t .Wl
0360-3016/91 $3.00 Copyright 0 1991 Pergamon Press plc
IRI. J. Radiatron Oncology Ed Ph.vs., Vol. 2 I, pp. 299-310 Printed ,n the U.S.A. All rights reserved.
l Original Contribution
THE RESULTS OF RADIATION THERAPY FOR ISOLATED REGIONAL RECURRENCE AFTER MASTECTOMY FREDERICK SCHWAIBOLD, D.O.,’
BARBARA L. FOWBLE, M.D.,’
LOCAL
LAWRENCE J. SOLIN, M.D.,’
DELRAY J. SCHULTZ, M.S.2 AND ROBERT L. GOODMAN,
M.D.’
‘University of Pennsylvania School of Medicine and The Fox Chase Cancer Center; and ‘Department University of Pennsylvania, Cancer Center, Philadelphia, PA
of Statistics,
Between1967 and 1988 128 patientswithisolatedlocal-regional recurrence of breast cancer after mastectomy were treated with definitive radiation therapy. Recurrence was confined to a single site in 108 patients and multiple sites in 20. The chest wall was the most common location (86) and the supraclavicular region was the second most common (20). Surgical treatment for recurrence prior to irradiation consisted of excision of all gross disease in 78 patients and incisional biopsy in 49 patients. Irradiation was directed to the entire chest wall in 19% of patients with isolated chest wall recurrences and to the chest wall and regional nodes in 81%. In patients with isolated nodal failures, treatment was directed to the nodal site and chest wall in 87% and to the regional site alone in 13%. Patients with multiple sites received treatment to the chest wall and regional nodes in all cases. Electively treated sites usually received 4500-5000 cGy. Following excision of chest wall disease, the median dose was 6000 cGy. Gross disease on the chest wall received a median dose of 6100 cGy. Gross disease in nodal sites received a median dose of 5600 cGy; 66 patients received systemic therapy at recurrence. The 5-year actuarial local-regional control was 43%. In a multivariate analysis only the estrogen receptor status of the recurrence remained signifficant (p = .002). The 5-year actuarial survival was 49% with a relapse-free survival of 24%. In a multivariate analysis for survival, the disease-free interval (p = .007), local regional control (p = .006), and excisional biopsy for recurrence (p = .03) remained significant. In a multivariate analysis for relapse-free survival, the disease-free interval (p = .03), excisional biopsy (p = .OOOl), and the extent of axillary nodal involvement (p = .007) remained significant. In the subgroup of patients with a disease-free interval 2 24 months, excisional biopsy, and local regional control, the 5-year survival was 61% with a relapse-free survival of 59%. This subgroup represents 18% of the entire group and has a relatively good prognosis after recurrence. Post-mastectomy
local regional recurrence, Radiotherapy,
Prognostic factors.
University of Pennsylvania and the Fox Chase Cancer Center. Treatment of local regional recurrence and results in terms of local regional control, survival, and relapse-free survival are presented. Prognostic factors associated with a relatively favorable survival after recurrence are identified and optimal treatment of local regional recurrence is discussed.
INTRODUCTION
Local regional recurrence after mastectomy with or without adjuvant chemotherapy has an overall incidence of 10% and represents approximately 30% of all first recurrences (7, 19, 22, 24, 40). Local regional recurrence can be the source of substantial morbidity for patients with uncontrolled disease. Bedwinik et al. (6) noted that 62% of patients with uncontrolled disease experience one or more distressing symptoms for the duration of their life. Treatment for local regional recurrence has consisted of excision alone, radiation therapy with or without excision, systemic therapy alone, or combined with surgery and/ or irradiation. In an attempt to define the local control rates and survival for patients with local regional recurrence treated with radiation therapy (with or without systemic treatment), we have reviewed the experience of the
METHODS
AND
MATERIALS
Between 1967 and 1986,128 women with isolated local regional recurrence after mastectomy were treated with definitive radiation at the University of Pennsylvania and the Fox Chase Cancer Center. Patients with simultaneous or antecedent metastatic disease were excluded from this report. The median age at mastectomy was 57 (range 3 l-
Reprint requests to: Barbara Fowble, M.D., Department of Radiation Oncology, Hospital of the University of Pennsylvania, 2 Donner, 3400 Spruce St., Philadelphia PA 19 104.
Accepted for publication
299
25 January 199 1.
Advertiser Index INTERNATIONAL JOURNAL OF RADIATION ONCOLOGY BIOLOGY PHYSICS Volume 21, Number 2, July, 1991
Bucky X-Ray International, Inc. C & C Nuclear Systems Inc. CompHealth Gammex, Inc. Guthrie Clinic Haynes Radiation limited, Inc. Medi Physics Medicalibrations, Inc. Mitsubishi Nuclear Associates Oldelf t Osler Institute Philips Medical Systems Physician Services of America Radiarium Corporation Radiation Oncology Computer Sytems, Inc. Roerig Siemens Medical Systems, Inc. Small Systems Group, Inc. Theratronics International Limited Triangle Radiation Oncology Associates, Inc. URDC, Inc. University of Pennsylvania Varian For information on placing advertising, please contact: Arnold Kranzler, Rosemarie Fazzolari, David Tully or Raffaella Zastenchik at (914) 592-7700
xii
44OA,T-T Cover 4 ZZ: Cover 3
440E ... Ill Cover 2 Xi, xv
A-4
.
4i~~
440H ... xlll Insert . 440; 42 440G 440A vi
RT
301
for local regional recurrence 0 F. SCHWAIBOLD et~11.
and local regional control were performed using the Mantel Cox Test (32). The following factors were analyzed for local regional control, overall survival, and relapse-free survival: initial clinical and treatment-related factors: age, menopausal status, clinical T stage, hormone receptor status, initial axillary nodal status and chemotherapy at mastectomy: clinical factors of recurrence: site of recurrence, diseasefree interval, recurrent hormone receptor status, recurrence size and number of chest wall nodules; treatmentrelated factors: surgery for recurrence, radiation dose (for excised and gross disease) of region treated, local regional control, and systemic therapy for recurrence. A simultaneous analysis of these factors was performed using a multivariate Cox regression analysis (30).
LOCAL AND I OR REGIONAL CONTROL
viva1 curves
30 20 10 -
oi
. 0
, 10
.
