Selective avoidance of lymphatic radiotherapy in the conservative management of early breast cancer

Radiotherapy and Oncology, 25 (1992) 83-88

© 1992 Elsevier Science Publishers B.V. All rights reserved. 0167-8140/92/$05.00

83

RADION 01040

Selective avoidance of lymphatic radiotherapy in the conservative management of early breast cancer P. J. H o s k i n a'*, B. Rajan a, S. Ebbs c, D. Tait a, S. Milan b a n d J. R. Y a r n o l d a Departments of ~Radiotherapy and Oncology, and b Computing, The Royal Marsden Hospital, Sutton, Surrey, UK and ° Department of Surgery, Mayday University Hospital, Croydon, Surrey, UK

(Received 18 April 1991, revision received 7 May 1992, accepted 27 May 1992)

Key words: Breast cancer; Radiotherapy; Lymph node irradiation

Summary In view of the morbidity and potential mortality associated with routine post-operative lymph node radiotherapy in women with early stage breast cancer, an attempt has been made to select patients in whom radiotherapy can be withheld. Three hundred and forty-seven consecutive patients treated wide local excision plus or minus axillary surgery have been evaluated. Only 20?0 were subsequently given radiotherapy to regional nodes. Relapse in the axilla, the supraclavicular fossa or at both these sites have occurred in 16 patients so far, 12 of whom were successfully treated. Systemic relapse was seen in eight of these patients occurring with one exception before or within 3 months of node relapse. Only four have persisting symptoms as a result of nodal relapse. So far, a policy involving selective lymphatic radiotherapy in women treated for early breast cancer appears justified.

Introduction

Breast conservation is established as a suitable alternative to mastectomy for the majority of patients with early breast cancer [ 7,17]. Locoregional radiotherapy reduces the incidence of relapse in the breast and node sites after conservative surgery. There is little evidence that local treatment benefits patients overall in terms of survival [6,17] although a recent prospective randomised study has suggested that in high risk (T3, T 4 or node positive) patients following mastectomy and node sampling, irradiation confers a disease-free and possible overall survival advantage independent of the effect of adjuvant systemic therapy [ 11]. However, recent survival analysis have indicated an excess mortality in long-term survivors associated with postmastectomy irradiation of the chest wall and lymphatic pathways using old-fashioned techniques [3,9,10].

Even if the risk of treatment-related death is much reduced with modern radiotherapy techniques, lymphatic irradiation is still associated with significant morbidity in a proportion of women. Side-effects include arm oedema, shoulder stiffness, brachial plexus neuropathy and match-line fibrosis [19]. Complications are more likely if radiotherapy encompasses a surgically dissected axilla [4]. In view of the significant morbidity and potential mortality associated with routine post-operative lymphatic radiotherapy, an attempt has been made to identify women in whom lymphatic treatment is better withheld. A treatment policy was developed based on an analysis of risk factors predisposing to lymph node recurrence and iatrogenic morbidity [19]. Women in whom the risks of complications following prophylactic radiotherapy were thought to exceed the risk and morbidity of lymphatic recurrence were not recom-

Address for correspondence: Dr. J. R. Yarnold, Department of Radiotherapy and Oncology, The Royal Marsden Hospital, Downs Road, Sutton,

Surrey SM2 5PT, UK. * Current address: Department of Radiotherapy, The Royal London Hospital, Whitechapel, London E1 1BB, UK.

84 mended lymphatic radiotherapy. An audit of treatment outcome has examined the levels of relapse and physical morbidity associated with this treatment policy.

