prevalent during the later time periods (Table 1; cf. Ref. 6). In cases diagnosed between 1993 and 1997, 29,810 out of 69,764 (42.7%) received radiation therapy, compared with 19,542 out of 64,534 (30.3%) between 1988 and 1992 and 14,098 out of 56,687 (24.9%) between 1983 and 1987 (Table 1). Overall, radiation therapy was used less in the elderly. Receiving radiation were 36,419 out of 104,381 (34.8%) women aged 40 –59, 34,890 out of 120,062 (29.0%) women aged 60 –79, and 4,111 out of 30,405 (13.5%) women aged 80⫹ (Table 1). Two hundred sixty-three cases were identified as having a subsequent sarcoma, regardless of site, 1 year or more after the diagnosis of breast cancer. Of these cases, 176 (67%) had not received radiation, and 87 (33%) had received radiation. The median latent period for developing second malignant sarcoma was 6 years (range: 1–21 years). Table 2 shows the histology of subsequent sarcomas occurring in the breast cancer cases (with ICD-0 codes collapsed to 10 individual sarcoma subtypes). The most prevalent sarcoma category occurring after breast cancer was leiomyosarcoma (22.1%), followed by malignant fibrous histiocytoma (15.2%), angiosarcoma (13.7%), and liposarcoma
Sarcoma after breast cancer
● J. YAP et al.
1233
Table 2. Secondary sarcomas developing after primary breast cancer* Secondary sarcoma†
ICD-0 codes
Fibrosarcoma MFH DFSP Liposarcoma Leiomyosarcoma Angiosarcoma Osteosarcoma Chondrosarcoma ESS Not specified
Other Totals
8810 8802, 8830 8832 8850-5 8890-1 9120,9170 9180-1 9220, 9231, 9240 8930-1 8800, 8801, 8900–8901, 8940-1, 8951, 9044, 9140 9040–2,9250
No RT (no. [%])
RT out-of-field (no. [%])
RT in-field (no. [%])
All cases (no. [%])
Cases of primary sarcoma‡
6 (3.4) 25 (14.2) 13 (7.4) 20 (11.4) 48 (27.3) 10 (5.7) 6 (3.4) 6 (3.4) 12 (6.8) 20 (11.4)
0 (0.0) 8 (18.6) 1 (2.3) 3 (7.0) 8 (18.6) 1 (2.3) 0 (0.0) 4 (9.3) 3 (7.0) 7 (16.3)
3 (6.8) 7 (15.9) 0 (0.0) 0 (0.0) 2 (4.5) 25 (56.8) 1 (2.3) 3 (6.8) 0 (0.0) 2 (4.5)
9 (3.4) 40 (15.2) 14 (5.3) 23 (8.7) 58 (22.1) 36 (13.7) 7 (2.7) 13 (4.9) 15 (5.7) 29 (11.0)
466 (2.9) 1567 (9.8) 1115 (6.9) 1173 (7.3) 3588 (22.4) 344 (2.1) 357 (2.2) 560 (3.5) 792 (4.9) 1182 (7.4)
8 (18.6) 43
1 (2.3) 44
10 (5.7) 176
19 (7.2) 263
925 (5.8) 16046
* Absolute number and percent of sarcomas occurring after primary breast cancer are tabulated. The (column) percentages indicate the proportion of the total number of secondary sarcomas for any category (no RT, in-field, out-of-field, all cases). † Secondary sarcomas were scored as occurring within (or adjacent to) or outside the radiation field, as described under “Materials and Methods.” ‡ For comparison purposes, the percentage incidence of each histology was also determined using all cases of primary sarcoma (n ⫽ 16,046) in the SEER database. Abbreviations: MFH ⫽ malignant fibrous histiocytoma; DFSP ⫽ dermatofibrosarcoma protruberans; ESS ⫽ endometrial stromal sarcoma.