1.1. 20
TINE
30
I. 40
8. 50
I 60
FROY RECURRENCE (MONTHS)
Fig. 1. Local and/or regional control.
RESULTS Local regional control
The failure rate in uninvolved sites with and without treatment is presented in Table 2. Elective treatment to uninvolved sites resulted in a 95% control rate. Twenty patients did not receive treatment to the supraclavicular region and one (5%) failed. Of note, 12/20 (60%) had pathologically negative axillary lymph nodes at mastectomy. Patients treated to the supraclavicular region had more advanced axillary disease. Thirty-nine percent had negative nodes and 33% had 24 positive nodes. Three patients did not receive elective treatment to the chest wall and two (67%) failed. Figure 1 demonstrates a 5-year actuarial local regional control rate of 43% for the entire group. Tables 3-5 present factors analyzed for local regional control. In a univariate analysis the following factors correlated significantly with local regional control: menopausal status, initial nodal status, disease-free interval, ER of recurrence, and excisional biopsy. Local regional control for excised patients
and those patients with gross disease is discussed separately below. Excised patients. The overall 5-year actuarial local regional control for the entire excised group was 48%. This is significantly better than unexcised patients, who had a control rate of 34% (p = .OO1). In general, excised patients had somewhat smaller recurrences than patients not undergoing excision. Seventy-six percent of those excised had disease measuring 14 cm compared to 5 1% of patients not excised. Of note, in patients excised the size of the recurrence did not appear to influence subsequent local regional control. Patients with lesions measuring 12 cm were controlled in 60% compared to 56% in patients with recurrences > 2 cm (range > 2-7 cm) (p = .78). In addition, higher doses of radiation after excision did not appear to improve local regional control rates. Patients receiving 4500-5000 cGy were controlled 55% compared to 43% in patients receiving >5000 cGy (p = .97). Following ex-
Table 2. Patterns of 2nd local regional failure after radiotherapy Location of 2nd local regional failure Excision of gross disease Site of initial recurrence Chest wall alone Nodal alone CW + nodal Gross residual disease Chest wall alone Nodal alone Chest + nodal
No. Tx
No. failed (%)
cw (%)
Nodal (%)
CW + nodal (W)
60 9 9
19 (31) 4 (44) 5 (55)
12 (20) O (0) 4 (44)
2 (3) 3 (33) O (0)
5 (8) I(111 l(Il)
26 15 9
16 (61) 7 (47)
11 (42) 2 (13)
1 (4) 4 (27)
4 (15) 1 (7)
4 (44)
3 (33)
0
I(II)
No.
Uninvolved at 1st failure Chest wall s’ clav Axilla Internal mammary
22 107 114 124
Treated
Failed (%)
Untreated
Failed (%)
19 87 51 30
1 (5) 6 (7) 2 (4) O (0)
3 20 63 94
2 (67) 1 (5) 3 (5) 2 (2)
302
1. J. Radiation Oncology 0 Biology 0 Physics
for 5 year actuarial local clinical and treatment related factors
Table 3. Prognostic factors analyzed
regional control-initial
5 yr. local
Factors
Number of patients
regional control %
40 85
57 37
.21
31 82
65 34
.05
59 26
49 53
.83
35 34 21 28
48 52 65
.48
53 24 40
43 59 33
49 76
32 50
p value
Age 150 >50 Menopausal status Pre/peri Post Clinical T Stage* Tl, T2 T3, T4 Initial receptor status ER+ ERPR+ PRInitial nodal status 0
l-3 24 Chemotherapy
.03
at
mastectomy Yes No
.61
* Significant in a multivariate
July 199 I, Volume 2 I, Number 2
>4 cm for a dose response analysis. However, with doses ranging from 4600 to 7000 cGy, the local regional control was 30%. The addition of chemotherapy did not significantly improve local regional control. However, in patients with gross disease, the administration of hormonal therapy resulted in a trend of improved control of 40% compared to 3 1% for patients not receiving chemotherapy (p = .lS). Table 2 demonstrates the pattern of failure after radiation therapy: 16/26 patients with gross disease on the chest wall failed, I 1 involving the chest wall only, 4 with the chest wall, and nodal failure and 1 nodal-only failure. Thus, 15116 (94%) subsequent failures had the chest wall as a component of disease. When the chest wall was involved as a single site or component of multiple sites, the subsequent chest wall failure rate was 61%. Multivariate analysis In an attempt to identify independent prognostic factors for local regional control, a multivariate analysis was performed. Only the ER of the recurrence (p = .002) remained significant.
.06
analysis, p = .002.
cision, chemotherapy given after recurrence improved local regional control. Patients excised had a control rate of 78% compared to 38% in patients not receiving chemotherapy (p = .15). However, this observation did not quite reach statistical significance. The number of chest wall nodules (single vs multiple) did not significantly correlate with local regional control after excision, 56% cGy for single nodules compared to 48% in patients with multiple nodules (p = .29). The pattern of second local regional recurrence can be seen in Table 2. For patients with chest wall disease alone, 17/19 (89%) who failed had a further component of chest wall failure and only 2 (11%) patients developed a regional failure alone. In 9 patients having nodal sites alone excised, 4 (44%) failed and all failed with a regional component. Patients having disease excised from both sites had a chest wall component in 5/9 (55%) and a nodal component in l/9 (11%). Unexcised patients. The initial complete response rate to irradiation was 74%. In patients who achieved a complete response, the 5-year local regional control was 37%. For the entire group of patients with gross disease, the overall 5-year actuarial local regional control was 34%. The size of the recurrence and dose of radiation given influenced local regional control. In patients with gross disease measuring 14 cm (range < l-4 cm) and receiving 26000 cGy, the local regional control was 57% compared to 15% in patients receiving ~6000 cGy (p = .04). There were too few patients (11) with gross disease measuring
Overall survival Figure 2 shows that the overall 5-year actuarial survival was 49% with relapse-free survival of 24%. Tables 6-9
Table 4. Prognostic factors analyzed for 5 year actuarial local regional control-clinical factors of recurrence
Factors Site of recurrence cw Nodal CW + nodal Mastectomy scar involved Yes No Disease free interval 2 cm Gross 52 cm 22 cm Number of CW nodules Single Multiple
Number of patients
5 yr. local regional control @)
84 20 18
45 45 49
.I8
56 22
42 76
.I2
65 60
42 48
28 35 16 24
60 25 71 41
,004
40 20
60 56
.78
18 16
43 24
.34
55 47
46 42
.08
* Significant in a multivariate analysis, p = .002.