TABLE II Details of lymphatic treatment. No Radiotherapy to Radiotherapy Total radiotherapy supraclavicular to axilla and fossa supraclavicular fossa

Patients and methods Three hundred and forty-seven consecutive patients treated with breast conserving techniques between January 1984 and June 1989 were evaluated. The women had a median age of 55 years (range24-78) and 107 women were < 50 years of age at presentation. The survivors were followed for a median of 35 months (range 3.5-98). The clinical stages at presentation are presented in Table I. The planned treatment policy included a macroscopic local excision of the primary tumour and whole breast radiotherapy in all women. In patients under the age of 50, most surgeons performed a level 2 axillary dissection rather than a sampling procedure. Women aged 50 or more with clinically positive axillae were recommended to have a level 2 axillary dissection. Radiotherapy policy involved treatment to the breast in all patients but never to the internal mammary chain. An arbitrary decision was made to offer supraclavicular fossa radiotherapy to women with more than four histologically positive axillary lymph nodes in the surgical specimen. No axillary radiotherapy was considered unless surgery was considered incomplete and positive axillary lymph nodes were recovered. Women over the age of 50 with clinically normal axillae were offered a watch policy (i.e. no surgery or radiotherapy) in all regional lymph node areas. Node positive women under 50 were offered 6 months adjuvant chemotherapy, and all women over 50 were offered adjuvant Tamoxifen for a minimum of 2 years. The actual surgical treatment of the lymphatic pathways is summarised in Table II. Two hundred and forty-eight (71 ~/o)had axillary surgery, usually described as a level 2 or complete axillary dissection. Ninety-nine (29~o) women underwent no axillary surgery, representing 90 women over 50 years of age with N 0 _ l a lymph nodes at presentation and nine node negative women under 50 years of age. In these 99 women, sur-

TABLE I T s t ~ e andclinical n o d e s t ~ u s ~ present.ion. No

N1

Nx

Total

T1 T2

81 198

17 48

0 3

98 249

Total

279

65

3

347

No axillary surgery 94 Axillary surgery 184

0

5

99

39

24

248

Total

39

29

347

278

gery was confined to wide local excision of the breast lump. Post-operative radiotherapy to the breast was given to all patients. The actual radiotherapy treatment to the lymphatic pathways is also shown in Table II. Two hundred and seventy-eight (80 ~o) women were given no lymphatic radiotherapy. Thirty-nine patients had radiotherapy confined to the supraclavicular fossa, 35 of whom had positive axillary lymph nodes. Twenty-nine patients had radiotherapy to both the axilla and supraclavicular fossa, mainly in women with positive axillary lymph nodes after a sampling procedure, but including four women who were pathologically node negative and five women > 50 years who were clinically node negative. Apart from these nine women, the only other protocol violations involved five patients who did not receive supraclavicular fossa radiotherapy despite > 4 positive axillary nodes and four patients who received radiotherapy to the supraclavicular fossa despite negative axillary surgery. Radiotherapy was delivered using 6 MV X-rays with the patient lying supine or on an inclined plane. Both arms were abducted to 90 ° and supported on an adjustable pole attached to the treatment couch [20]. The same position was maintained for all treatment fields. An isocentric technique was used to cover the breast with tangential fields. Couch twist was used to facilitate matching of the tangential beams along their superior border. Treatment to the supraclavicular fossa involved a direct 6 MeV photon field matched to the upper border of the tangential beams. In patients having axillary radiotherapy as well, the anterior field was enlarged and a smaller axillary field was used to raise the dose in the axillary mid-plane. Doses were prescribed as an applied dose to the supraclavicular fossa and a mid-plane dose to the axilla. The standard dose prescription to the lymphatic pathways was 50 Gy in 25 fractions given over 5 weeks. Since 1986, patients were randomised with informed consent into a fractionation study com-

85 t00

t00

90

90

80

80

70

70

--i

60

60

c~

50

50

40

40 o:

30

30

20

20

10

10

0

''''"'""t"'"'

0

.....

1

I' . . . . . . . . . .

I ......

2

3

'''"1'"

........

4

I .........

5

0

"l'"''"""l

6

''"'"'"l

0

7

cancer,

~-

70

~

6O

~

5O

~

4o

N

3o

~

20

"['''''''''"1

3

...........

4

{p>0.1)

I ...........