(8.7%). Dermatofibrosarcoma protruberans (DFSP), a relatively low-grade neoplasm, made up 5.3% of the secondary sarcomas. Osteosarcoma and chondrosarcoma, taken together, made up 7.6% of cases. These proportions were closely comparable to those observed for primary sarcomas (n ⫽ 16,046) in the SEER database (Table 2), with the
significant exception of angiosarcoma (2.1% overall and 13.7% after breast cancer). For further analysis, the secondary sarcomas were grouped according to the anatomic locations of occurrence (Table 3). Thoracic and breast sarcoma locations were classified as “in-field” (i.e., within or adjacent to the radiation
Table 3. Anatomic locations of secondary sarcoma* Anatomic location †
Head and neck Thoracic‡ Breast§ Abdominal㛳 Upper extremity¶ Lower extremity** Pelvis†† Other Total
No RT (no. [%])
RT out-of-field (no. [%])
RT in-field (no. [%])
All RT cases (no. [%])
Total
11 (6.3) 25 (14.2) 5 (2.8) 37 (21.0) 17 (9.7) 35 (19.9) 42 (23.9) 4 (2.3) 176
4 (9.3) 0 (0.0) 0 (0.0) 8 (18.6) 5 (11.6) 11 (25.6) 15 (34.9) 0 (0.0) 43
0 (0.0) 28 (63.6) 11 (25.0) 0 (0.0) 5 (11.4) 0 (0.0) 0 (0.0) 0 (0.0) 44
4 (4.6) 28 (32.2) 11 (12.6) 8 (9.2) 10 (11.5) 11 (12.6) 15 (17.2) 0 (0.0) 87
15 (5.7) 53 (20.2) 16 (6.1) 45 (17.1) 27 (10.3) 46 (17.5) 57 (21.7) 4 (1.5) 263
* Absolute number and percent of sarcomas are tabulated according to anatomic location. The (column) percentages indicate the proportion of the total number of sarcomas for any category (no RT, out-of-field, in-field, all cases). Secondary sarcomas were scored as occurring within (or adjacent to) or outside the radiation field, as described under “Materials and Methods.” † Category includes cases categorized as parotid gland, mandible, larynx, skin of the head, skull, and soft tissue of the head, face, or neck. ‡ Includes cases categorized as lung, soft tissue of the thorax or trunk, pleural, rib, sternum, clavicle, and skin of the thorax or trunk. Breast sarcoma is excluded from this category. § Includes all breast sarcomas only. 㛳 Includes cases categorized as colon, gallbladder, stomach intestinal tract, kidney, peritoneum, retroperitoneum, small intestine, soft tissue of the abdomen. ¶ Includes cases categorized as skin of the upper extremity, soft tissue of the upper extremity, and long bones of the upper extremity. ** Includes skin of the lower extremity, soft tissue of the lower extremity, and long bones of the lower extremity. †† Includes cases categorized as uterus, uterine cervix, uterine corpus, endometrium, female genital tract, myometrium ovary, soft tissue of the pelvis, and vagina.
1234
I. J. Radiation Oncology
● Biology ● Physics
Volume 52, Number 5, 2002
Fig. 1. (a) Cumulative incidence of sarcoma in patients after primary breast cancer. The incidence of second malignant sarcoma per 1,000 patients having radiotherapy (hash-marked line) or no radiotherapy for breast cancer treatment (solid line) is plotted against time. (b) Cumulative incidence of angiosarcoma after primary breast cancer. The incidence of second malignant angiosarcoma per 1,000 patients having radiotherapy (hash-marked line) or no radiotherapy (solid line) for breast cancer treatment.