p value
.04
.I1
RT
for local regional recurrence 0 F. SCHWAIBOLD
Table 5. Prognostic factors analyzed for 5 year actuarial local regional control-treatment related factors
Factors Chemo for recurrence Yes No Excised Yes No Gross Yes No Hormones for recurrence Yes No Excised Yes No Gross Yes No Surgery for recurrence Excision Incision Total dose-excised 4500-5000 cGy >5000 cGy Total dose-gross; 54 cm 2 cm (p = .02). Patients undergoing excisional biopsy prior to irradiation had improved survival compared to patients incisionally biopsied, 54% versus 40% (p = .002). Patients that were controlled had a superior 5-year survival to uncontrolled patients: 64% versus 35% (p = .OOl). For all patients, neither chemotherapy nor hormonal therapy at the time of the local regional recurrence influenced the overall survival. Patients receiving chemotherapy had a 5-year survival of 54% compared to 47% in those not treated with chemotherapy (p = .57). Systemic therapy at recurrence did not alter survival in excised patients. However, for patients with gross disease the use of hormones at recurrence significantly improved survival: 73% compared to 25% in patients not receiving chemotherapy (a = .05). Multivariate
analysis
In an attempt to identify independent prognostic factors for survival, a multivariate analysis was performed. Al-
OVERALL VS. RELAPSE
FREE SURVIVAL
TIME FROM RECURRENCE (YONTHS)
Fig. 2. Overall vs. relapse free survival.
I. J. Radiation Oncology 0 Biology0 Physics
304
July 1991, Volume 21, Number 2
Table 6. Prognostic factors analyzed for 5 year actuarial overall and relapse-free survival-initial and treatment related factors
Factors
5 yr. overall survival (W)
42 86
58 50
.72
26 24
.25
39 83
67 45
.35
35 18
.47
60 27
52 43
.29
36 35 22 29
78 48 49 66
.02
42 24 26 48
.I4
54 24 42
52 69 29
.03
51 77
52 50
.18
p value
.79
.62
.01*
analyzed in the model. All three factors remained significant; the disease-free interval 2 24 months (p = .007), use of excisional biopsy (p = .03), and local regional conThis subgroup of patients represented 18% trol (p = .006).
Table 7. Prognostic factors analyzed for 5-year actuarial overall and relapse-free survival-clinical Number of patients
5 yr. overall survival (%)
86 21 18
63 40 21
.20
57 23
63 41
66 62
Site of recurrence cw Nodal CW + nodal Mastectomy scar involved Yes No DFI ~24 mos. 224 mos. Recurrent receptors ER+ ERPR+ PRSize of recurrence 12 cm >2 cm No. chest wall nodules Single Multiple * Significant in multivariate
29 18
p value
analysis.
though there were several factors significant in a univariate analysis, due to the small numbers of patients with all data available for multivariate analysis, only disease-free interval, excisional biopsy, and local regional control were
Factors
5 yr. relapse-free survival
Number of patients
Age 4 nodes (p = .03). Similar results were reported by Chen et al. (9) who reported a 29% 5-year disease-free survival when l-3 nodes were positive versus 11% with >4 nodes. Mango et al. (3 1) noted a 42% 5year survival with l-3 nodes versus 18% when 24 nodes were involved. The disease-free interval is consistently identified as a major prognosticator of survival after local regional recurrence. Survival for patients with a disease free interval of t24 months ranges from lo-39% versus 37-70% in those with longer intervals ( 1, 3, 6, 12, 14, 15, 3 1, 35). The disease-free interval and initial stage of disease appear related. Patients with more advanced disease (Stage I, II, vs Ill) at mastectomy have shorter intervals to recurrence than those with less advanced disease (4, 23). The hormone receptor status at the time of initial surgery predicts for survival after local regional recurrence. Patients with estrogen receptor positive tumors survived longer after local regional recurrence than patients who
308
I. J. Radiation Oncology 0 Biology0 Physics
were estrogen receptor negative, 78% versus 48%, respectively (p = .02). The hormonal status of the recurrent disease also was a significant predictor of survival with estrogen receptor positive tumors having a superior survival than estrogen receptor negative tumors. The progesterone receptor was predictive of survival after recurrence with positive patients having a 5-year survival of 88% versus 46% (p = .006). The extent of the recurrent disease provides prognostic information. In the present series patients with a single nodule had a 5-year survival of 77% compared to 44% when there were multiple nodules (p = .O 1). Others have noted that a single nodule on the chest wall is associated with a 5-year survival of 40-60% compared to a 5-year survival ranging from lo-26% when there are multiple nodules present (6, 17, 18, 27, 3 1, 39). The size of the largest recurrence is also significant. Smaller lesions are associated with better survival than larger recurrences. Lesions I 2 cm had a 70% 5-year survival versus a 39% survival for larger recurrences (p = .02). Bedwinik et al. (6) reported a 5 1% 5-year survival for recurrence I 1 cm versus 18% for lesions measuring l-3 cm. Similarly, Mango et al. (3 1) reported improved survival in patients with lesions < 1 cm compared to larger recurrences. The influence of the site of recurrence on survival is unclear. Patients with chest wall only, nodal only, or chest wall plus nodal disease had a 5-year survival of 63%, 40%, and 21%, respectively (p = .20). Thus we found no statistically significant differences in survival for single sites of involvement versus multiple sites. However, some authors report reduced survival for nodal recurrences (15, 25,4 1) while others report no difference (4,6, 10, 17, 18). When multiple sites of involvement occur, several authors have reported reduced survival compared to patients with single sites of disease (1, 14, 25). Chemotherapy is commonly used at the time of a local regional recurrence because the majority of patients ultimately develop distant metastasis (1, 6, 37, 39, 42). In this series chemotherapy given at the time of recurrence did not appear to influence survival for the entire group. Patients receiving chemotherapy had a 5-year survival of 54% compared to 47% in patients not treated with chemotherapy (p = .57). However, in excised patients chemotherapy improved the overall and relapse-free survival. Patients receiving chemotherapy had an overall and relapse-free survival of 72% and 53% compared to 48% and 19% in patients not receiving chemotherapy. Although the differences did not reach statistical significance, the magnitude of the difference suggests that the improvement in overall and relapse-free survival could be clinically important. To date, there are no prospective randomized trials which have evaluated the effects of the combination of chemotherapy and radiation. Results of retrospective reviews are variable. Janjan et al. (27) noted no significant difference in 5-year survival for chemotherapy and radiation therapy vs radiation alone. Toonkel et al. (41) reported a 5-year survival of 35%
July I99 I, Volume 2 I, Number 2
