Contralateral breast cancer and other second malignancies in patients treated by breast-conserving therapy with radiation

Inr. J. Radiation Oncology Biol. Phys.. Vol. 15. pp. 277-284 Printed in the U.S.A. All rights reserved.

0360-3016/88 $3.00 + .OO Copyright 0 1988 Pergamon Press plc

??Original Contribution

CONTRALATERAL BREAST CANCER AND OTHER SECOND MALIGNANCIES IN PATIENTS TREATED BY BREAST-CONSERVING THERAPY WITH RADIATION

JOHN M. KURTZ, M.D.,2 ROBERT AMALRIC, M.D.,’ HENRI BRANDONE, M.D.,’ YVES AYME, M.D.,’ AND JEAN-MAURICE SPITALIER, M.D.’

‘Departments of Surgery and Radiotherapy, Institut Paoli-Calmettes, Marseille, France; and 2Department of Radiation Oncology, University Hospital, Basel, Switzerland

Metachronous contralateral breast cancers and other second malignancies were evaluated in 2,850 patients treated between 1960 and 1981 primarily with radiotherapy (RT) either alone or following breast-conserving surgery. One hundred eighty-four contralateral cancers were observed in 22,491 patient-years of observation (818 per 10’ patient-years), with a cumulative probability of 4.5% at 57.9% at 10, and 11% at 15 and 20 years. Compared to patients with unilateral tumors, those destined to develop contralateral cancers were younger (mean age 51.9 vs 56.6) and more often gave a family history of breast cancer. Contralateral breast cancers were more frequent for more extensive tumors (T3 10% vs Tl-2 6%; with inflammatory signs 10.6% without 6%), and in patients with ipsilateral local recurrence (with 9.1%, without 5.6%). Patients with contralateral cancers had a significantly less favorable survival experience (15-year actuarial survival after primary therapy 42%) than patients without contralateral cancer (15-year survival 65.5%). In early stage patients treated with conservative surgery and RT, contralateral cancer was not propostically more favorable than ipsilateral breast recurrence. Among 72 other second malignancies (320 per 10 patient-years) were 2 soft tissue sarcomas in the irradiated area. This corresponds to an incidence of 21 cases per 10’ patient-years for survivors beyond the fifth year. The possible influence of RT on contralateral cancers and other second malignancies is discussed.

Contralateral breast cancer, Breast-conserving therapy, Radiation-induced cancer.

INTRODUCTION

Breast-conserving therapy differs from the primary radical surgical management of mammary carcinoma not only in the preservation of the diseased breast, but in the requirement for breast irradiation. Since radiation is an established breast carcinogen,” and many patients treated by primary radical surgery do not require radiotherapy (RT), it might be argued that patients who are treated with breast conservation are subjected to serious potential risks not necessarily shared by their surgically- treated counterparts.

Discounting the local side effects of RT, which are sel- dom of functional consequence, these risks fall into three categories: first, the possible induction of contralateral breast cancer due to scattered radiation to the opposite breast;‘,16 Secondly, the induction of sarcomas in the high-dose region of the breast and chest wall;“Y3’ Finally, a potential increase in the rate of second malignancies in other anatomical regions might be postulated.” The scant literature on this subject is based upon studies of

atomic bomb survivors,25 on the use of X ray therapy for benign diseases,6*8,20,30 or upon the use of post-operative therapy following mastectomy, often using ortho-voltage or imprecisely-defined treatments.7.‘6,17,24,29,34 To our knowledge, the question of second malignancies has never been evaluated formally in a large group of patients treated exclusively with breast conservation using mega-voltage therapy.

METHODS AND MATERIALS

The study group comprises 2,850 patients with operable breast cancers in clinical Stages I-III, treated between September 1960 and December 1981 at the Can- cer Institute and associated clinics in Marseille, using RT alone or following a breast-conserving operative proce- dure. According to the staging system of the American Joint Committee,’ 680 were in clinical Stage I, 1,743 in Stage II, and 427 in Stage III. One thousand four hundred eighty-three patients, primarily in Stages T2N 1 and III, were treated with RT alone, following cytologic or

Presented at the Annual Meeting of the American Society Reprint requests to: J.-M. Spitalier, M.D., Clairval Saugeaie, for Therapeutic Radiology and Oncology, Boston, October 22, 36, Bd du Redon, F-l 3009, Marseille, France. 1987. Accepted for publication 2 March 1988.