I"'''''""l

5

6

7

Results

The overall survival and lymph node relapse-free interval are shown in Figs. 1 and 2. The actuarial risk of lymph node relapse at 5 years is 10~o and of death is 11 ~o. Sixteen patients suffered lymphatic relapse in either axilla, the supraclavicular fossa or at both sites, see Table III. Freedom from axillary lymph node relapse according to surgical treatment of the axilla is shown in Fig. 3. Of the 14 axillary recurrences, eight developed in the 224 women following axillary surgery alone. One appeared among the four women treated by lymphatic radiotherapy and five developed in the 94 women who had neither surgery nor radiotherapy to the axilla, see Table IV. Eight patients relapsed at distant sites as well as nodal sites; seven patients also relapsed in the breast, including three who relapsed in breast, node and distant sites concomitantly. Distant relapse was seen prior to or within 3 months of node relapse in all except one patient who relapsed 15 months after node relapse.

100

80

2

signiricanl

on all patients, including those who also relapse in the breast and at distant sites.

paring three alternative schedules; 39 Gy in 13 fractions over 5 weeks versus 42.9 Gy in 13 fractions over 5 weeks versus 50 Gy in 25 fractions over 5 weeks. Twenty women under the age of 50 years with positive axillary lymph nodes received 6 months of adjuvant cytotoxic chemotherapy consisting of either cyclophosphamide, methotrexate and 5-fluorouracil (CMF) or methotrexate, mitozantrone and mitomycin C (MMM). One hundred and sixty-seven women over the age of 50 years were prescribed Tamoxifen 20 mg daily for at least 2 years. Patient details were recorded on a computer database updated at each out-patient clinic visit. This database was used to generate survival analyses based on the method of Kaplan-Meier. Nodal relapse is reported

N

I .........

nol

Fig. 3. Actuarial axillary relapse-free survival in 347 women according to axillary surgery.

Fig. 1. Actuarial overall survival in 347 women with early stage breast

9o

...........

!

is

TIHE SINCE SURGERY IYEARS)

T I ~ SINCE SUR6ERY {YEARS)

~

No surgery Surgery

qO c a s e s Z48 c a , e s I)irrere.ce

TABLE Il! Pattern of relapse in 16 women with lymph node recurrence. Relapse in axilla

~

0

...........

0

[ ...........

I ...........

I ...........

t ...........

I ...........

I ...........

1

1

2

3

4

5

6

7

TINE SINCE SUR6ERY (YEARS)

Fig. 2. Actuarial lymph node relapse-free survival in 347 women with early stage breast cancer.

Present Relapse in supraclavicular fossa

Present Absent Total

Absent

Total

5 9

2 331

7 340

14

333

347

86 TABLE IV Prior treatment at 21 sites of lymphatic relapse in 16 patients. (a) Axillary relapse Stage

Prior treatment

Total

Axillary relapse

Axillary RT

No axillary RT

nodal disease representing a successful locoregional salvage rate of 56~o. Of the eight patients relapsing in node sites without evidence of distant metastases salvage treatment has failed in only one case. The median follow-up from relapse in 33 months (range 16-52). Discussion

No axillary surgery No

99

6

1

5

3 4 1 0

0 0 0 0

3 4 1 2

Axillary surgery No 1-3 + >4+ Nx

138 74 34 2

(b) SCF relapse Stage

Prior treatment SCF relapse

Total

SCF RT

No SCF RT

0 1 1 0

1 4 0 2

No axillary surgery No

99

0

Axillary surgery No 1-3 + >4+ Nx

138 74 34 2

1 5 1 0

The morbidity of lymph node relapse in 16 patients is shown in Table V. Subsequent management depended upon co-existing sites of relapse; blood-borne metastases were present in 8/16 patients at the time of lymphatic relapse. Patients without evidence of haematogenous dissemination were treated by axillary surgery in six patients, with post-operative radiotherapy to the axilla and supraclavicular fossa in one patient and by supraclavicular irradiation alone in two patients. Four patients in this group received hormone therapy and two received chemotherapy in addition to local treatment. Palliative, local or systemic treatment was given to patients with distant metastases. At death or last follow-up four patients still had progressive nodal disease and three had stable asymptomatic nodal disease. The remaining nine patients had no assessable TABLE V Morbidity of lymphatic relapse in 16 patients. At time of relapse Local pain Neuropathy Lymphoedema Symptom-free Total