for patients who received radiation alone versus 38% for the radiation plus chemotherapy group. There are less data on the impact of hormones given at recurrence. Deutsch et al. ( 14) found no difference in systemically treated patients (chemotherapy +- hormones). This series found that patients with gross disease appeared to have improved survival with the use of hormones at recurrence. Beck et al. (4) compared the results of patients treated with systemic therapy only to those receiving radiation only and reported a 5-year disease free survival of 25% for the systemically treated group versus 6% for the group receiving radiation only. Thus, the role of systemic therapy at local regional recurrences requires further investigation. In this series ultimate local regional control was associated with survival. Controlled patients had a 5-year survival of 64% versus 35% when uncontrolled (p = .OOl). Although factors that impact on local regional control might also contribute to survival, obtaining local regional control was significant in a multivariate analysis. In addition, the improved survival cannot be due to lead time bias because patients having the event (local regional recurrence) had a shorter survival than controlled patients. Several authors have previously noted that controlled patients have 5-year survivals ranging from 26-63% vs O37% for uncontrolled patients (9, 10, 12, 25, 27, 31, 39). The impact of excision on survival is difficult to assess since the size and extent of disease determine resectability and are also prognostic factors. Nevertheless, excision is associated with improved local regional control (I, 4). This series found that excised patients had a survival of 54% versus 40% in those not excised (p = .002). Similarly, Andry et al. (3) reported an improved 5-year survival for excised patients, 49% versus 29% when unexcised. An attempt was made to determine independent predictors of overall and relapse-free survival. Analyzing only the disease-free interval, excisional biopsy, and local regional control, all remained significant. Thus, the importance of the other factors as independent predictors remains uncertain. Similarly, for relapse-free survival, the disease-free interval, excisional biopsy, and pathologic status of the axilla remained significant. In conclusion, the optimal treatment of a local regional recurrence includes excisional biopsy followed by comprehensive radiation therapy to a volume that includes the entire chest wall and probably the supraclavicular fossa. Routine inclusion of the axilla and internal mammary region does not appear necessary and their inclusion may increase complications. Doses of 4500 cGy are adequate for electively treated sites. For excised regions a minimum dose of 5000 cGy is required. For gross disease higher doses are required. Lesions 5 4 cm require a minimum dose of 6000 cGy. No dose response is evident for larger lesions with up to a 50% failure rate despite doses of 7000 cGy (25). Despite the use of comprehensive irradiation and adequate doses, recurrence after radiation therapy occurs in 23-57% of patients. Efforts to improve the local regional control with the use of hyperthermia are in progress and remain investigational in previously
RT for local regional recurrence 0 F. SCHWAIBOLD
untreated patients. Although the majority of patients develop distant metastases, the role of chemotherapy at the time of local regional recurrence requires further investigation. In the present series, excised patients had an improvement in local regional control, overall survival, and relapse-free survival. Although these differences did not reach statistical significance, the magnitude of the difference suggests that the improvement may be clinically important for the decision-making process.
et ul.
309
The 5-year survival after a local regional recurrence ranges from 1O-50% with a disease free survival of 2 l26%. However, in a multivariate analysis, favorable factors for local regional control, overall and relapse-free survival have been identified. A favorable subgroup of patients comprising 18% of the entire group and having a 5-year survival of 6 1% and a relapse free survival of 59% is identified. Thus, a fatalistic approach to patients with isolated local regional recurrence is not justified.
REFERENCES 1. Aberzik, W. J.; Silver, B.; Henderson, 1. C.; Kady, B.; Harris, J. R. The use of radiotherapy in patients with isolated locoregional recurrence of breast carcinoma after mastectomy. Cancer 58:1214; 1986. 2. Allegra, J. C.; Barlock, A.; Huff, K. K.; Lippman, M. E. Changes in multiple or sequential estrogen receptor determinations in breast cancer. Cancer 45:792; 1980. 3. Andry, G.; Suciu, S.; Vito, P.; Faverly, D.; t’Hooft, A.; Verhesst, A.; Nogaret, J. M.; Mattheie, M. W. Locoregional recurrences after 649 modified radical mastectomies: incidence and significance. Eur. J. Surg. Oncol. 15:476; 1989. 4. Beck, T. M.; Hart, N. E.; Woodard, D. A.; Smith, C. E. Local or regionally recurrent carcinoma ofthe breast: Results of therapy in 121 patients. J. Clin. Oncol. 1:400; 1983. 5. Bedwinik, J. M.; Fineberg, B.; Lee, J.; Ocwieza, M. Analysis of failures following local treatment of isolated locoregional recurrence of breast cancer. Int. J. Radiat. Oncol. Biol. Phys. 7581; 1981. 6. Bedwinik, J. M.; Lee, J.; Fineberg, B.; Ocwieza, M. Prognostic indicators in patients with isolated local regional recurrence of breast cancer. Cancer 47:22; 198 1. 7. Bonadonna, G.; Valagussa, P.; Rossi, A.; Tancini, G.; Brambilla, C.; Zambetti, M.; Veronesi, U. Ten year experience with CMF-based adjuvant chemotherapy in resectable breast cancer. Breast Cancer Res. Treat. 5:95; 1985. 8. Campbell, F. C.; Elston, C. W.; Blarney, R. W.; Morris, A. H.; Nicholson, R. I.; Griffiths, K.; Haybittle, J. L. Quantitative oestradiol receptor values in primary breast cancer and response of metastases to endocrine therapy. Lancet 2: 1317; 1981. 9. Chen, K. K. Y.; Montague, E. D.; Oswald, M. J. Results of irradiation in treatment of locoregional breast cancer recurrence. Cancer 56: 1269; 1985. 10. Chu, F. C. H.; Lin, F. J.; Kim, J. H.; Huh, S. H.; Garmatis, C. .I. Locally recurrent carcinoma of the breast: results of radiation therapy. Cancer 37:2677; 1976. 1 1. Danckers. U. F.; Hamann, A.; Savage, J. L. Postoperative recurrence of breast cancer after 32 years: a case report and review of the literature. Surgery 47:656; 1960. 12. Danoff, B. F.; Coia, L. R.; Cantor, R. I.; Pajak, T. F.; Kramer, S. Locally recurrent breast carcinoma: the effect of adjuvant chemotherapy on prognosis. Radiology 1147:849; 1983. 13. Demaree, E. W. Local recurrence following surgery for cancer of the breast. Ann. Surg. 134:863; I95 1. 14. Deutsch, M.; Parsons, J. A.; Mittal, B. B. Radiation therapy for local regional recurrent breast carcinoma. Int. J. Radiat. Oncol. Biol. Phys. 12:2061; 1986. 15. Di Pietro, S.; Bertario, L.; Piva, L. Prognosis and treatment of loco-regional breast cancer recurrences: critical considerations on 120 cases. Tumori 66:33 1; 1980. 16. Donegan, W. L.; Perez-Mesa, C. M.; Watson, F. R. A biostatistical study of locally recurrent breast cancer. Surg. Gynecol. Obstet. 122:529; 1966. 17. Fentiman, I. S.; Matthews. P. N.; Davison, 0. W.; Millis,
18.