211

278 I. J. Radiation Oncology 0 Biology ??Physics August 1988, Volume 15, Number 2

histologic confirmation of malignancy. The remainder of patients ( 1,367) were irradiated following surgical re- moval of the primary tumor, with or without limited ax- illary dissection. The evolution of treatment policies at Marseille, a description of techniques, as well as the long- term results of therapy, have been presented in previous publications.2-4,32

RT to breast and nodal areas was delivered using a tel- ecobalt unit in 1,116 patients and a telecesium unit in 1,734 patients. Telecesium therapy was used primarily before 1976, and in a minority of patients until 1979. Supplemental irradiation was always given to the tumor bed, the lower axilla, and selectively to the internal mammary area, usually using lo- 15 Mev electrons from a be- tatron. The primary boost to the tumor bed typically delivered 20-25 Gy in 2-2.5 weeks to the reference isodose.

Breast RT was delivered using 2 opposed tangential fields, with the patients in the lateral decubitus position. This technique has previously been described in detail and illustrated photographically,3x4 being similar to the technique developed at Institut Curie by Baclesse.36 Us- ing this technique, the breast flattens out into a rather uniform thickness, so that a central-axis midplane dose calculation can be used with open fields. In this fashion 50 Gy in 25 fractions in 5 weeks were delivered on teleco- balt, or 60 Gy in 30 fractions in 6 weeks on telecesium. It is evident that the chest wall receives a lesser dose using this technique, depending upon the patient’s anatomical configuration. Typical isodose distributions have been published.3.4

An estimation of the dose delivered to the opposite breast was performed on a patient during actual treatment on 6oCo. Thermo-luminescent dosimeters were placed on the skin of the contralateral breast at 5 cm and 10 cm from the medial edge of the tangential field. Aver- age measurements of 4.8% and 10.8% of the midplane central axis dose were registered at the lateral and medial points, respectively. This corresponds to 2.4 Gy and 5.4 Gy for a treatment course of 50 Gy. The lateral decubitus position does not permit convenient measurement on a phantom for estimation of dose at depth within the breast.

Patient data are maintained in a personal computer by one of the senior authors (JMS), with continuous updat- ing of each patient’s status, including registration of local-regional recurrences, metastases, contralateral breast cancers, and other second malignancies. All patients in this study have a minimum follow-up of 5 years for living patients with a maximum of 26 years.

Actuarial calculations were performed using the life- table method, as recommended by the American Joint Committeee5 Incidence densities were computed as described by Hankey et al. I6 As computed from life tables, the total follow-up for the 2,850 patients amounts to 22,49 1 patient-years, including 9,148 patient-years beyond the fifth year, 2,447 patient-years beyond the tenth year, and 413 patient-years beyond the fifteenth year.

Breast cancer-specific survival (hereafter referred to sim- ply as “actuarial survival”) was calculated by treating patients lost to follow-up or dead of intercurrent disease while clinically free of cancer as censored observations at the time of last consultation. Patients dying of unknown cause were counted as cancer deaths. Differences between actuarial curves were tested for significance using the logrank test.26

RESULTS

Contralateral breast cancers (CBC) From the study group of 2,850 patients, one hundred

eighty-four (6.5%) were observed to have developed proven metachronous cancer in the opposite breast, including 90 of the 1,483 patients (6.1%) treated with RT alone and 94 of the 1,367 patients (6.9%) treated with RT following breast-conserving surgery. This represents an incidence density of 8 18 cases per 100,000 patient- years. Patients with synchronous bilateral cancers had been systematically excluded from registration at the beginning of RT. An additional 15 patients gave a history of prior treatment of the opposite breast. The cumulative actuarial probability of contralateral breast cancer (CBC) as a function of time was 4.5% at 5 years, 7.9% at 10 years, and 11% at 15 years. There were no CBC observed after the fifteenth year.

Patients developing CBC tended to be younger (mean age 5 1.9 years) compared to patients with unilateral cancer (mean age 56.6 years). Potential risk factors are tabulated in Table 1. Of 1,828 patients for whom information

Table 1. Potential risk factors for contralateral breast cancers

Contralateral breast cancer

Yes No Relative

risk pvalue

Age <45 245

Family history of breast cancer?

Yes No

Clinical tumor size >5 CM 15 CM

Inflammatory signs Yes No

Tumor location Medial/Central Lateral

Ipsilateral breast recurrence

Yes No

55 608 1.41 129 2058 1 .o* <0.05

34 382 1.28 90 1322 1 .o* >o. 1

40 367 1.66 144 2299 1 .o* <O.Ol

17 144 1.71 167 2522 1 .o* <0.05

59 701 1.30 125 1965 1.0* 0.1

64 637 1.64 120 2029 1 .o* <o.oo 1

* Reference category. t Only available for 1828 patients.

Second malignancies after breast RT 0 J. M. KURTZ et al. 219

regarding family breast cancer history had been recorded, 8.2% of patients with a positive family history developed CBC, compared with 6.4% of patients with a negative family history. There was a tendency for larger tumors to become bilateral (T3, 10% versus Tl-2, 6%), as well as tumors with localized erythema or edema (10.6%). Patients with centrally or medially located tumors had a tendency to have more CBC than did patients with cancers in the lateral half of the breast.