treatment portal); upper extremity lesions were classified as in-field or out-of-field based on whether they were ipsilateral to the primary breast cancer. Other locations were classified as out-of-field. Cases were approximately equally distributed between in-field and out-of-field classifications. Of the secondary sarcoma patients not receiving radiotherapy, the most prevalent histology was leiomyosarcoma (27.3%), followed by malignant fibrous histiocytoma (14.2%). DFSP totaled 7.4% of cases, whereas 6.8% were osteosarcoma/chondrosarcoma, and 5.7% were angiosarcoma. This contrasted sharply with relative numbers of histologies in patients having in-field radiotherapy, of which 56.8% were angiosarcoma (Table 2). Out of 36 angiosarcomas, 25 occurred in-field. The increase in angiosarcoma in
patients having in-field radiotherapy was compensated for by the relative decreases in liposarcoma (0%), leiomyosarcoma (4.5%), and DFSP (0%). Leiomyosarcoma, as expected because of anatomic location, was more common outside the field of radiation (Table 2). Fifteen out of 58 leiomyosarcomas were uterine (data not shown). The cumulative incidence of sarcoma in the breast cancer cases is shown in Fig. 1a. The cumulative incidence of sarcoma at 15 years for cases receiving radiation therapy, regardless of site, was 3.2 per 1,000 (SE [standard error] ⫽ 0.4) compared to 2.3 per 1,000 (SE ⫽ 0.2, p ⫽ 0.001) without radiotherapy. The cumulative incidence of angiosarcoma was determined separately for breast cancer cases receiving radiation therapy, as shown in Fig. 1b. The cu-
Sarcoma after breast cancer
● J. YAP et al.
1235
Fig. 2. Overall survival after diagnosis of sarcoma in breast cancer patients treated with or without radiotherapy. Kaplan–Meier survival curves are shown for sarcoma treated with in-field radiation (solid line) and out-of-field radiation (hash-marked line) vs. without radiation (interlaced line). p ⫽ 0.6.
mulative incidence of angiosarcoma at 15 years was 0.9 per 1,000 for cases receiving radiation (SE ⫽ 0.2) and 0.1 per 1,000 for cases not receiving radiation (SE ⬍ 0.1). All but two of the angiosarcomas occurred within 3 to 7 years from diagnosis of the primary breast cancer. The remaining two occurred at 1.9 and 12.5 years from diagnosis. The overall survival of primary breast cancer cases after the development of a secondary sarcoma is shown in Fig. 2. Survival rates from diagnosis of second sarcoma were similar for cases that received radiotherapy (in-field or out-offield secondary sarcomas) or did not receive radiotherapy (p ⫽ 0.6). Five-year survival rates were 27.7% for sarcomas occurring within the radiation field (n ⫽ 46), 30.1% for outside the radiation field (n ⫽ 76), and 35% (n ⫽ 272) for nonradiotherapy cases. The differences observed were not statistically significant (p ⫽ 0.60). The median survival after diagnosis of sarcoma was 2.3 years for all patients. DISCUSSION Ionizing radiation is known to be a potent carcinogen (12); malignancy induced by radiation may result from natural sources or a radiation accident, or it may be a side effect of cancer therapy (13). Solid cancers associated with exposure to ionizing radiation may have latency periods of up to 40 years (13). Approximately 6.7% of new cancers represent second primary cancers in cancer survivors (14), and breast cancer survivors constitute up to 25% of all cancer survivors (15, 16). In addition to cancer recurrence and second malignant neoplasms, breast cancer survivors remain at long-term risk for other late effects of therapy, such as heart disease, premature menopause, and economic and psychosocial problems (14, 15, 16). A number of investigators have pointed out an increased