After subsequent treatment

3 2 2 9

1 1 2 12

16

16

Long-term survival data following mastectomy show an excess late mortality in those patients receiving postoperative radiotherapy [3,9,10]. This is mainly attributed to an excess of cardiovascular deaths resulting from attempted irradiation of the internal mammary chain. However, recent data comparing coronary artery doses from such techniques with modern tangential field megavoltage techniques demonstrates that a significant dose to the left anterior descending (LAD) coronary artery may still be delivered even when IMC is excluded [8]. There is also the possibility of radiation induced malignancy in long-term survivors, one recent series having found a relative risk for soft tissue sarcoma within the irradiated field of 1.8 [16]. Radical treatment of regional nodes, whether by surgery or irradiation is based upon early Halstedian concepts of the regional nodes as part of the stepwise spread of breast cancer from the primary site. This is no longer supported by data comparing radical treatment with more conservative approaches. For example, the NSABP found no difference in metastatic rate or survival between those patients having radical mastectomy or total mastectomy with regional radiotherapy including the axilla and those having total mastectomy alone leaving node areas untreated (although it is of note that 35 ~o of the latter group in fact had axillary surgery) [6]. Data from the Milan study comparing radical mastectomy with conservative surgery and radiotherapy includes a subgroup of node positive patients randomised to receive additional irradiation to the SCF and IMC. Disease-free survival at 5 years in these groups was 48.4~o in those receiving nodal radiotherapy compared to 73.9~o in those who did not but no difference in overall survival was seen [ 17]. Treatment of the lymphatic pathways does not influence survival because they are part of the generalised metastatic process. This is demonstrated by the finding that true isolated node relapse is far less frequent than node relapse in association with clinically apparent systemic disease; 64Yo of axillary relapses and 85~o of supraclavicular relapses in the European Curietherapy Group series [12] had associated distant metastases, in keeping with the figure of 5 0 ~ in this series. In the absence of an identifiable subgroup of women whose survival is improved by lymphatic radiotherapy, these findings highlight the

87 importance of confining lymphatic radiation fields to the minimum required in order to avoid unnecessary radiation-induced morbidity and mortality [15,19]. It is worth noting that whereas the risk of histologically positive axillary nodes is 30-40~o, in early stage breast cancer the rate of axillary relapse at 5 years in untreated lymph node areas in randomised trials may be only one-third of the rate of histological involvement [2,5]. The incidence of internal mammary chain node relapse is even lower compared to the incidence of histological involvement based on extended radical mastectomy [18]. The risk of supraclavicular adenopathy is closely linked to axillary involvement, being < 5 ~o at 10 years in node negative patients and rising to > 10To at 10 years in node positive patients [18]. In this retrospective review of our clinical practice, an axillary relapse rate of 5 ~o in 248 women submitted to axillary surgery was expected. With few exceptions, the operations were regarded as therapeutic by the surgeons but recurrences demonstrate the incompleteness of some resections. Post-operative radiotherapy may have prevented most of these eight axillary recurrences but at a considerable cost to the whole group at risk. A recent review of 1624 patients reports an axillary relapse rate following axillary dissection of 0.9~o when post-operative radiotherapy to the axilla was given and 2.2~o where no axillary irradiation was given [14]. In this series, seven patients relapsed in the supraclavicular fossa, five coinciding with axillary relapse and four with failure at systemic sites. Not a single instance of parasternal recurrence was observed. Successful local salvage rates after regional nodal failure of between 47 and 61 ~o are reported [ 12,14], in keeping with the 56~o seen in this series. The physical and psychological morbidity of regional recurrence in women submitted to a watch policy must be compared with the anticipated morbidity associated with routine prophylactic measures involving surgery and/or radiotherapy. A wide range of side-effects have been reported, some minor and others severe. The incidence and severity of these side-effects reported are also highly variable but it is certainly clear that combining surgery and radiotherapy is more damaging than either modality alone. Two European studies highlight the range of morbidities reported. The report of the European Curietherapy Group indicates very low rates of complications associated with radiotherapy and/or axillary dissection alone in 2884 patients [ 12]. The rates of arm oedema associated axillary dissection, axillary dissection plus radiotherapy or radiotherapy only were 2, 9 and 2~o, respectively. On the other hand, a very detailed review of 558 patients treated at the Institut Gustave Roussy reported higher rates of complications

