Im. J Radiation Oncology Bid. Phys.. Vol. 12, pp. 1575-1582 0360-3016/86 $3.00 + $4

Printed in the U.S.A. All rights reserved. Copyright 0 1986 Pergamon Journals Ltd.

??Original Contribution

EDEMA OF THE ARM AS A FUNCTION OF THE EXTENT OF AXILLARY SURGERY IN PATIENTS WITH STAGE I-II CARCINOMA OF THE

BREAST TREATED WITH PRIMARY RADIOTHERAPY

DAVID LARSON, M.D., PH.D.,’ MARK WEINSTEIN, M.D.,2 ITZHAK GOLDBERG, M.D.,’ BARBA.RA SILVER, B.A.,’ ABRAM RECHT, M.D.,’ BLAKE CADY, M.D.,3

WILLIAM SILEN, M.D.2 AND JAY R. HARRIS, M.D.’

‘Joint Center for Radiation Therapy, Department of Radiation Therapy; *Department of Surgery, Beth Israel Hospital; 3Department of Surgery, New England Deaconess Hospital; Harvard Medical School Boston, MA

Edema of the arm. can be a significant complication following treatment of breast cancer. To determine the risk of arm edema and factors associated with this risk in patients treated with primary radiotherapy, we reviewed the records of 475 women with early breast cancer treated between 1968 and 1980. During this period, the use of axillary surgery prior to radiation gradually increased, and all patients received full axillary irradiation until late in the series. Based on the surgeon’s report, the extent of axillary surgery was classified as either a sampling, a lower dissection, or a full dissection. Edema of the arm was scored on clinical grounds and ranged from mild hand swelling to an increased arm circumference of 8 cm. At 6 years, the actuarial risk of developing arm edema was 8% for the entire study population. This risk was 13% for 240 patients who had axillary surgery and 4% for 235 patients not undergoing axillary surgery (p = 0.006). For patients undergoing axillary surgery, the risk of arm edema was 37% with full dissection compared to 5% with sampling (p = 0.0003), and 8% with lower dissection (p = 0.03). The risk of arm edema at 6 years was 28% if more than ten nodes were removed, and 9% if one to ten nodes were removed (p = 0.03). However, the extent of axillary dissection was a stronger predictor of subsequent edema than was the number of nodes obtained. The role of axillary irradiation could not be evaluated since 91% of patients received axillary irradiation. The use of chemotherapy, the site or size of the primary tumor, clinical nodal status, patient age and weight, type of suture, the use of a drain, and subsequent local or distant failure did not appear to be significant risk factors. We conclude that the combination of full dissection and full axillary irradiation results in an unacceptably high risk of arm edema.

Breast cancer, Arlm edema, Axillary surgery, Axillary irradiation.

INTRODUCTION

The conventional treatment for early breast cancer has been radical or modified radical mastectomy. However, several retrospective and prospective clinical trials of pri- mary radiation therapy have demonstrated local control and survival rates comparable to those achieved by mas- tectomy. 1,3,7,10.15.21,28,33,35,39,40,45,47 Given these results and

the cosmetic and psychological advantages of breast pres- ervation, many patients and physicians now prefer treat- ment with primary radiation therapy. For patients treated with this approach, the cosmetic and functional outcomes are of particular importance.

Examples of adverse cosmetic outcome in the breast following primary radiation therapy are breast edema, re- traction, and telangiectasia. Factors known to be asso-

ciated with these complications include extensive breast surgery, extensive radiation therapy, and adjuvant che- motherapy.2,23 Several examples of poor functional and cosmetic results in tissues adjacent to the breast have been reported previously3,9,‘3,20332*36 and some, such as brachial plexus neuropathy, are known to be associated with high- dose axillary irradiation. 23*43 However, treatment factors associated with ipsilateral arm edema have not been fully defined. Such edema is bothersome to patients and is poorly managed by currently available techniques.

In this paper we present data from the Joint Center for Radiation Therapy (JCRT) on the development of edema of the arm in patients undergoing primary radiotherapy for early breast cancer. The goals of this study were: (a) to assess the actuarial risk of developing edema of the arm, (b) to relate the risk of edema to the extent of axillary

Presented in part at the annual meeting of the American So- ciety for Therapeutic Rad:iology and Oncology, Washington, D.C., 1984.

Reprint requests to: Jay R. Harris, M.D. Accepted for publication 24 March 1986.

1575

1576 I. J. Radiation Oncology 0 Biology 0 Physics September 1986, Volume 12. Number 9

surgery as judged by the surgeon’s report and by the num- stripping of the axillary vein is associated with an increased ber of nodes removed, and (c) to evaluate the role of other risk of arm edema. In the ensuing analysis and discussion factors, including axillary irradiation and chemotherapy, we also employ the term limited dissection, meaning less in the development of edema. than full dissection, that is, sampling or lower dissection.