Patients with CBC were more likely to have ipsilateral breast recurrence. Of the patients treated by breast-conserving surgery and RT, 16/94 ( 17%) with CBC developed ipsilateral breast recurrence compared with 136/ 1,273 (10.7%) of patients without CBC. For patients treated with RT alone, 49/90 (54%) of patients with CBC had their primary tumors uncontrolled by initial treatment, compared with 500/ 1,393 (36%) of patients without CBC.

Ultimate prognosis for patients with CBC was clearly less favorable than that of patients whose cancers re- mained unilateral. Of patients with CBC, 79/ 184 (43%) have thus far developed distant metastases, compared to 706/2,666 (26.5%) of patients with unilateral tumors. This is reflected in the actuarial survival curves as measured from time of initial therapy (Fig. l), with 15year survivals of 42% and 65.5% with and without CBC. This survival disadvantage applied to the relatively favorable patients treated by conservative surgery and RT, as well as to the more advanced cases treated with RT alone (separate curves not shown).

The prognosis of CBC was compared to that of ipsilat- era1 breast recurrence for 1,350 Stage I and II patients treated by conservative surgery and RT. As measured from the time of initial therapy, there was no significant

difference between the survival curves (Fig. 2). Contra- lateral cancer thus does not appear to represent a more favorable event than ipsilateral breast recurrence.

Other second malignancies Other second cancers were recorded in 72 of the 2,850

patients (2.5%). For 22,49 1 patient-years of observation, this corresponds to an incidence density of 320 cases per 100,000 patient-years. The types of tumors observed are listed in Table 2.

Two of the second malignancies were soft tissue sarcomas arising in the primarily irradiated area, one in the axilla and one in the para-areolar area of the breast. Both tumors were malignant fibrous histiocytomas, occurring at 5 and 5.5 years post-therapy, and leading in both in- stances to the death of the patient. One patient had also received adjuvant chemotherapy at time of primary treatment. For 22,49 1 patient-years of observation, this represents 9 cases per 100,000 patient-years.

DISCUSSION

Contralateral cancer, the most common second malignancy suffered by breast cancer patients, has been the subject of numerous publications, including hospital- based 12.15,1W4J8.29 and population_based series. 163 17.27.34

Based on data from the cancer registries of Connecticut, Denmark, and Birmingham, women having once been treated for breast cancer have a three-fold relative risk of developing a second primary breast cancer, compared with their counterparts in the population at large.‘6,17,27,34 The overall incidence density of CBC ranges from 533 to 8 18 cases per 100,000 patient-years (Table 3). This wide range of values is doubtless a reflection of numerous fac-

P < 0.01

0 11, I1 I I I u I I

0 2 4 6 a IO 12 14

YEARS FOLLOWING PRIMARY TREATMENT

-_b 2666 2250 963 220 WIIHOUI CONlRALAlERAL CANCER

- 184 161 9? 25 VI IH CONlRALAIERAL CANCER

Fig. 1. Actuarial cancer-specific survival for patients with unilateral vs metachronous contralateral breast cancers, as measured from the time of primary treatment. Numbers along the time axis represent the number of patients alive at the beginning of each interval.

280 I. J. Radiation Oncology 0 Biology 0 Physics August 1988, Volume 15, Number 2

p I NS

q , , , ( , , ( , , , , , , , ,

0 2 4 6 6 IO 12 14

TIME AFTER PRIMARY TREATMENT

- 93 87 55 19 COMIRALIIERIL BREASr CINCf - 147 I26 77 26 IPSILAIERAL ,,AWtARY RECURf

Fig. 2. Actuarial cancer-specific survival for clinical Stage I and II patients treated by breast-conserving surgery and radiotherapy, comparing patients with contralateral breast cancers and ipsilateral breast recurrence, measured from time of primary treatment. Numbers along the time axis represent the number of patients alive at the beginning of each interval. Error bars represent 95% confidence limits.

tors, including geographical differences, varying compo- sition of the populations studied, differing followup times, divergent criteria for defining CBC, and perhaps accuracy of reporting. Treatment may also play a certain role, as discussed below.

Of greater clinical usefulness than incidence density is the cumulative actuarial probability as a function of time, which is tabulated in Table 4 for various hospital- based series, including the present one. This illustrates that the risk of CBC for surviving patients remains rather constant, at least during the first 2 decades, or perhaps decreases slightly with time.

The incidence density of our series (8 18 per 100,000 patient-years) is one of the highest values yet reported.

Table 2. Second malignancies after treatment of primary breast cancer

No.