associated with these treatments [4]. For example, the probability of complications (including oedema, reduced mobility, pain on movement, sensory and/or motor deficit, pectoral sclerosis) for axillary dissection, axillary dissection plus radiotherapy or radiotherapy alone were 7.2~o (5.0-10.2), 33.7~o (25.1-43.5) and 26.1~o (13.8-43.8), respectively, figures in brackets representing 95 ~o confidence limits. Clinically apparent brachial plexus injury is reported in < 1/% of patients after axillary surgery and irradiation [12,13 ]. These figures compare with a crude incidence of less than 1 ~o of our patients so far having persistent symptoms as a result of nodal relapse following a selective policy for nodal irradiation. It is difficult in evaluating the relative roles of surgery and irradiation, particularly in multicentre reviews, to isolate the interaction between the two treatment modalities at a pathophysiological level from the influence of individual surgical and radiotherapy techniques. For example, a wide range of dose-fractionation schedules have been employed in the treatment of breast cancer employing total doses to the axilla of between 40 and 75 Gy and fraction sizes of 2 to 4 Gy. Fraction size has been implicated in the high incidence of brachial plexus neuropathy seen in one series [ 13]. Radiotherapy techniques which fail to account for divergence of matching beams in all planes will inevitably lead to the risk of either overlap, and increased dose-related morbidity, or cold areas increasing the likelihood of relapse. One advantage in avoiding the axilla following surgical dissection is that it obviates the need for matching fields across this area which may in itself be a major factor in reducing treatment morbidity rather than any interaction between the two treatment modalities. An obvious question is how the morbidity and effectiveness of axillary clearance as sole treatment compares to axillary sampling with radiotherapy confined to node positive patients. A small randomised trial from Edinburgh has attempted to address this issue [1]. It randomised 94 women to either axillary clearance alone or axillary sampling with radiotherapy in node positive patients. There was a non-significant trend for objective and subjective measures of arm morbidity to be worse in the group randomised to axillary sampling plus selective radiotherapy. A larger trial is needed to resolve the relative merits of these alternative approaches. In summary, we have tried to avoid prophylactic lymphatic radiotherapy whenever possible. In situations where treatment of the lymphatic pathways is judged necessary we have avoided the combination of radiotherapy with surgery at the same site. The policy of selective lymphatic irradiation has resulted in less than 20 ~o of patients with early breast cancer receiving

88 post-operative irradiation to the nodal areas following lumpectomy and axillary dissection. Careful analysis of the relapsing patients demonstrates that for the majority of patients nodal relapse is associated with relapse at other sites and no additional criteria for extending the indications for nodal irradiation have so far been identified. The early outcome of a watch policy in clinically node negative women over 50 with an actuarial rate of node relapse at 5 years of 10~o is encouraging and nodal salvage has been successful. Assessments have been confined to physical complications, without attempting to compare the impact of treatment and

disease-related morbidity on life quality, including the emotional shock of experiencing isolated lymphatic relapse requiring further treatment. These aspects require further evaluation, but meanwhile it is our impression that the majority of women gain from this selective approach.

Acknowledgement We are indebted to Miss Jane Regan for the preparation of this manuscript.