METHODS AND MATERIALS

Between July 1968 and December 1980, 525 patients underwent primary radiotherapy for early breast cancer (UICC clinical Stages I and II) at the Joint Center for Radiation Therapy. Eleven of these patients had bilateral breast cancer, therefore, a total of 536 breast cancers were treated. Fifty cases were excluded from the present study because operative notes or pathology reports from axillary surgery were not available. In addition, eleven patients whose axillary surgery was performed using a subclavicu- lar rather than a lateral incision were excluded. The study population thus consisted of 475 breast cancers, of which 240 had axillary surgery for staging purposes and 235 did not. Median follow-up for patients treated with axillary surgery was 45 months (range, 27-94 months). Median follow-up for patients not having axillary surgery was 63 months (range, 29- 16 1 months). All patients underwent breast biopsy prior to any therapy. Clinical staging was performed by either a surgeon or a radiotherapist or both.

The extent of axillary surgery was selected at the dis- cretion of the surgeon. Each surgeon typically performed the same procedure in all patients, except in some in- stances in which a more limited procedure was performed when grossly-positive nodes were encountered. The work of 56 different surgeons is represented in this series.

The extent of axillary surgery was categorized according to the number of nodes removed as noted in the final pathology report, and also according to the surgeon’s de- scription as contained in the dictated operative report. The extent of axillary surgery was classified as a sampling, a lower dissection, or a full dissection. Definitions of these terms are as follows:

The technique of radiotherapy employed has been pre- viously described.42 All patients received radiotherapy at the JCRT. In all cases, radiotherapy was delivered with tangential fields to the ipsilateral breast and chest wall. In some cases the internal mammary nodes were treated, either with deeper tangential breast fields or, in patients with large tangential field separation, with an anterior “hockey-stick” field. In 350 patients, breast irradiation was followed by an additional boost to the primary lo- cation by external beam photons or electrons or by an interstitial implant. Four hundred forty-five patients had additional treatment to the ipsilateral supraclavicular re- gion, of whom 430 (91% of the total population) had treatment to both the ipsilateral supraclavicular and ax- illary regions with a separate anterior field matched to the tangential fields. The medial supraclavicular field border extended to the pedicles of the ipsilateral vertebral bodies. The lateral field margin typically extended to the coracoid process of the scapula for the supraclavicular field (dose calculated at a depth of 3 cm) or to the outer two-thirds of the humeral head for the full supraclavicular and ax- illary field (dose calculated at a depth of 5 cm). In some patients receiving axillary irradiation, an additional ax- illary boost was delivered through a single posterior ax- illary field or through an en-face axillary field. Until 1979, all patients received supraclavicular and axillary irradia- tion. After that, the decision to treat or not treat the nodal regions was usually based on the findings of the axillary dissection: for patients with negative axillary nodes and a lateral primary, treatment was given to the breast alone; for patients with negative axillary nodes and a medial primary, the axilla was not treated but the internal mam- mary nodes were treated, for patients with positive axillary nodes, all regional lymph node areas, including the axilla, were treated. In most patients, radiation was delivered with 180 or 200 rad fractions, 5 days per week. All treat- ments were given with X rays produced on 4 MV or 8 MV linear accelerators. The median axillary dose for the 430 cases involving axillary irradiation was 4600 rad (range, 4000-7500 rad). The number of cases, according to axillary surgery and axillary irradiation is listed in Ta- ble 1.

Sampling: A minimal axillary procedure involving the resection of a portion of the lower axillary contents without formal boundaries, varying from a simple node biopsy to excision of the axillary fat pad. A minor por- tion of the axillary contents is removed. The axillary vein is not visualized. Lower dissection: A resection of the lower axillary contents with formal boundaries, extending from the latissimus dorsi laterally, the tail of the breast inferiorly, the axillary vein superiorly and the lateral border of the pectoralis minor medially. A moderate portion of the axillary contents is removed. The ax&u-y vein,may be visualized but is not stripped. Full dissection: Resection of at least the entire level I and II axillary contents, from the latissimus dorsi muscle laterally, the tail of the breast inferiorly, the axillary vein superiorly, and up to or beyond the medial border of the pectoralis minor muscle medially. A ma- jor portion of the axillary contents is removed. The axillary vein is stripped.

Although these definitions are not standard, they do cor- respond to the types of axillary surgery performed at our institutions and allow us to examine the hypothesis that removal of a large portion of the axillary contents with

Eighty-eight patients received chemotherapy as part of their initial therapy. Of these, 65 underwent axillary sur- gery and had pathologically-positive nodes, whereas 23 received chemotherapy without undergoing axillary sur- gery. Cytoxan, methotrexate, and 5-FU (CMF) were used

Stage I-II carcinoma of the breast 0 D. LARSON et al. 1577

Table 1. Number of breast cancer patients according to axillary surgery and axillary irradiation

Axillary irradiation

Axillary surgery

Full dissection Lower dissection Sampling None

Total

Yes

33 90

2.::

4.30 (91%)

No Total

16 49 16 106 IO 85 3 235

45 415

in 57 patients, with the remaining 31 patients receiving one of 11 other single or (double-agent therapies. Only two patients received prednisone as part of their chemo- therapy.