19.
20.
2 I. 22.
23.
24.
25.
26.
27.
28. 29.
30. 3 1.
32.
33.
R. R.; Hayward, J. L. Survival following local skin recurrence after mastectomy. Br. J. Surg. 72: 14; 1985. Fentiman, I. S.; Lavelle, M. A.; Caplan, D.; Milla, N.; Millis, R. R.; Hayward, J. L. The significance of supraclavicular fossa node recurrence after radical mastectomy. Cancer 57: 908; 1986. Fisher, B.; Fisher, E. R.; Redmond, C. Ten year results from the National Surgical Adjuvant Breast and Bowel Project (NSABP). Clinical trials evaluating the use of L-phenylalanine mustard (L-Pam) in the management of primary breast cancer. J. Clin. One. 4:929: 1986. Fletcher, G. H. Local results of irradiation in the primary management of localized breast cancer. Cancer 29:545; 1972. Fletcher, G. H.; Montague, E. D. Radical irradiation of advanced breast cancer. Am. J. Roentgenol. 93:573; 1965. Fowble, B.; Gray, R.; Gilchrist, K.; Goodman, R. L.: Taylor, S.: Tormey, D. C. Identification of a subgroup of patients with breast cancer and histologically positive nodes receiving adjuvant chemotherapy who may benefit from postoperative radiotherapy. J. Clin. Oncol. 6: 1107: 1988. Gilliand, M. D.; Barton, R. M.; Copeland, E. M. The implications of local recurrence of breast cancer as the first site of therapeutic failure. Ann. Surg. 197:284; 1983. Goldhirsch. A.: Gelba, R. Adjuvant treatment for early breast cancer: The Ludwig breast cancer studies in NC1 monograph. Proceedings of the NIH Consensus Development Conference on Adjuvant Chemotherapy and Endocrine Therapy for Breast Cancer Institute 55; 1986. Halverson, K. J.; Perez, C. A.: Kuske, R. R.: Garcia. D. M.: Simpson, J. R.: Fineberg, B. Isolated local-regional recurrence of breast cancer following mastectomy: radiotherapeutic management. Int. J. Radiat. Oncol. Biol. Phys. 19: 851; 1991. Jackson, S. M. Carcinoma of the breast-the significance of supraclavicular lymph node metastases. Clin. Radiol. 17: 107; 1966. Janjan, N. A.; McNeese, M. D.; Buzdar, A. U.; Montague, E. D.; Osward, M. J. Management of local regional recurrent breast cancer. Cancer 58: 1552; 1986. Kaplan, E. L.; Meier, P. Nonparametric estimation from incomplete observations. J. Am. Stat. Assoc. 53:457: 1958. Karabali-Dalomaga, S.; Souhami, R. L.; O’Higgins. N. J.; Soumilas, A.; Clark, C. G. Natural history and prognosis of recurrent breast cancer. Br. Med. J. 2:730; 1978. Lawless, J. F. Statistical models and methods for lifetime data. New York: John Wiley and Sons; 1982. Mango, L.; Bignardi, M.; Micheletti, E.; Bardelli. D.; Plebani, F. Analysis of prognostic factors in patients with isolated chest wall recurrence of breast cancer. Cancer 60:240; 1987. Mantel, N. Evaluation of survival data and two new rank order statistics arising in its consideration. Cancer Chemother. Rep. 50: 163; 1966. Oliver, D. R.; Sugarbaker, E. D. The significance of skin
310
34.
35.
36. 37.
38. 39.
40.