Colon and rectum Stomach Ovary Endometrium Cervix Vulva Head and neck Skin (excluding melanoma) Melanoma Kidney Non-Hodgkin’s lymphoma Acute leukemia Chronic leukemia Lung Sarcoma; breast/chest wall UnslXcii%d

9 1 7 7 7 2 3 9 2 3 1 1 1

: 16

This most likely reflects the liberal definition of CBC in this study. The problems inherent in distinguishing between “true” second primary breast cancers and metastases from the original primary tumors have been discussed by Robbins and Berg in their classic paper.28 Vari- ous authors have attempted to refine the definition by eliminating from consideration patients with recurrent disease,24 excluding patients diagnosed within an arbitrary interval following the primary cancer,7*‘8*34 or by using strict histologic criteria.7~‘8~28 Rather than resorting to such arbitrary measures, we have chosen to include all patients with histologically or cytologically proven cancer in the opposite breast, regardless of the clinical set- ting. A lower incidence density could have been obtained by making certain adjustments, since 4 1 of the 184 CBC in this series occurred during the first 12 months, and 64 of them involved patients whose ipsilateral breast disease was uncontrolled by primary therapy.

Numerous observations point to the likely role of lymphatic or hematogenous spread in the pathogenesis of some CBC. First, there is an apparent association between disease extent and risk of bilateralization. In a recent review of the subject from Institut Gustave- Roussy, I2 20% of patients presenting with distant metastases developed tumor formation in the opposite breast by 2 years, whereas 25% of T4 tumors and 22% of inflammatory cancers had contralateral manifestations by 3 years. Data from the Connecticut Tumor Registry indi- cate that patients with positive nodes have an incidence of CBC which is about one-third greater than that of negative-node patients. I6 The correlation between tumor extent and bilateralization is also demonstrable in the present series, with T3 tumors and those with localized inflammatory changes being at higher risk, despite the


Table 3. Contralateral breast cancers: Incidence density in cases per lo5 patient-years

Source

Connecticut Tumor RegistryI Irradiated Not irradiated

Danish Tumor Registry34 Irradiated Not irradiated

Memorial Hos~.~* Columbia Univ.” Marseille

Patient-years Number CBC Incidence density

63,096 544 862 208,428 1383 663

249,63 1 1172 469 95,943 668 696 12,818 91 710 12,201 87 713 22,49 1 184 818

shorter survival time generally associated with these patients. Unlike the Villejuif series,12 T4 tumors and true inflammatory cancers were not included in the present analysis.

A further indicator of the influence of the primary tumor process on CBC is the clear correlation with ipsilat- era1 breast recurrence (Table 1). It is possible that common factors lead to a greater tendency to both local recurrence and bilateralization. Alternatively, one might postulate that the inability to control the primary cancer without the use of salvage surgery may result in a higher frequency of opposite breast cancers. Of the 64 CBC associated with uncontrolled ipsilateral breast disease, 49 (76.6%) occurred in patients treated by radiotherapy alone, without primary tumor excision, a treatment method with a particularly high local recurrence rate.2 Excess contralateral cancers may represent a previously unrecognized disadvantage of certain breast-conserving techniques which are associated with inadequate local control. We have virtually abandoned radiation therapy alone as a primary treatment in recent years, in favor of breast-conserving surgery followed by RT.32

A third possible influence of the primary tumor on CBC formation concerns primary tumor location. The slightly higher frequency for tumors in the medial half of the breast (Table 2) may reflect lymphatic dissemination across the midline. A similar 30% greater risk for medial tumors has been noted by Robbins and Berg.28

The most widely studied risk factor for CBC is patient age at time of primary therapy. Patients younger than 45

Table 4. Cumulative actuarial risk of contralateral breast cancer (%)

Years after primary therapy

Number of Center patients 5 10 15 20

Memorial Hos~.~’ 1458 3.6 6.6 9.2 12.9 Columbia Univ.” 1361 3 6 11 15 London, Ontario24

Irradiated 1120 4 8 11 23 Not irradiated 369 4 11 17 17

Villejuifi2 1111 7 14 Marseille 2850 4.5 7.9 11 11

have a 5-fold risk of a second breast malignancy relative to comparable controls in the normal population.‘5*‘6,34 These elevated relative risks, however, are of little perti- nence to daily clinical practice, since they are at least par- tially a reflection of the relatively low incidence of breast cancer before the fifth decade. Thus, it is important to note that the absolute risks are higher for younger women as well. Incidence densities on the order of 1000 cases per 100,000 patient-years have been recorded for patients contracting primary breast cancer before the age of 45.15*16 Young patients with positive nodes appear to be subjected to risks as high as 1400 per lo5 patient- years.16 The higher CBC risk for younger patients was also apparent in the present series, as was the slightly higher risk for patients with a family history of breast cancer, particularly for patients whose mothers or sisters had been afflicted.155’8~21~23