Edema of the arm was assessed clinically in periodic follow-up by radiotherapists at the JCRT or by referring surgeons. The 28 patients who developed arm edema were examined in 208 follow-up visits (median number of visits 7.4, range 1- 18), not including routine examinations prior to the diagnosis of arm edema. Systematic quantification of edema, such as volumetric measurement of the arm, was not routinely employed. Edema ranged from mild symptomatic swelling of the hand to an increased circum- ference of the arm and/or forearm of eight cm. Edema persisted in all but four of the 28 patients; these four pa- tients, whose arm edema resolved at 9-17 months, were counted in the analysis as having edema. For all patients who developed edema, various standard recommenda- tions for management of edema were made, including elevation, pressure stockings, diuretics, pumps, avoidance of hand or arm trauma, etc. Patient compliance with these recommendations could not be evaluated. In no case was surgery undertaken to improve lymphatic flow. For pur- poses of statistical analy:sis, arm edema was scored at the time of onset, measured from the start of radiotherapy. The actuarial risk of developing edema was calculated using the Kaplan-Meier method.26 The statistical signif- icance of differences between actuarial curves was cal- culated using the two-tailed logrank test.3’ A multivariate analysis of factors possibly related to the development of arm edema was performed using the Cox model.‘*

R.ESULTS

Arm edema developed after treatment in 28 of 475 patients (5.9%). The actuarial risk of developing arm edema was 8% at six yea.rs for the entire population. The time course for the development of arm edema is relatively long in many cases, however no patient had the onset of edema more than 6 years after the start of radiotherapy. In this group of patients, arm edema was either mild or moderate. No patient developed severe edema.

The risk of arm edema was evaluated in relation to the use and extent of axillary surgery (Table 2). The actuarial risk of edema at six years was 13% for patients undergoing

axillary surgery, whereas the risk was 4% if no axillary surgery was performed (p = 0.006) (Fig. 1). When patients

To evaluate the relative importance of the extent of

were grouped according to the extent of surgery, this risk was 37% for full dissection, 8% for lower dissection (p

surgery and the number of nodes obtained as predictors

= 0.03), and 5% for sampling (p = 0.0003). Since the risk

of subsequent edema, we compared the 6 year risk of

of edema was not significantly different for patients un- dergoing lower dissection and sampling, even when strat-

edema for the four possible subgroups involving the extent

ified by other variables, these groups were combined and termed limited dissection. When patients were grouped according to the number of nodes obtained, the risk at six years was 28% if more than IO nodes were obtained, and 9% if I- IO nodes were obtained (p = 0.03).

Table 2. Actuarial risk of arm edema at 6 years

Patient subgroup No. %

cases edema p-value

All patients 475

Axillary surgery 240 No axillary surgery 235

Full dissection 49 Lower dissection 106 Sampling 85 No axillary surgery 235

> 10 nodes I-IO nodes

74 166

Full dissection, > IO nodes 26 Full dissection, I- IO nodes 23 Limited dissection, > IO nodes 48 Limited dissection, I- 10 nodes 143

Full dissection, nodes negative 36 Full dissection, nodes positive 13 Limited dissection, nodes negative 126 Limited dissection, nodes positive 65

Full dissection, axillary irradiation Full dissection, no axillary

irradiation Limited dissection, axillary

irradiation Limited dissection, no axillary

irradiation

33 36

16 -

165

26 11

Full dissection, axillary irradiation >4600 rad

Full dissection, axillary irradiation ~4600 rad

Limited dissection, axillary irradiation >4600 rad

Limited dissection, axillary irradiation ~4600 rad

IO

23

65 7

Full dissection, chemotherapy Full dissection, no chemotherapy Limited dissection, chemotherapy Limited dissection, no

chemotherapy

126

:: 54

137 6

8

1:10.006

37 8 0.03 5 0.0003

“b 0.03

40 37 10 5 \ 0.0004

29 38

6 \ 0.008 11

6 0.0002

8

31 32

8

1578 1. J. Radiation Oncology 0 Biology 0 Physics September 1986, Volume 12, Number 9

50 -

P 40-

8

Q 30-

$ ae 201

p=O 006

10 - No ox~llory surgery (235)

0

0. 2 4 6 0 IO 12

TIME cja?aK~

Fig. 1. Actuarial risk of developing edema of the arm within 6 years after treatment with primary radiotherapy.

of surgery (full vs. limited dissection) and the number of nodes obtained (l-10 vs. more than 10). This simulta- neous comparison demonstrated that the extent of surgery rather than the number of nodes obtained is more strongly associated with subsequent edema (Table 2). For patients undergoing full dissection, the risk of edema was similar whether more than 10 nodes or 1 - 10 nodes were obtained (40% vs. 37%, p = NS). For patients undergoing limited dissection, the risk of edema was also similar whether more than 10 or l- 10 nodes were obtained (10% vs. 5%, p = NS). The risk of arm edema was significantly higher for patients who underwent full dissection and had l-10 nodes removed compared to those who underwent limited dissection and had l-10 nodes removed (37% vs. 5%, p = 0.0004).