I. J. Radiation Oncology 0 Biology 0 Physics recurrences following radical mastectomy. Surg. Gynecol. Obstet. 85360; 1947. Olson, C. E.; Ansfield, F. J.; Richards, M. J.; Ramirez, G.; Davis, H. L. Review of soft tissue recurrence of breast cancer irradiated with and without actinomycin D. Cancer 39: 198 1; 1977. Patanaphan, V.; Salazar, 0. M.; Poussin-Rosillo, H. Prognosticators in recurrent breast cancer: a 15 year experience with irradiation. Cancer 54:228; 1984. Pawlias, K. T.; Dockerty, M. B.; Ellis, F. H. Late local recurrent carcinoma ofthe breast. Ann. Surg. 148: 192; 1958. Shimkin, M. B.; Lucia, E. L.; Low-Beer, V. A.; Bell, H. G. Recurrent cancer of the breast: Analysis of frequency, distribution and mortality at the University of California Hospital, 1918 to 1947, inclusive. Cancer 7:29; 1954. Spratt, J. S. Locally recurrent cancer after radical mastectomy. Cancer 20: 105 1; 1967. Stadler, B.; Kogelink, D. Local control and outcome of patients irradiated for isolated chest wall recurrences of breast cancer. Radiother. Oncol. 8: 105; 1987. Stefanik, D.; Goldberg, R.; Byrne, P.; Smith, F.; Uneo, W.; Smith, L.; Hatter, K.; Bachenheimer, L.; Beiser, C.; Drit-
July 1991, Volume 21, Number 2 schilo, A. Local regional failure in patients treated with adjuvant chemotherapy for breast cancer. J. Clin. Oncol. 3: 660-665; 1985. 4 1. Toonkel, L. N.; Fix, I.; Jacobson, L. H.; Wallach, C. B. The significance of local recurrence of carcinoma of the breast. Int. J. Radiat. Oncol. Biol. Phys. 9:33; 1983. 42. Tormey, D. C.; Weinberg, V. E.; Holland, J. F.; Weiss, R. B.; Glidwell, 0. J.; Perloff, M.; Falkson, G.; Falkson, H. C.; Henry, P.; Leone, L. A.; Rafla, S.; Ginsberg, S. J.; Silver, R. T.; Blom, J.; Carey, R. W.; Schein, P. S.; Lesnick, G. J. A randomized trial of five and three drug chemotherapy and chemoimmunotherapy in women with operable node positive breast cancer. J. Chn. Oncol. 1: 138; 1983. 43. Tough, 1. C. K. The significance of recurrence in breast cancer. Br. J. Surg. 53:897; 1966. 44. Valagussa, P.; Bonadonna, G.; Veronesi, U. Patterns of relapse and survival following radical mastectomy: analysis of 7 16 consecutive patients. Cancer 4 1: 1170; 1978. 45. Zimmerman, K. W.; Montague, E. D.; Fletcher, G. H. Frequency anatomical distribution and management of local recurrences after definitive therapy for breast cancer. Cancer 19:67: 1966.
0360-3016/91 $3.00 Copyright 0 1991 Pergamon Press plc
IRI. J. Radiatron Oncology Ed Ph.vs., Vol. 2 I, pp. 299-310 Printed ,n the U.S.A. All rights reserved.
l Original Contribution
THE RESULTS OF RADIATION THERAPY FOR ISOLATED REGIONAL RECURRENCE AFTER MASTECTOMY FREDERICK SCHWAIBOLD, D.O.,’
BARBARA L. FOWBLE, M.D.,’
LOCAL
LAWRENCE J. SOLIN, M.D.,’
DELRAY J. SCHULTZ, M.S.2 AND ROBERT L. GOODMAN,
M.D.’
‘University of Pennsylvania School of Medicine and The Fox Chase Cancer Center; and ‘Department University of Pennsylvania, Cancer Center, Philadelphia, PA
of Statistics,
Between1967 and 1988 128 patientswithisolatedlocal-regional recurrence of breast cancer after mastectomy were treated with definitive radiation therapy. Recurrence was confined to a single site in 108 patients and multiple sites in 20. The chest wall was the most common location (86) and the supraclavicular region was the second most common (20). Surgical treatment for recurrence prior to irradiation consisted of excision of all gross disease in 78 patients and incisional biopsy in 49 patients. Irradiation was directed to the entire chest wall in 19% of patients with isolated chest wall recurrences and to the chest wall and regional nodes in 81%. In patients with isolated nodal failures, treatment was directed to the nodal site and chest wall in 87% and to the regional site alone in 13%. Patients with multiple sites received treatment to the chest wall and regional nodes in all cases. Electively treated sites usually received 4500-5000 cGy. Following excision of chest wall disease, the median dose was 6000 cGy. Gross disease on the chest wall received a median dose of 6100 cGy. Gross disease in nodal sites received a median dose of 5600 cGy; 66 patients received systemic therapy at recurrence. The 5-year actuarial local-regional control was 43%. In a multivariate analysis only the estrogen receptor status of the recurrence remained signifficant (p = .002). The 5-year actuarial survival was 49% with a relapse-free survival of 24%. In a multivariate analysis for survival, the disease-free interval (p = .007), local regional control (p = .006), and excisional biopsy for recurrence (p = .03) remained significant. In a multivariate analysis for relapse-free survival, the disease-free interval (p = .03), excisional biopsy (p = .OOOl), and the extent of axillary nodal involvement (p = .007) remained significant. In the subgroup of patients with a disease-free interval 2 24 months, excisional biopsy, and local regional control, the 5-year survival was 61% with a relapse-free survival of 59%. This subgroup represents 18% of the entire group and has a relatively good prognosis after recurrence. Post-mastectomy
local regional recurrence, Radiotherapy,
Prognostic factors.
University of Pennsylvania and the Fox Chase Cancer Center. Treatment of local regional recurrence and results in terms of local regional control, survival, and relapse-free survival are presented. Prognostic factors associated with a relatively favorable survival after recurrence are identified and optimal treatment of local regional recurrence is discussed.
INTRODUCTION
Local regional recurrence after mastectomy with or without adjuvant chemotherapy has an overall incidence of 10% and represents approximately 30% of all first recurrences (7, 19, 22, 24, 40). Local regional recurrence can be the source of substantial morbidity for patients with uncontrolled disease. Bedwinik et al. (6) noted that 62% of patients with uncontrolled disease experience one or more distressing symptoms for the duration of their life. Treatment for local regional recurrence has consisted of excision alone, radiation therapy with or without excision, systemic therapy alone, or combined with surgery and/ or irradiation. In an attempt to define the local control rates and survival for patients with local regional recurrence treated with radiation therapy (with or without systemic treatment), we have reviewed the experience of the
METHODS
AND
MATERIALS
Between 1967 and 1986,128 women with isolated local regional recurrence after mastectomy were treated with definitive radiation at the University of Pennsylvania and the Fox Chase Cancer Center. Patients with simultaneous or antecedent metastatic disease were excluded from this report. The median age at mastectomy was 57 (range 3 l-
Reprint requests to: Barbara Fowble, M.D., Department of Radiation Oncology, Hospital of the University of Pennsylvania, 2 Donner, 3400 Spruce St., Philadelphia PA 19 104.