incidence of soft-tissue and bone sarcoma after treatment for breast cancer (1–5, 17–24). This may occur in the breast after breast conservation therapy or in the chest wall after mastectomy, with or without postoperative adjuvant radiotherapy. Large single-institution retrospective studies have estimated cumulative incidence rates for breast cancer patients receiving radiation at about 2 per 1,000 at 10 years. The Gustave Roussy Institute study included 6,919 patients followed for more than 1 year, with 11 radiation-induced sarcomas identified (19). Cumulative incidences of 0.2% at 10 years, 0.43% at 20 years, and 0.78% at 30 years were reported. The latency period ranged from 4 to 24 years (mean: 9.5 years). Remarkably similar rates of radiationassociated sarcoma have been reported in different institutional studies. Pierce et al. (2) found three in-field sarcomas in 1,624 patients at the Joint Center for Radiation Therapy and calculated a crude incidence of 0.18%; Hatfield and Schulz (3) estimated a rate of 0.22% based on 5 radiationinduced sarcomas in 2,250 patients, and Phillips and Sheline (4) found a similar rate of 0.22% with 1 sarcoma out of 445 patients. Doherty et al. (5) described 4 in-field bone sarcomas for 3,199 patients with 18 –28-year follow-up (0.26% percentage risk). Zucali et al. (17) reported 3 cases of soft-tissue sarcoma among 3,295 patients (2 out of 3 were angiosarcoma). Kurtz et al. (18) found 2 out of 2,850 patients with in-field soft-tissue sarcoma among patients treated with breast conservation therapy in Marseilles. Previous population studies indicate that breast cancer patients who do not receive radiotherapy are also at increased risk of second malignant sarcoma. Both the data from Denmark (Ref. 21, n ⫽ 54,964, 1943–1980) and from the Connecticut tumor registry (Ref. 22, n ⫽ 41,109, 1935– 1982) would suggest a relative risk of about 2.1 to 2.3 for second malignant sarcoma in all breast cancer survivors. For
1236
I. J. Radiation Oncology
● Biology ● Physics
patients receiving radiotherapy, the relative risk of sarcoma was approximately doubled, although in-field and out-offield occurrences could not be distinguished. Karlsson et al. reported 19 sarcomas vs. 8.7 expected for the Swedish Cancer Registry data (1960 –1980, n ⫽ 13,490) (24). A total of 263 subsequent sarcomas were identified within 274,572 breast cancer patients identified from the SEER data in the current study. For patients receiving RT, the cumulative incidence of second malignant sarcoma was 3.2 per 1,000 at 15 years compared to 2.3 per 1,000 in patients not receiving radiotherapy. Risk factors for second malignancy in unirradiated patients may be related to underlying genetic susceptibility, environmental exposures, chemotherapeutic agents, or to surgically related treatmentrelated sequelae, such as lymphedema or stasis. Lymphedema has been implicated in the Stewart-Treves syndrome (25) of lymphangiosarcoma occurring postmastectomy. Whether lymphangiosarcoma formation is the result of lymph stasis or postmastectomy radiation (or some combination of the two) is still unknown. Importantly, angiosarcoma may also occur after breast conservation therapy. The subject of angiosarcoma in the intact breast was recently reviewed in detail by Marchal et al. (20), who surveyed cases at 11 French national cancer centers. In the world’s literature, these authors identified 52 cases of angiosarcoma in irradiated intact breasts reported before December 1997. Of 18,115 breast carcinomas treated using a conservation approach, 9 post-therapy breast angiosarcomas were reported by the French centers. In the current study, angiosarcoma (including hemangiosarcoma, angiosarcoma, and lymphangiosarcoma) constituted 5.7% of all histologies encountered for unirradiated patients compared with 56.8% of all histologies occurring within a previously radiated field for breast cancer (p ⬍ 0.0001 [Table 1]). This level of risk is supported by a literature review by Pendlebury et al. (1), who found 46% of in-field soft-tissue sarcomas to be angiosarcomas. In the Swedish Cancer Registry data (23, 24), 6 out of 18 secondary sarcomas were angiosarcomas. Five out of 6 angiosarcomas had lymphedema, and 6 out of 6 had received radiotherapy. These findings favor a causative or a synergistic role for radiation. A number of genetic mechanisms for the formation of and/or predisposition to sarcoma formation may exist. Exposure to radiation and/or chemical carcinogens may be superimposed on pre-existing genetic susceptibility. In Li-