There was no difference in the rate of development of arm edema between patients with pathologically negative nodes and patients with pathologically positive nodes when stratified by the extent of axillary surgery (Table 2). Patients who underwent limited dissection had an actu- arial risk of edema of 6% at six years if their nodes were negative and 11% if their nodes were positive (p = NS). Among patients who underwent full dissection, the risk of edema was 29% if the nodes were negative and 38% if the nodes were positive (p = NS).

A comparison of subgroups involving the extent of sur- gery and axillary irradiation was also performed (Table 2). This comparison was limited because only 16 patients underwent full dissection without axillary irradiation and the maximum follow-up in this group was only 4 years. The 4 year actuarial risk of edema was 12% for this group. For the 33 patients who underwent full dissection and also received axillary irradiation, the 6 year risk of edema was 36%, significantly higher than the 6% risk for the 165 patients who underwent limited dissection and received axillary irradiation (p = 0.0002). The risk of edema was 11% for the 26 patients undergoing limited dissection and no axillary irradiation.

Subgroups were compared for extent of surgery and dose of axillary irradiation (Table 2). Among patients un- dergoing full dissection, there appeared to be an increased risk of arm edema with increased dose, although the number of patients in each group followed for 6 years was small, and the results were not statistically significant. For the 23 patients treated with a full dissection and ax- illary irradiation of 4400 to 4600 rad, the 6 year risk of arm edema was 19% compared to 72% for the 10 patients treated with full dissection and 4800 to 7050 rad (p = 0.21). Among patients treated with limited dissection and axillary irradiation there was no apparent dose-re- sponse relationship for edema of the arm, although no patient received more than 5600 rad, and the risk of edema was low (7% versus 8%).

Results were also compared for subgroups involving the extent of surgery and the use of chemotherapy (Table 2). This comparison demonstrated a significantly higher risk of edema in patients undergoing full dissection whether or not they received chemotherapy (3 1% and 32%) compared to those undergoing a limited dissection whether or not they received chemotherapy (8% and 6%).

Multivariate analysis of all the aforementioned variables plus age, clinical nodal status, primary size and location, the use of sutures, the use of drains, and the development of subsequent local and distant failure was performed. The only significant variables were the extent of surgery and the number of nodes removed. These results parallel those of the univariate analysis.

The yield of axillary nodes for each of the three surgical procedures was evaluated. For full dissection, the median number of nodes obtained was 11 (range, 3-24). For lower dissection, the median number of nodes obtained was 9 (range, l-35). For sampling, the median number of nodes obtained was 5 (range, l- 14). The percentage of patients with pathologically-positive nodes for each category of surgical procedure and clinical axillary nodal group was determined. For patients with a clinically-negative axilla, the percentage with positive nodes was 2 1% for sampling, 22% for lower dissection, and 16% for a full dissection (all p = NS). For patients with a clinically-positive axilla, the percentage of patients with positive nodes was 70% for sampling, 94% for lower dissection, and 100% for full dis- section (all p = NS). Note that 50% of patients with clin- ically-positive axillary nodes underwent a sampling (as opposed to a larger surgical procedure) compared to only 32% of patients with clinically-negative nodes (p = 0.04). This observation suggests that surgeons commonly de- creased the extent of axillary surgery when grossly-positive axillary nodes were encountered.

DISCUSSION

Management of the axilla is an important component of the overall staging and treatment plan for patients with early breast cancer. For patients who undergo primary radiation therapy, axillary management typically includes

Stage I-II carcinoma of the breast 0 D. LARSON et al. 1579

axillary surgery, ranging from a sampling to a full dissec- tion, and may involve axillary irradiation, commonly in the range of 4500-5000 rad delivered in 5 weeks. The chief purpose of axillary surgery is to establish the presence or absence of nodal involvement, to determine whether adjuvant systemic therapy is indicated. It is generally agreed that the procedure should be extensive enough to provide accurate assessment of nodal status. However, the extent of axillary surgery necessary to achieve this goal remains controversial. Some surgeons advocate full dissection. They argue that such a procedure provides the most accurate assessment of prognosis and best defines patients most likely to benefit from adjuvant systemic therapy, since the total number of positive axillary nodes is identified.34 Furthermore, it is argued, such a procedure obviates the need for axillary irradiation since axillary recurrences are uncommon after radical mastec- tomy. 16~24,46,49 Also, some data suggest that sampling pro- cedures are likely to miss positive nodes in a substantial number of cases. Davies et al. 14, in a study involving post- operative dissections of full axillary dissection specimens, reported that a sampling .would have missed positive nodes in 6 of 43 patients ( 14%), although we can only speculate on how closely such postoperative samplings correspond to actual sampling procedures. Kissin et d2’, in a study involving full dissection immediately following sampling, reported that sampling failed to identify positive nodes in 6 of 35 patients ( 17%), although the samplings might have been less thorough than usual since they were performed with the knowledge that full dissection would follow.