Accepted for publication
299
25 January 199 1.
Advertiser Index INTERNATIONAL JOURNAL OF RADIATION ONCOLOGY BIOLOGY PHYSICS Volume 21, Number 2, July, 1991
Bucky X-Ray International, Inc. C & C Nuclear Systems Inc. CompHealth Gammex, Inc. Guthrie Clinic Haynes Radiation limited, Inc. Medi Physics Medicalibrations, Inc. Mitsubishi Nuclear Associates Oldelf t Osler Institute Philips Medical Systems Physician Services of America Radiarium Corporation Radiation Oncology Computer Sytems, Inc. Roerig Siemens Medical Systems, Inc. Small Systems Group, Inc. Theratronics International Limited Triangle Radiation Oncology Associates, Inc. URDC, Inc. University of Pennsylvania Varian For information on placing advertising, please contact: Arnold Kranzler, Rosemarie Fazzolari, David Tully or Raffaella Zastenchik at (914) 592-7700
xii
44OA,T-T Cover 4 ZZ: Cover 3
440E ... Ill Cover 2 Xi, xv
A-4
.
4i~~
440H ... xlll Insert . 440; 42 440G 440A vi
RT
301
for local regional recurrence 0 F. SCHWAIBOLD et~11.
and local regional control were performed using the Mantel Cox Test (32). The following factors were analyzed for local regional control, overall survival, and relapse-free survival: initial clinical and treatment-related factors: age, menopausal status, clinical T stage, hormone receptor status, initial axillary nodal status and chemotherapy at mastectomy: clinical factors of recurrence: site of recurrence, diseasefree interval, recurrent hormone receptor status, recurrence size and number of chest wall nodules; treatmentrelated factors: surgery for recurrence, radiation dose (for excised and gross disease) of region treated, local regional control, and systemic therapy for recurrence. A simultaneous analysis of these factors was performed using a multivariate Cox regression analysis (30).
LOCAL AND I OR REGIONAL CONTROL
viva1 curves
30 20 10 -
oi
. 0
, 10
.
1.1. 20
TINE
30
I. 40
8. 50
I 60
FROY RECURRENCE (MONTHS)
Fig. 1. Local and/or regional control.
RESULTS Local regional control
The failure rate in uninvolved sites with and without treatment is presented in Table 2. Elective treatment to uninvolved sites resulted in a 95% control rate. Twenty patients did not receive treatment to the supraclavicular region and one (5%) failed. Of note, 12/20 (60%) had pathologically negative axillary lymph nodes at mastectomy. Patients treated to the supraclavicular region had more advanced axillary disease. Thirty-nine percent had negative nodes and 33% had 24 positive nodes. Three patients did not receive elective treatment to the chest wall and two (67%) failed. Figure 1 demonstrates a 5-year actuarial local regional control rate of 43% for the entire group. Tables 3-5 present factors analyzed for local regional control. In a univariate analysis the following factors correlated significantly with local regional control: menopausal status, initial nodal status, disease-free interval, ER of recurrence, and excisional biopsy. Local regional control for excised patients
and those patients with gross disease is discussed separately below. Excised patients. The overall 5-year actuarial local regional control for the entire excised group was 48%. This is significantly better than unexcised patients, who had a control rate of 34% (p = .OO1). In general, excised patients had somewhat smaller recurrences than patients not undergoing excision. Seventy-six percent of those excised had disease measuring 14 cm compared to 5 1% of patients not excised. Of note, in patients excised the size of the recurrence did not appear to influence subsequent local regional control. Patients with lesions measuring 12 cm were controlled in 60% compared to 56% in patients with recurrences > 2 cm (range > 2-7 cm) (p = .78). In addition, higher doses of radiation after excision did not appear to improve local regional control rates. Patients receiving 4500-5000 cGy were controlled 55% compared to 43% in patients receiving >5000 cGy (p = .97). Following ex-
Table 2. Patterns of 2nd local regional failure after radiotherapy Location of 2nd local regional failure Excision of gross disease Site of initial recurrence Chest wall alone Nodal alone CW + nodal Gross residual disease Chest wall alone Nodal alone Chest + nodal
No. Tx
No. failed (%)
cw (%)
Nodal (%)
CW + nodal (W)
60 9 9
19 (31) 4 (44) 5 (55)
12 (20) O (0) 4 (44)
2 (3) 3 (33) O (0)
5 (8) I(111 l(Il)
26 15 9
16 (61) 7 (47)
11 (42) 2 (13)
1 (4) 4 (27)
4 (15) 1 (7)
4 (44)
3 (33)
0
I(II)
No.
Uninvolved at 1st failure Chest wall s’ clav Axilla Internal mammary
22 107 114 124
Treated
Failed (%)
Untreated
Failed (%)
19 87 51 30
1 (5) 6 (7) 2 (4) O (0)
3 20 63 94
2 (67) 1 (5) 3 (5) 2 (2)
302
1. J. Radiation Oncology 0 Biology 0 Physics
for 5 year actuarial local clinical and treatment related factors
Table 3. Prognostic factors analyzed
regional control-initial
5 yr. local
Factors
Number of patients
regional control %
40 85
57 37
.21
31 82
65 34
.05
59 26
49 53
.83
35 34 21 28
48 52 65
.48
53 24 40
43 59 33
49 76
32 50
p value
Age 150 >50 Menopausal status Pre/peri Post Clinical T Stage* Tl, T2 T3, T4 Initial receptor status ER+ ERPR+ PRInitial nodal status 0
l-3 24 Chemotherapy
.03
at
mastectomy Yes No
.61
* Significant in a multivariate
July 199 I, Volume 2 I, Number 2
>4 cm for a dose response analysis. However, with doses ranging from 4600 to 7000 cGy, the local regional control was 30%. The addition of chemotherapy did not significantly improve local regional control. However, in patients with gross disease, the administration of hormonal therapy resulted in a trend of improved control of 40% compared to 3 1% for patients not receiving chemotherapy (p = .lS). Table 2 demonstrates the pattern of failure after radiation therapy: 16/26 patients with gross disease on the chest wall failed, I 1 involving the chest wall only, 4 with the chest wall, and nodal failure and 1 nodal-only failure. Thus, 15116 (94%) subsequent failures had the chest wall as a component of disease. When the chest wall was involved as a single site or component of multiple sites, the subsequent chest wall failure rate was 61%. Multivariate analysis In an attempt to identify independent prognostic factors for local regional control, a multivariate analysis was performed. Only the ER of the recurrence (p = .002) remained significant.