Several recent authors have addressed the possible connection between contralateral breast cancer and post-operative RT. These retrospective analyses have variously concluded that the frequency in irradiated patients is similar to,24 greater than,‘6**7 or even less than that of unirradiated patients.29*34 This lack of unanimity suggests either that the contribution of therapeutic irradiation to CBC induction is small, or that any relation- ship is at best complex. This topic has recently been re- viewed with skepticism by Levitt and Mande1.22

Based on data from atomic bomb survivors, and on patients treated by X ray therapy for benign diseases 6s8,20,25,30 radiation has come to be accepted as a breas; carcinogen. According to current notions, radiation induced breast cancer requires a latent period of 10 or perhaps 20 years, affects preferentially younger women, and is dose dependent within the dose range studied (tens to hundreds of rads). For women between 20 and 40 years of age, a rate of breast cancer induction of 6.6 cases per rad per million patient-years has been estimated, beginning 10 years after radiation exposure.20 The quality of available data for patients older than 40 at time of exposure is considered inadequate for reliable estimates2’ but the risk is generally assumed to be less for older patients.

Pertaining to contralateral cancers, there have been thus far virtually no clinical data supporting this picture of breast cancer induction by therapeutic irradiation.


Data from the cancer registries of Connecticut and Den- mark provide conflicting evidence. In the Connecticut data,16,” in which RT was used in 28% of the patients treated, the incidence density and relative risk were higher overall for irradiated patients, whereas in the Dan- ish study,34 where 68% of patients were coded as having been irradiated, the overall risk was markedly lower in the irradiated patients (Table 3). Such differences are as likely a result of divergent selection criteria for RT as a direct effect of RT itself. These problems are discussed in detail by Hankey et aLI6 and by Storm and Jensen.34 Moreover, the higher risk for irradiated patients in Con- necticut (1.4 relative to nonirradiated patients) mainly reflected a risk differential occurring during the first 5 years of followup.‘6~‘7 In addition, a consistent age effect could not be shown.16 These findings are not consistent with an irradiation effect. The Connecticut data, together with the corresponding values from our series, are presented in Table 5.

The risk of CBC in irradiated Danish patients is with 469 per lo5 patient-years one of the lower rates reported in the literature.34 Although this may reflect the strict his- tological criteria apparently used in this study, it is possible that this represents in part a beneficial effect of post- operative RT. A lower risk was noted in irradiated versus unirradiated patients treated at M.D. Anderson Hospi- ta1,29 suggesting that post-operative RT to chest wall and internal mammary areas may impede subsequent dissemination of cancer across the midline. Such an effect would be expected to influence primarily the CBC rate during the first 5 or 10 years of observation, which is the case in the Danish study. That this phenomenon was not apparent in the Connecticut study may be related to divergent selection criteria for RT, which was used rather sparingly in the Connecticut patients.‘6,‘7

breast. 7V13,22,35 Technical factors influencing contralateral breast dose have been elegantly analyzed by Fraass et aZ.22 It is likely that the lateral decubitus technique used to treat the patients in the present series results in the largest opposite breast dose of any commonly used technique, because of the difficulty in shielding the opposite breast during treatment of the lateral tangent. In fact, using this technique in the treatment of synchronous bilateral cancers, we have found it necessary to reduce the applied breast dose by 10 Gy to avoid unacceptable acute skin reactions. Assuming a contralateral breast dose between 2.4 and 5.4 Gy, and applying the guidelines reported by Land,” one might expect an excess of 158 to 356 CBC per lo5 patient-years, beginning after the tenth year, at least for the patients exposed before the age of 40. However, CBC rates in this series are similar to those of other hospital-based studies (Table 4), and there is no evidence of increasing risk with time. Furthermore, the incidence density during the second decade is similar for our patients and the unirradiated Connecticut patients (Table 5). Nonetheless, the number of patient-years beyond 15 years is small, and longer follow-up will be required to precisely evaluate the possibility of later radiation effects.