Advocates of less extensive surgery, on the other hand, argue that lesser procedures are as likely as full dissection to detect positive nodes and are less morbid. Steele et d4’, in a randomized study, reported 42% of patients had positive nodes following rampling (mean number of nodes = 4.8) compared to 40% following full dissection (mean number of nodes = 20.6), although the clinical nodal sta- tus of patients in each group was not reported. The present study, in which patients were categorized according to clinical nodal status, also failed to demonstrate statistically significant differences in the percentage of patients having positive nodes for different extents of axillary surgery ( 16- 22% pathologically positive if clinically negative; 70- 100% pathologically positive if Iclinically positive). Rosen et d3*, in a study involving 1228 patients undergoing mastec- tomy, reported that discontinuous or “skip” metastases (not showing progressive spread from level I to II to III) occurred in only 3% of 508 patients with positive nodes. Since half of those with “skip” metastases had tumor lim- ited to level II, the potential risk from such patterns of spread was considered negligible for women undergoing axillary surgery that includes level II.

These arguments favoring differing degrees of axillary surgery are difficult to msolve for several reasons. First, the number of nodes obtained may vary depending on the technique used by the pathologist. For example, com- pared to conventional manual dissection, “clearing” a

specimen by removing fat may increase by 30% the num- ber of nodes recovered, although the chance that these additional nodes will be positive is 10w.~’ Furthermore, serial subsectioning of lymph nodes may increase by up to 30% the probability of detecting nodal involvement.48 At present it is unknown whether less extensive axillary surgery, together with “clearing” and/or serial subsec- tioning would give a higher likelihood of detecting nodal disease, compared to conventional handling of a more extensive axillary specimen. Second, difficulties exist in defining the precise location from which nodes have been removed. For example, a common terminology used to describe the extent of axillary surgery is that of levels. Tissue lying lateral to, beneath, or medial to the pectoralis minor muscle is referred to as level I, II, or III, respectively. However, these divisions do not represent precise ana- tomic boundaries. It is therefore difficult to be absolutely certain of the precise level from which a given node was taken. A more anatomic classification may be used to identify subgroups of axillary nodes, with the axilla divided into external mammary, scapular, central, axillary vein and subclavicular areas.” Under this system also, how- ever, the precise location of a given node may be uncer- tain. Unfortunately, terms such as “sampling”, which are not defined anatomically, may well have different meaning for different surgeons. Finally, less extensive surgery might be performed by some surgeons if there is a high degree of suspicion of axillary involvement. In this study, for example, 50% of patients with clinically positive nodes underwent a sampling, compared to 32% of patients with clinically negative nodes. For these reasons, it is difficult to compare the number of nodes obtained in different studies addressing the optimum extent of axillary surgery.

Another point to consider regarding axillary treatment is tumor control at that site. There is little debate regarding the effectiveness of either axillary irradiation or full axillary dissection in providing control in the axilla. It is known that up to 40% of patients with a clinically negative axilla harbor axillary disease.16 In the absence of axillary surgery or axillary irradiation, approximately 20% of patients will develop clinical evidence of axillary metastases.16 With axillary irradiation of 4500-5000 rad in 5 weeks, or with a full axillary dissection, the axillary recurrence rate is O- 2%.22,28,33,47 The likelihood of axillary tumor control after a lesser surgical procedure has not been well-documented.

A still unresolved issue is the influence of axillary treat- ment on the likelihood of patient survival. Results from prospective randomized trials have failed to demonstrate a significant survival advantage for patients who received effective axillary treatment compared to patients with no axillary treatment.8s’7

A final point to consider in the management of the axilla is the risk of complications following axillary treat- ment. According to the results of this study, the overall risk of arm edema for patients with breast cancer treated with primary radiotherapy is relatively low, but edema is permanent for most patients. Only four of 28 patients in

1580 I. J. Radiation Oncology 0 Biology 0 Physics September 1986, Volume 12, Number 9

this series had resolution of their edema (at 9- 17 months). Although the risk was assessed by the use of actuarial rather than crude statistics, and by the inclusion of the four patients who developed transient edema, the actuarial risk of arm edema was only 8% at six years. This is similar to the (crude) incidence reported in other studies, in the range of 1.5 to 11%. 7,9.22,29,33,35,37,45 For comparison, the

risk of arm edema in patients treated with mastectomy was reported to vary from 6.7 to 62.5%.5 More recent studies report an incidence of 2.76 and 8.9%.” Such a wide range may reflect varying criteria for the diagnosis of arm edema as well as evolving surgical techniques.

bers of patients with adequate follow-up in this latter group, the difference in risk is not significant. We are therefore unable to determine the impact of radiotherapy on the risk of edema in patients undergoing full dissection. For patients undergoing limited dissection, however, no increased risk was associated with radiotherapy.