Volume 52, Number 5, 2002
Fraumeni syndrome (26, 27), a condition with associated germline mutation in the tumor suppressor gene p53, patients become highly susceptible at a young age to the development of sarcomas and other cancers, including leukemia, brain tumors, and breast cancer (10 –13). Similarly, the retinoblastoma tumor suppressor gene Rb is related to predisposition to osteogenic sarcoma among patients affected by bilateral retinoblastoma (28). Interestingly, the osteosarcomas are more common in the field of radiation, but also occur with an increased frequency in nonirradiated sites in cases with a germline mutation. Although not specifically associated with sarcomas, the genes BRCA1 and BRCA2 may confer increased sensitivity to the carcinogenic effects of radiation with an elevated risk of second primary breast cancers (29). BRCA1-deficient stem cells are hypersensitive to ionizing radiation and seem to be important in the rad50-mediated response to DNA repair (30). It has been proposed that radiation-induced sarcomas are biologically more aggressive than their spontaneously occurring counterparts. We failed to identify any trend toward poorer survival among breast cancer patients with sarcomas that were within the radiation field 5 years from diagnosis. Five-year survival for sarcoma subsequent to breast cancer was generally poor (27–35%). In summary, our observations indicate that breast cancer patients receiving radiation therapy have a statistically increased risk of developing subsequent sarcoma. However, the magnitude of this difference is very small. Because the incidence rates of radiation-induced cancer are low, an accurate estimate of these rates requires a large number of cases. Limitations of the SEER data include the unavailability of information related to the type of chemotherapy received, radiation energy or dose, and information regarding the exact radiation portals. Ascertainment and follow-up rates are relatively high, and survivals can be accurately estimated. Long-term risk associated with cancer therapies takes on increased importance when disease control and survival rates are high, as with early breast cancer. With effective screening, an increasing number of patients will undergo conservative and multimodality approaches to breast cancer treatment and be at risk for radiation-induced sarcoma. It is important to recognize that radiotherapy is an effective treatment modality, and the small difference in these incidence rates does not supersede the benefit of radiotherapy.
REFERENCES 1. Pendlebury SC, Bilous M, Langlands AO. Sarcomas following radiation therapy for breast cancer: A report of three cases and a review of the literature. Int J Radiat Oncol Biol Phys 1995;31:405– 410. 2. Pierce SM, Recht A, Lingos TI, et al. Long term radiation complications following conservative surgery and radiation in patients with early stage breast cancer. Int J Radiat Oncol Biol Phys 1992;23:915–923. 3. Hatfield PM, Schulz MD. Post-irradiation sarcoma including 5
cases after X ray therapy of breast carcinoma. Radiology 1970;96:593– 602. 4. Phillips TL, Sheline GE. Bone sarcoma following radiotherapy. Radiology 1993;81:992–996. 5. Doherty MA, Rodger A, Langlands AO. Sarcoma of the bone following therapeutic irradiation for breast carcinoma. Int J Radiat Oncol Biol Phys 1985;12:103–106. 6. Chuba PJ, Simon MS. Trends in primary surgical and radiation therapy for localized breast cancer in the Detroit Metro-
Sarcoma after breast cancer
7.
8. 9. 10. 11. 12.
13. 14.
15. 16. 17. 18.
19.