Interestingly, patients followed for more than 15 years in the Danish series show a higher relative risk for irradiated than unirradiated patients. This excess risk shows an inverse dependence on age and seems to persist beyond 30 years’ follow-up. To our knowledge, these represent the only data in the literature consistent with the possible induction of CBC by therapeutic irradiation. The relative risk for the radiated patients followed beyond 10 years was 1.3 with respect of the unirradiated patients.34

Concern about cancer induction has stimulated inter- est in determination of radiation dose to the opposite

Previous studies have suggested that, stage for stage, survival after treatment of CBC is similar to that following therapy of a primary unilateral breast cancer.‘5,‘9 Nevertheless, it should be recognized that CBC represents an additional threat to survival, above that posed by the first primary. Using matched pair analysis, Rob- bins and Berg estimated this survival deficit to be about 20K2* We have chosen the simple device of calculating the actuarial survival from time of initial therapy for patients with and without CBC, as a measure of its impact on the course of the original disease. Statistically, this is subject to criticism in that a bias presumably results in favor of patients with contralateral cancers, who must have survived long enough to manifest a second tumor. Nonetheless this shows a distinct survival deficit associated with CBC. Interestingly, this deficit is similar to that associated with ipsilateral breast recurrence (Fig. 2). Al- though in our experience ipsilateral breast recurrence following conservative surgery and RT is the more common event,34 it is not justified to lavish preferential diag- nostic attention on the treated breast. Life-long surveil- lance of both breasts is mandatory.

Contralateral breast cancer represents between 40-

Table 5. Contralateral breast cancers: Comparison with data from the Connecticut Tumor Registry-Incidence density (cases/ lo5 patient-years) as function of interval

Connecticut Tumor Registry16~‘7 Irradiated Not irradiated

Marseille Cancer Institute (all irradiated)

O-4 yr

962 657

914

5-9 yr

155 691

731

IO-19 yr

753 640

533

20-29 yr

635 719

30+ yr

555 496


50% of second malignancies in breast cancer patients.10,17.34 According to Connecticut and Denmark data, other second malignancies should be expected with an incidence density of 659-695 cases per lo5 patient- years.“.‘7 An excess malignancy rate has been identified for both irradiated and unirradiated breast cancer patients for certain tumor types.17 These included melanoma, cancers of the ovary, uterus, thyroid, colon, lung, and connective tissue, as well as non-lymphocytic leuke- mias. Significantly more common among irradiated patients were cancers of the esophagus, kidney, connective tissue, non-Hodgkin’s lymphomas, chronic lymphatic leukemia, and acute non-lymphocytic leukemia. These increased risks for breast cancer patients could be con- firmed by the Danish data for melanoma, cancers of the ovary, lung, connective tissue, and acute lymphocytic leukemia.” In addition an excess of bone tumors was found among the Danish patients, for whom the data for irradiated and unirradiated subgroups were not sepa- rately tabulated. Esophageal cancer, reported elsewhere as occurring with increased frequency in irradiated breast cancer patientsI was not more common in the Danish breast cancer patients, despite the fact that 69% had received RT.

It is surprising to have identified only 72 patients with other metachronous malignancies in our series, about half the expected number. This may reflect underreport- ing of second cancers, which unlike local recurrences and contralateral cancers, might suffer from being considered less relevant oncologic events by the reporting physi- cians. There is nothing striking in the list of second tumors (Table 2) deserving of special comment, except for the 2 soft tissue sarcomas in the high-dose irradiated area, which must be accepted as treatment-induced. Both tumors occurred in the region of supplemental elec- tron beam irradiation, having received prescribed doses of 70 Gy and 80 Gy, respectively. Both were malignant fibrous histiocytomas, the most common histology for RT-induced sarcomas. ’ ‘-3’ Although numerous reports of soft tissue sarcomas following post-mastectomy RT have appeared in the literature, we are aware of only 2 previously reported cases following breast-conserving therapy.9*3’

Since this represents an extremely serious complica- tion of treatment, efforts to quantitate this risk are clearly needed. In the Connecticut data 23 soft tissue sarcomas were identified among breast cancer patients, with incidence densities of 14.3 and 6.7 per lo5 patient-years for irradiated and unirradiated patients, respectively.17 No distinction was made, however, between tumors within and without the irradiated area, although 8/23 (35%)

were lymphangiosarcomas arising from the arms and shoulders, generally considered a surgical complica- tion.33 Since a minimum 5-year latency period is presumably required for sarcoma induction to express itself, it seems more appropriate to take only the patient-years of observation beyond the fifth year into consideration. From the Connecticut data, this yields 7 sarcomas in 28,385 patient-years (24.7 per lo5 patient-years) for irradiated and 7 sarcomas per 107,823 patient-years (6.5 per lo5 patient-years) for the unirradiated patients. The excess sarcoma risk of 18.2 per lo5 patient-years corresponds well with the 2 cases we observed in 9 148 patient- years beyond 5 years (21.9 per lo5 patient-years). This implies that for every 1000 5-year survivors of breast- conserving therapy, about 2 therapy-induced soft tissue sarcomas might be expected for each decade of subsequent follow-up. Since these data are based on a small number of events, these conclusions must be viewed with some skepticism. For example, it should be mentioned that the observed sarcoma incidence density in the Dan- ish Cancer Registry (in which 68% of patients received RT) was 8.5 cases per lo5 patient-years beyond 5 years, considerably lower than in the Connecticut data.