In this study, 430 of 475 patients (9 1%) received axillary irradiation and the risk of subsequent edema was clearly related to the use of axillary surgery. The risk of mild or moderate arm edema was 13% at six years following ax- illary surgery compared to only 4% when no axillary sur- gery was performed. Furthermore, the degree of risk was related to the extent of axillary surgery, as judged either by the surgeon’s report or by the number of nodes ob- tained. This risk was 37% for full dissection, 8% for lower dissection, and 5% for sampling; it was 28% if more than 10 nodes were obtained but only 9% if fewer than I- 10 nodes were obtained. A relatively low risk was associated with limited dissection, independent of the number of nodes obtained. Conversely, a high risk was associated with full dissection, also independent of the number of nodes obtained. These results indicate that the risk of arm edema in patients receiving axillary irradiation is related to the extent of axillary surgery, and that this risk is most clearly defined by the surgeon’s description of the pro- cedure performed.

In this study we were unable to demonstrate an in- creased risk of arm edema associated with chemotherapy. This is in contrast to a recent studyI in which 27 patients with early breast cancer who received radiation and com- pleted 12 cycles of adjuvant chemotherapy consisting of CMF or CMF plus prednisone were compared to 19 pa- tients who received radiation alone. It was reported that six of the patients receiving adjuvant chemotherapy sub- sequently developed arm edema whereas none of the pa- tients in the control group developed arm edema. The reason for the difference between those results and ours is not apparent, although it may be related to the number of cycles of chemotherapy given, a variable which we did not evaluate.

Axillary irradiation may contribute to the risk of arm edema in patients undergoing axillary surgery. It has been well-documented that axillary irradiation following radical mastectomy significantly increases the risk of arm edema.4~‘8~25~44 In this study, the combination of full dis- section and axillary irradiation resulted in a 36% risk of arm edema at 6 years. For patients undergoing full axillary dissection followed by 4600 rad or more to the axilla the risk was 43%; for such patients receiving less than 4600 rad the risk was 14%. However, because of limited num-

Our findings have implications for the radiotherapist with regard to optimal management of the axilla in pa- tients selected for primary radiotherapy. As discussed above, the optimal extent of axillary surgery is contro- versial, and reasonable surgeons are likely to continue to perform a range of procedures. It is important for the radiotherapist to determine the extent of axillary surgery performed. If a limited dissection has been performed, axillary irradiation may be required to assure local tumor control, particularly when positive nodes are obtained. It appears, from this study, that such irradiation does not substantially contribute to the risk of arm edema. How- ever, if a large portion of the axillary contents has been removed with stripping of the axillary vein, there is a high risk of arm edema if axillary irradiation is added. Since axillary irradiation is rarely required to ensure local tumor control in patients undergoing full dissection, and may contribute to arm edema, it should be avoided. Only in special circumstances where nearly all recovered nodes are positive or extensive extranodal disease is found should axillary irradiation be considered for local control follow- ing full dissection.

REFERENCES

1. Amalric, R., Santamaria, F., Robert, F., SeigIe, J., Altschuler, C., Kurtz, J.M., Spitalier, J.M., Brandone, H., Ayme, Y., Pollet, J.F., Burmeister, R., Abed, R.: Radiation therapy with or without primary limited surgery for operable breast cancer. A 20 year experience at the Marseilles Cancer In- stitute. Cancer 49: 30-34, 1982.

2. Beadle, G.F., Silver, B., Botnick, L., Hellman, S., Harris, J.R.: Cosmetic results following primary radiation therapy for early breast cancer. Cancer 54: 29 1 l-29 18, 1984.

3. Bedwinek, J.M., Perez, C.A., Kramer, S., Brady, L.W., Goodman, R., Grundy, G.: Irradiation as the primary man- agement of stage I and II adenocarcinoma of the breast. Cancer Clin. Trials 3: 1 l- 18, 1980.

4. Brismar, B., Ljungdahl, I.: Postoperative lymphoedema after treatment ofbreast cancer. Acta Chir. &and. 149: 687-689, 1983.

5. Britton, R.C., Nelson, P.A.: Causes and treatment of post- mastectomy lymphedema of the arm: report of I 14 cases. JAMA 180: 95, 1962.

6. Budd, DC., Cochran, R.C., Sturtz, D.L., Fouty, W.J.: Sur- gical morbidity after mastectomy operations. Am. J. Surg. 135: 218-220, 1978.

7. Calle, R., Pilleron, J.P., Schlienger, P., Vilcoq, J.R.: Con- servative management of operable breast cancer. Ten years experience at the Fondation Curie. Cancer 42: 2045-2053, 1978.

Stage I-II carcinoma of the breast 0 D. LARSON et al. 1581

8.

9.

10.