politan area 1973–1992;. Int J Radiat Oncol Biol Phys 1997: 38:103–107. Surveillance, Epidemiology, and End Results (SEER) Program Public-Use CD-ROM (1973–1996), National Cancer Institute, DCCPS, Cancer Surveillance Research Program, Cancer Statistics Branch, April, 1999. Percy C, Can Holten V, Muir C, editors. International classification of diseases for oncology. 2nd edition. Geneva (Switzerland): World Health Organization; 1990. Kaplan EL, Meier P. Non-parametric estimation from incomplete observations. J Am Stat Assoc 1958;53:457– 481. Lee ET. Statistical methods for survival analysis. Belmont, CA: Lifetime Learning Publications; 1980. Snedecor GW, Cochran WG. Statistical methods. 7th edition. The Iowa State University Press. 1980. Chakborty R, Little MP, Sancaranarayanan K. Cancer predisposition, radiosensitivity, and the risk of radiation induced cancers. IV: Prediction of risk in relatives of cancer-predisposed individuals. Radiat Res 1998;149:493–507. Health effects of exposure to low levels of ionizing radiation (BEIR V). Washington, DC: National Research Council; 1990. Meadows AT, Varrichio C, Crosson K, et al. Research issues in cancer survivorship: Report of a workshop sponsored by the Office of Cancer Survivorship, National Cancer Institute. Cancer Epidemiol Biomarkers Prev 1998;7:1145–1151. Ganz PA, Greendale GA, Kahn B, et al. Are older breast cancer survivors willing to have hormone replacement therapy? Cancer 1999;86:814 – 820. Landis SH, Murray T, Bolden S, Wingo PA. Cancer statistics 1999. CA Cancer J Clin 1999;49:8 –31. Zucali R, Merson M, Placucci M, et al. Soft tissue sarcoma of the breast after conservative surgery and irradiation for early mammary cancer. Radiother Oncol 1994;30:271–273. Kurtz JM, Almaric R, Brandone H, et al. Contralateral breast cancer and other second malignancies in patients treated by breast-conserving therapy with radiation. Int J Radiat Oncol Biol Phys 1988;15:277–284. Taghian A, De Vathaire F, Terrier P, et al. Long-term risk of
● J. YAP et al.
20.
21.
22.
23.
24.
25.
26.
27. 28.
29.
30.
1237
sarcoma following radiation treatment for breast cancer. Int J Radiat Oncol Biol Phys 1991;21:361–397. Marchal C, Weber B, De LaFontan B, et al. Nine breast angiosarcomas after conservative treatment for breast carcinoma: A survey from French comprehensive cancer centers. Int J Radiat Oncol Biol Phys 1999;44:113–119. Ewertz M, Mouridsen H. Second cancer following cancer of the female breast in Denmark 1943–1980. National Cancer Inst Monograph 1985;68:325–329. Harvey E, Brinton L. Second cancer following cancer of the breast in Connecticut 1935– 82. National Cancer Inst Monograph 1981;68:99 –112. Karlsson P, Holmberg E, Samuelsson A, et al. Soft tissue sarcoma after treatment for breast cancer—a Swedish population-based study. Eur J Cancer 1998;34:2068 –2075. Karlsson P, Holmberg E, Johansson KA, et al. Soft tissue sarcoma after treatment for breast cancer. Radiother Oncol 1996;38:25–31. Stewart FW, Treves N. Classics in oncology: Lymphangiosarcoma in postmastectomy lymphedema. A report of six cases in elephantiasis chirurgica. CA Cancer J Clin 1981;31: 284 –299. Li FP, Fraumeni JF Jr, Mulvihill JJ, et al. A cancer family syndrome in twenty-four kindreds. Cancer Res 1988;48:5358 – 5362. Li FP, Fraumeni JF Jr. Familial breast cancer, soft-tissue sarcomas, and other neoplasms. Ann Intern Med 1975;83:833. Chuba PJ, Hamre M, Blackburn R, et al. Second malignant neoplasms in retinoblastoma (Abstr). Eur J Cancer 1999; 35(Suppl. 4):S112. Turner BC, Harrold E, Matloff E, et al. BRCA1/BRCA2 germline mutations in locally recurrent breast cancer patients after lumpectomy and radiation therapy: Implications for breast-conserving management in patients with BRCA1/ BRCA2 mutations. J Clin Oncol 1999;17:3017–3024. Zhong Q, Chen C, Li S, et al. Association of BRCA1 with the hRad50-hMre11-p95 complex and the DNA damage response. Science 1999;285:747–750.