In summary, this study of CBC in patients treated primarily by RT yielded risks similar to those observed in previously published series based on radical surgical therapy. The actuarial risk as a function of time was similar to that of other hospital-based series, and risk factors operative in our patients were similar to those previously identified in the literature. Nevertheless, based on our experience, as well as upon a review of the literature, we have identified several unique and interesting aspects of contralateral cancers as they relate to breast-conserving therapy with RT: (a) We have shown a clear association with ipsilateral breast recurrence, so that failure to control the primary disease without salvage surgery may result in an excess of contralateral tumor formation. (b) Contralateral cancer appears to have a similar adverse impact on the course of the disease as ipsilateral breast recurrence. (c) Data from the literature were presented suggesting the possibility of a reduction in contralateral cancers during the first 5- 10 years due to the routine use of RT compared to surgical therapy alone. (d) The possibility of an excess of late contralateral cancers due to RT should remain a topic for further study; our data thus far lend no support to this. (e) Finally, we have observed 2 soft tissue sarcomas in the breast or chest wall in our patients and have estimated this risk as being approxi- mately 20 per lo5 patient-years for survivors beyond the fifth year.

REFERENCES

1. Adami, H.O., BergstrBm, R., Hansen, J.: Age at first primary as a determinant in the incidence of bilateral breast cancer. Cancer 55: 643-641, 1985.

2. Amalric, R., Santamaria, F., Robert, F., Seigle, J., Alt-

schuler, C., Pietra, J.C., Amahic, F., Kurtz, J.M., Spitalier, J.M., Brandone, H., Ayme, Y., Pollet, J.F., Bressac, C., Fondarai, J.: Conservation therapy of operable breast cancer: Results at five, ten, and fifteen years in 22 16 consecu-


3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17,

18.

19.

20.

tive cases. In Conservative Management of Breast Cancer, Harris, J.R., Hellman, S., Silen, W. (Eds.). Philadelphia, J.B. Lippincott Co. 1983, pp. 15-2 1. Amalric, R., Spitalier, J.M.: Cesiumtherapie Curative des Cancers du Sein. Paris, Masson, 1973. Amalric, R., Spitalier, J.M.: Radiation as the sole method of treatment in carcinoma of the breast. In Handbuch der medizinischen Radiologie, Band XIX, Zuppinger, A., and Hellriegel, W. (Eds.). Heidelberg, Springer-Verlag. 1982, pp. 301-346. American Joint Committee: Manualfor Staging of Cancer, 2nd edition. Philadelphia, J.B. Lippincott Co. 1983, pp ll-21,127-132. Baral, E., Larsson, L.E., Mattson, B.: Breast cancer following irradiation of the breast. Cancer 40: 2905-29 10, 1977. Basco, V.E., Coldman, A.J., Elwood, J.M., Young, M.E.J.: Radiation dose and second breast cancer. Br. J. Cancer 52: 319-325,1985. Boice, J., Monson, R.: Breast cancer in women after re- peated fluoroscopic examinations of the chest. J. Nat. Can- cer Inst. 59: 823-832, 1977. Delouche, G., Bachelot, F., Premont, M., Kuttz, J.M.: Conservation treatment of early breast cancer: Long-term results and complications. Znt. J. Radiat. Oncol. Biol. Phys. 13: 29-34, 1987. Ewertz, M., Mouridsen, H.T.: Second cancer following cancer of the female breast in Denmark, 1943-80. Nat. CancerZnst. Monograph 68: 325-329, 1985. Ferguson, D.J., Sutton, H.G., Jr., Dawson, P.J.: Late effects of adjuvant radiotherapy for breast cancer. Cancer 54: 2319-2323, 1984. Fontaine, F., Contesso, G., Mouriesse, H., Travagli, J-P., Sarrazin, D: Bilatkralisation des cancers du sein. Bull. Can- cer(Paris)72: 127-138, 1985. Fraass, B.A., Roberson, P.L., Lichter, AS.: Dose to the contralateral breast due to primary breast irradiation. Znt. J. Radiat. Oncol. Biol. Phys. 11: 485-497, 1985. Goffman, T.E., McKeen, E.A., Curtis, R.E., Schein, P.S.: Esophageal carcinoma following irradiation for breast cancer. Cancer 52: 1808- 1809, 1983. Haagensen, C.D.: Bilateral carcinoma of the breast. In Breast Carcinoma: Risk and Detection, Haagensen, C.D. (Ed.). Philadelphia, W.B. Saunders Co. 198 1, pp. 339-379. Hankey, B.F., Curtis R.E., Naughton, M.D., Boice, J.D., Jr., Flannery, J.T.: A retrospective cohort analysis of second breast cancer risk for primary breast cancer patients with an assessment of the effect of radiation therapy. J. Nat. Cancer Inst. 70: 797-804, 1983. Harvey, E.B., Brinton, L.A.: Second cancer following cancer of the breast in Connecticut, 1935-82. Nat. Cancer Inst. Monog. 68: 9-l 12, 1985. Hislop, T.G., Elwood, J.M., Coldman, A.J., Spine& J.J., Worth, A.J., Ellison, L.G.: Second primary cancers of the breast: Incidence and risk factors. Br. J. Cancer 49: 79-85, 1984. Khafagy, M.M., Schottenfeld, D., Robbins, G.F.: Progno- sis of the second breast cancer. The role of previous exposure to the first primary. Cancer 35: 596-599,1975. Land, C.E.: Low-dose irradiation. A cause of breast cancer? Cancer 46: 868-873, 1980.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