Il.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

Cancer Research Carnpaign Working Party. Cancer Re- search Campaign trial for early breast cancer: a detailed update at the tenth year. Lancet 2: 55-64, 1980. Chu, A.M., Cope, O., Russo, R., Wang, C.C., Shulz, M.D., Wang, C., Rodkey, G.: Treatment of early stage breast cancer by limited surgery and radical irradiation. Int. J. Radiat. Oncol. Biol. Phys. 6: 25-30, 1980. Cope, O., Wang, C.A., Chu, A., Wang, C., Shulz, M., Cas- tleman, B., Long, J., Sohier, W.D.: Limited surgical excision as the basis of a comprehensive therapy for cancer of the breast. Am. J. Surg. 131: 400-406, 1976. Corneillie, P., Gruwez, J.A., Lerut, T., Van Elst, F.: Early and later postoperative sequelae after surgery for carcinoma of the breast. Acta Chk Belg. 84: 227-23 1, 1984. Cox, D.R.: Regression models and life tables. J. R. Stat. Sot. B. 34: 187-220, 1.972. Danoff, B.F., Goodman, R.L., Glick, J.H., Haller, D.G., Pajak, T.F.: The effect of adjuvant chemotherapy on cos- mesis and complications in patients with breast cancer treated by definitive irradiation. Int. J. Radiat. Oncol. Biol. Phys. 9: 1625-1630, 1983. Davies, G.C., Millis, lR.R., Hayward, J.L.: Assessment of axillary lymph node st,atus. Ann. Surg. 192: 148- 15 1, 1980. Fisher, B., Bauer, M., Margolese, R., Poisson, R., Pilch, Y., Redmond, C., Fisher, E., Wolmark, N., Deutsch, M., Mo- ntague, E.: Five-year results of a randomized clinical trial comparing total mast’ectomy and segmental mastectomy with or without radiation in the treatment of breast cancer. N. Engl. J. Med. 312: 665-673, 1985. Fisher, B., Montague, I:., Redmond, C., Barton, B., Vorland, D., Fisher, E.R., Deutsch, M., Schwarz G., Margolese, R., Donegan, W., Volk, I-I., Konvolinka. C.: Comparison of radical mastectomy with alternative treatments for primary breast cancer. A first report of results from a prospective randomized clinical trial. Cancer 39: 2827-2839, 1977. Fisher, B., Redmond, C., Fisher, E.R., Bauer, M., Wolmark, N., Wickerham, D.L., Deutsch, M., Montague, E., Mar- golese, R., Foster, R.: Ten-year results ofa randomized clin- ical trial comparing radical mastectomy and total mastec- tomy with or without radiation. N. Engl. J. Med. 312: 674- 681, 1985. Ganel, A., Engel, J., Sela, M., Brooks, M.: Nerve entrap- ments associated with postmastectomy lymphedema. Can- cer 44: 2254-2259, 19’79. Haagensen, C.D.: Diseases of the Breast, 2nd edition. Phil- adelphia, W.B. Saunders, 197 1. Harris, J., Hellman, S.: Primary radiation therapy for early breast cancer: The experience at the Joint Center for Ra- diation Therapy. In Alternatives to Mastectomy, Harris, J.R., Hellman, S., Silen, W. (Eds.). Philadelphia, J.B. Lippincott Co., 1982. Harris, J.R., Beadle, G.F., Hellman, S.: Clinical studies on the use of radiation therapy as primary treatment of early breast cancer. Cancer 53: 705-7 11, 1984. Harris, J.R., Botnick, L., Bloomer, W.D., Chaffey, J.T., Hellman, S.: Primary radiation therapy for early breast can- cer: The experience at the Joint Center for Radiation Ther- apy. Int. J. Radiat. On(col. Biol. Phys. 7: 1549-1552, 198 1. Harris, J.R., Levene, M.B., Svensson, G., Hellman, S.: Analysis of cosmetic results following primary radiation therapy for stages I and II carcinoma of the breast. Znt. J. Radiat. Oncol. Biol. Phys. 5: 257-261, 1979. Hayward, J.L.: The Guy’s trial of treatments of “early” breast cancer. World J. Surg. 1: 314-316, 1977. Jolly, P.C., Viar, W.: F!eduction of morbidity after radical mastectomy. Am. Surg. 47: 377-380, 1981. Kaplan, E.S., Meier, P.: Non-parametric estimation from