References 1 Aitken, R. J., Gaze, M. N., Rodger, A., Chetty, U. and Forrest, A. P . M . Arm morbidity within a trial of mastectomy and either nodal sample with selective radiotherapy or axillary clearance. Br. J. Surg. 76: 568-571, 1989. 2 Cancer Research Campaign. Management of early cancer of the breast. Report on an international multicentre trial supported by the Cancer Research Campaign. Br. Med. J. 1: 1035-1038, 1976. 3 Cuzack, J., Stewart, H., Pert, R. et al. Overview ofrandomised trials of postoperative adjuvant radiotherapy in breast cancer. Cancer Treat. Rep. 71: 15-29, 1987. 4 Dewar, J. A., Sarrazin, D., Benhamon, E. et al. Management of the axiUa in conservatively treated breast cancer: 592 patients treated at Institut Gustave-Roussy. Int. J. Radiat. Oncol. Biol. Phys. 13: 475-481, 1987. 5 Fisher, B., Montague, E., Redmond, C. et al. Comparison of radical mastectomy with alternative treatments for primary breast cancer. A first report of results from a prospective randomised clinical trial. Cancer 39: 2827-2839, 1977. 6 Fisher, B., Wolmark, N., Redmond, C. et al. Findings from NSABP Protocol No B-04: comparison of radical mastectomy with alternative treatments. II. The clinical and biologic significance of medial-control breast cancers. Cancer 48: 1863-1872, 1981. 7 Fisher, B., Redmond, C., Poisson, R. et al. Eight year results of a randomised clinical trial comparing total mastectomy and lumptectomy with or without irradiation in the treatment of breast cancer. N. Engl. J. Med. 320: 822-828, 1989. 8 Fuller, S. A., Smith, R. E. A. and Dobbs, H . J . Cardiac doses received during irradiation. Br. J. Cancer 64, Suppl. XV: 10, 1991. 9 Haybittle, J. L., Brinkley, D., Houghton, J., A'Hem, R. P. and Baum, M. Postoperative radiotherapy and late mortality: evidence from the Cancer Research Campaign trial for early breast cancer. Br. Med. J. 298: 1611-1615, 1989. 10 Jones, J. M. and Ribeiro, G . G . Mortality patterns over 34 years of breast cancer patients in a clinical trial of postoperative radiotherapy. Clin. Radiol. 40: 204-208, 1989.

11 Overgaard, M., Christensen, J., Johansen, H., Nybo-Rasmussen, A., Rose, C., van der Kooy, P. et al. Evaluation of radiotherapy in high-risk breast cancer patients: report from the Danish breast cancer co-operative group trial (DBCG 82). Int. J. Radiat. Oncol. Biol. Phys. 19: 1121-1124, 1990. 12 Pierquin, B., Mazeron, J.-J. and Glaubiger, D. Conservative treatment of breast cancer in Europe: report of the Groupe Europeen de Curietherapie. Radiother. Oncol. 6: 187-198, 1986. 13 Powell, S., Cooke, J. and Parsons, C. Radiation-induced brachial plexus injury: follow-up of two different fractionation schedules. Radiother. Oncol. 18: 213-220, 1990. 14 Recht, A., Pierce, S. M., Abner, A. et al. Regional nodal failure after conservative surgery and radiotherapy for early stage breast carcinoma. J. Clin. Oncol. 9: 988-996, 1991. 15 Strender, L.-E., Lindahl, J. and Larsson, L.-E. Incidence of heart disease and functional significance of changes in the electrocardiogram 10 years after radiotherapy for breast cancer. Cancer 57: 929-934, 1986. 16 Taghian, A., de Vathaire, F., Terrier, P. et al. Long-term risk of sarcoma following radiation treatment for breast cancer. Int. J. Radiat. Oncol. Biol. Phys. 21: 361-367, 1991. 17 Veronesi, U., Sacozzi, R., Del Vecchio, M. et al. Comparing radical mastectomy with quadrantectomy, axillary dissection and radiotherapy in patients with small cancers of the breast. N. Engl. J. Med. 305: 6-11, 1981. 18 Veronesi, U. and Valagussa, P. Inefficacy of internal mammary dissection in breast cancer surgery. Cancer 47: 170-175, 1981. 19 Yarnold, J . R . Selective avoidance of lymphatic irradiation in the conservative management of breast cancer. Radiother. Oncol. 2: 79-92, 1984. 20 Yarnold, J. R. and Bloom, H. J . G . Radiation after limited and radical surgery. In: International Union Against Cancer (UICC), pp. 181-197. Current Treatment of Cancer Series. Editors: B. Hoogstraten, I. Burn and H. J. G. Bloom. Springer-Verlag, Berlin, 1989.