Lesser, M.L., Rosen, P.P., Kinne, D.W.: Multicentricity and bilaterality in invasive breast carcinoma. Surgery 91: 234-240,1982. Levitt, S.H., Mandel, J.S.: Breast carcinogenesis: Risk of radiation. Znt. J. Radiat. Oncol. Biol. Phys. 11: 142 l-1423, 1985. Lies, H.P., Jr.: Bilateral breast cancer. Surg. Clin. North Am. 58: 833-941,1978. McCredie, J.A., Inch, W.R., Alderson, M.: Consecutive primary carcinomas of the breast. Cancer 35: 1472-1477, 1975. McGregor, D.H., Land, C.D., Choi, K., Tokuoda, S., Klu, P.L., Wakabayashi, T., Beebe, B.W.: Breast cancer incidence among atomic bomb survivors, Hiroshima and Na- gasaki, 1950- 1969. J. Nat. Cancer Inst. 59: 799-8 11,1977. Peto, R., Pike, M.C., Armitage, P., Breslow, N.E., McPher- son, K., Peto, J., Smith, P.G., Cox, D.R., Howard, V., Mantel, N.: Design and analysis of randomized clinical tri- als which require prolonged observations of each patient. II. Analysis and examples. Br. J. Cancer 35: l-39, 1977. Prior, P., Waterhouse, J.A.: Incidence of bilateral tumors in a population-based series of breast cancer patients. I. two approaches to an epidemiological analysis. Br. J. Can- cer 37: 620-634, 1978. Robbins, G.F., Berg, J.W.: Bilateral primary breast cancers. A prospective clinicopathologic study. Cancer 17: 1501-1527,1964. Schell, S.R., Montague, E.D., Spanos, W.J., Jr., Tapley, NduV., Fletcher, G.H., Oswald, M.J.: Bilateral breast cancer in patients with initial stage I and II disease. Cancer 50: 1191-l 194,1982. Shore, R.E., Hemplemann, L.H., Kowaluk, E., Mansur, P., Pastemak, B., Albert, R., Hanghie, G.: Breast neo- plasms in women treated with x-rays for acure postpartum mastitis. J. Nat. Cancer Inst. 59: 8 13-822, 1977. Souba, W.W., McKenna, R.J., Jr., Benjamin, R., Ray- mond, A.K., Mountain, C.F.: Radiation-induced sarcomas of the chest wall. Cancer 57: 6 1 O-6 15, 1986. Spitalier, J.M., Gambarelli, J., Brandone, H., Ayme, Y., Hans, D., Brandone, J.M., Bressac, C., Amalric, R., Santa- maria, F., Robert, F., Seigle, J., Amalric, F., Pollet, J.F., Kurtz, J.M.: Breast-conserving surgery with radiation therapy for operable mammary carcinoma: A 25-year experience. WorldJ. Surg. 10: 1014-1020, 1986. Stewart, F.W., Treves, N.: Lymphangiosarcoma in post- mastectomy lymphedema: A report of six cases in elephan- tiasis chirurgica. Cancer 1: 64-8 1, 1948. Storm, H.H., Jensen, O.M.: Risk of contralateral breast cancer in Denmark 1943-80. Br. J. Cancer 54: 483-492, 1986. Svensson, G.K., Kase, K.R., Chin, L.M., Harris, J.R.: Dose to the opposite breast as a result of primary radiation therapy for carcinoma ofthe breast (Abstr.). Znt. J. Radiat. Oncol. Biol. Phys. 7: 1209, 198 1. Vilcoq, J.R., Calle, R., Schlienger, P.: Irradiation techniques for conservative treatment of localized breast cancer. In Conservative Management of Breast Cancer, Har- ris, J.R., Hellman, S., Silen, W. (Eds.). Philadelphia, J.B. Lippincott Co. 1983, pp. 2 13-224.

Documents

Contralateral breast cancer and other second malignancies in patients treated by breast-conserving therapy with radiation