27

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

incomplete observation. Am. Stat. Assoc. J. 53: 457-481, 1958. Kissin, M.W., Thompson, E.M., Price, A.B., Slavin, G., Kark, A.E.: The inadequacy of axillary sampling in breast cancer. Lancet 1: 1210-1212, 1982. Montague, E.D., Gutierrez, A.E., Barker, J.L., Tapley, N.D., Fletcher, G.H.: Conservative surgery and irradiation for the treatment of favorable breast cancer. Cancer 43: 105% 106 1, 1979. Montague, E.D., Schell, S.R., Romsdahl, M.D., Ames, F.C.: Conservation surgery and irradiation in the treatment of breast cancer. In Conservation Surgery and Radiation Therapy in the Treatment of Operable Breast Cancer, Vaeth, J.M. (Ed.). Basel, Munchen, Paris, London, New York, Sid- ney, Karger. 1983, pp. 69-75. Morrow, M., Evans, J., Rosen, P.P., Kinne, D.W.: Does clearing of axillary lymph nodes contribute to accurate stag- ing of breast carcinoma? Cancer 53: 1329- 1332, 1984. Peto, R., Pike, M.C., Armitage, P., Breslow, N.E., Cox, D.R., Howard, S.V., Mantel, N., McPherson, K., Peto, J., Smith, P.G.: Design and analysis of randomized clinical trials re- quiring prolonged observation of each patient. Br. J. Cancer 35: l-39, 1977. Pierquin, B.: Conservative treatment for carcinoma of the breast-Experience of Creteil. 10 years’ results. In Alter- natives to Mastectomy, Harris, J.R., Hellman, S., Silen, W. (Eds.). Philadelphia, J.B. Lippincott Co., 1982. Pierquin B., Owen, R., Maylin, C., Otmezguine, Y., Raynal, M., Mueller, W., Hannoun, S.: Radical radiation therapy of breast cancer. Znt. J. Radiat. Oncol. Biol. Phys. 6: 17- 24, 1980. Pigott, J., Nichols, R., Maddox, W.A., Balch, C.M.: Metas- tases to the upper levels of the axillary nodes in carcinoma of the breast and its implications for nodal sampling pro- cedures. Surg. Gynecol. Obstet. 158: 255-259, 1984. Prosnitz, L.R., Goldenberg, I.S., Packard, R.A., Levene, M.B., Harris, J.R., Hellman, S., Wallner, P.E., Brady, L.W., Mansfield, C.M., Kramer, S.: Radiation therapy as initial treatment for early stage cancer of the breast without mas- tectomy. Cancer 39: 9 17-923, 1977. Prosnitz, L.R., Weshler, Z., Goldenberg, I.S.: Primary ra- diotherapy for breast carcinoma instead of mastectomy: The Yale experience. In Alternatives to Mastectomy, Harris, J.R., Hellman, S., Silen, W. (Eds.). Philadelphia, J.B. Lippincott Co., 1982. Ray, G.R., Fish, V.J., Lee, R.H., Marzoni, F.A., Jr., Trol- lope, M.L., Hews, M., Gribble, M.: Biopsy and definitive radiation therapy in stages I and II carcinoma of the female breast. Int. J. Radiat. Oncol. Biol. Phys. 9: 23-28, 1983. Rosen, P.P., Lesser, M.L., Kinne, D.W., Beattie, E.J.: Dis- continuous or “skip” metastases in breast carcinoma. Anal- ysis of 1228 axillary dissections. Ann. Surg. 197: 276-283, 1983. Sarrazin, D., Le, M., Fontaine, F., Arriagada, R.: Conser- vative treatment versus mastectomy in T 1 or small T2 breast cancer-A randomized clinical trial. In Conservative Man- agement of Breast Cancer, Harris, J.R., Hellman, S., Silen, W. (Eds.). Philadelphia, J.B. Lippincott Co. 1983, pp. lOl- 111. Spitalier, J., Brandone, H., Ayme, Y., Amahic, R., Santa- maria, F., Seigle, J.: Cesiumtherapy of breast cancer. A five year report on 400 consecutive patients. Int. J. Radiat. On- col. Biol. Phys. 2: 231-235, 1977. Steele, R.J.C., Forrest, A.P.M., Gibson, T., Stewart, H.J., Chetty, U.: The efficacy of lower axillary sampling in ob- taining lymph node status in breast cancer: A controlled randomized trial. Br. J. Surg. 72: 368-369, 1985.

1582 1. J. Radiation Oncology 0 Biology 0 Physics September 1986. Volume 12. Number 9

42.

43.

44.

45.

Svensson, G.K., Bjamgard, B.E., Larsen, R.D.. Levene, M.B.: A modified three-field technique for breast treatment. Int. J. Radiat. Oncol. Biol. Phys. 6: 689-694, 1980. Svensson, H., Westling, P., Larsson, L.G.: Radiation-in- duced lesions of the brachial plexus correlated to the dose- time-fraction schedule. Acta Radiof. 14: 228-238, 1975.

Swedborg, I., Wallgren, A.: The effect of pre- and postmas- tectomy radiotherapy on the degree of edema, shoulder- joint mobility, and gripping force. Cancer 47: 877-881, 1981. Veronesi, U., Saccozzi, R., Del Vecchio, M., Banfi, A., Clemente, C., Delena, M., Gallus, G., Greco, M., Luini, A., Marubini, E., Muscolino, G., Rilke, F.: Comparing radical mastectomy with quadrantectomy, axillary dissection and

46.

47.

48.

49.

radiotherapy in patients with small cancers of the breast. N. Engl. J. Med. 305: 6- 11, 198 1. Veronesi. U.. Valagussa, P.: Inefficacy of internal mammary node dissection in breast cancer surgery. Cancer 47: 170- 175, 1981. Vilcoq, J.R., Calle, R., Stacey, P., Ghossein, N.A.: The out- come of treatment by tumorectomy and radiotherapy of patients with operable breast cancer. Int. J. Radiat. Oncol. Biol. Phys. 7: 1327- 1332, 198 1. Wilkinson, E.J., Hause, L.: Probability in lymph node sec- tioning. Cancer 33: 1269- 1274, 1974. Zimmerman, K.W., Montague, E.D., Fletcher, G.H.: Fre- quency, anatomical distribution and management of local recurrences after definitive therapy for breast cancer. Cancer 19: 67-74, 1966.