.06
analysis, p = .002.
cision, chemotherapy given after recurrence improved local regional control. Patients excised had a control rate of 78% compared to 38% in patients not receiving chemotherapy (p = .15). However, this observation did not quite reach statistical significance. The number of chest wall nodules (single vs multiple) did not significantly correlate with local regional control after excision, 56% cGy for single nodules compared to 48% in patients with multiple nodules (p = .29). The pattern of second local regional recurrence can be seen in Table 2. For patients with chest wall disease alone, 17/19 (89%) who failed had a further component of chest wall failure and only 2 (11%) patients developed a regional failure alone. In 9 patients having nodal sites alone excised, 4 (44%) failed and all failed with a regional component. Patients having disease excised from both sites had a chest wall component in 5/9 (55%) and a nodal component in l/9 (11%). Unexcised patients. The initial complete response rate to irradiation was 74%. In patients who achieved a complete response, the 5-year local regional control was 37%. For the entire group of patients with gross disease, the overall 5-year actuarial local regional control was 34%. The size of the recurrence and dose of radiation given influenced local regional control. In patients with gross disease measuring 14 cm (range < l-4 cm) and receiving 26000 cGy, the local regional control was 57% compared to 15% in patients receiving ~6000 cGy (p = .04). There were too few patients (11) with gross disease measuring
Overall survival Figure 2 shows that the overall 5-year actuarial survival was 49% with relapse-free survival of 24%. Tables 6-9
Table 4. Prognostic factors analyzed for 5 year actuarial local regional control-clinical factors of recurrence
Factors Site of recurrence cw Nodal CW + nodal Mastectomy scar involved Yes No Disease free interval 2 cm Gross 52 cm 22 cm Number of CW nodules Single Multiple
Number of patients
5 yr. local regional control @)
84 20 18
45 45 49
.I8
56 22
42 76
.I2
65 60
42 48
28 35 16 24
60 25 71 41
,004
40 20
60 56
.78
18 16
43 24
.34
55 47
46 42
.08
* Significant in a multivariate analysis, p = .002.
p value
.04
.I1
RT
for local regional recurrence 0 F. SCHWAIBOLD
Table 5. Prognostic factors analyzed for 5 year actuarial local regional control-treatment related factors
Factors Chemo for recurrence Yes No Excised Yes No Gross Yes No Hormones for recurrence Yes No Excised Yes No Gross Yes No Surgery for recurrence Excision Incision Total dose-excised 4500-5000 cGy >5000 cGy Total dose-gross; 54 cm 2 cm (p = .02). Patients undergoing excisional biopsy prior to irradiation had improved survival compared to patients incisionally biopsied, 54% versus 40% (p = .002). Patients that were controlled had a superior 5-year survival to uncontrolled patients: 64% versus 35% (p = .OOl). For all patients, neither chemotherapy nor hormonal therapy at the time of the local regional recurrence influenced the overall survival. Patients receiving chemotherapy had a 5-year survival of 54% compared to 47% in those not treated with chemotherapy (p = .57). Systemic therapy at recurrence did not alter survival in excised patients. However, for patients with gross disease the use of hormones at recurrence significantly improved survival: 73% compared to 25% in patients not receiving chemotherapy (a = .05). Multivariate
analysis
In an attempt to identify independent prognostic factors for survival, a multivariate analysis was performed. Al-
OVERALL VS. RELAPSE
FREE SURVIVAL
TIME FROM RECURRENCE (YONTHS)
Fig. 2. Overall vs. relapse free survival.
I. J. Radiation Oncology 0 Biology0 Physics
304
July 1991, Volume 21, Number 2
Table 6. Prognostic factors analyzed for 5 year actuarial overall and relapse-free survival-initial and treatment related factors
Factors
5 yr. overall survival (W)
42 86
58 50
.72
26 24
.25
39 83
67 45
.35
35 18
.47
60 27
52 43
.29
36 35 22 29
78 48 49 66
.02
42 24 26 48
.I4
54 24 42
52 69 29
.03
51 77
52 50
.18
p value
.79
.62
.01*
analyzed in the model. All three factors remained significant; the disease-free interval 2 24 months (p = .007), use of excisional biopsy (p = .03), and local regional conThis subgroup of patients represented 18% trol (p = .006).
Table 7. Prognostic factors analyzed for 5-year actuarial overall and relapse-free survival-clinical Number of patients
5 yr. overall survival (%)
86 21 18
63 40 21
.20
57 23
63 41
66 62
Site of recurrence cw Nodal CW + nodal Mastectomy scar involved Yes No DFI ~24 mos. 224 mos. Recurrent receptors ER+ ERPR+ PRSize of recurrence 12 cm >2 cm No. chest wall nodules Single Multiple * Significant in multivariate
29 18
p value
analysis.
though there were several factors significant in a univariate analysis, due to the small numbers of patients with all data available for multivariate analysis, only disease-free interval, excisional biopsy, and local regional control were
Factors
5 yr. relapse-free survival
Number of patients
Age 4 nodes (p = .03). Similar results were reported by Chen et al. (9) who reported a 29% 5-year disease-free survival when l-3 nodes were positive versus 11% with >4 nodes. Mango et al. (3 1) noted a 42% 5year survival with l-3 nodes versus 18% when 24 nodes were involved. The disease-free interval is consistently identified as a major prognosticator of survival after local regional recurrence. Survival for patients with a disease free interval of t24 months ranges from lo-39% versus 37-70% in those with longer intervals ( 1, 3, 6, 12, 14, 15, 3 1, 35). The disease-free interval and initial stage of disease appear related. Patients with more advanced disease (Stage I, II, vs Ill) at mastectomy have shorter intervals to recurrence than those with less advanced disease (4, 23). The hormone receptor status at the time of initial surgery predicts for survival after local regional recurrence. Patients with estrogen receptor positive tumors survived longer after local regional recurrence than patients who
308
I. J. Radiation Oncology 0 Biology0 Physics
were estrogen receptor negative, 78% versus 48%, respectively (p = .02). The hormonal status of the recurrent disease also was a significant predictor of survival with estrogen receptor positive tumors having a superior survival than estrogen receptor negative tumors. The progesterone receptor was predictive of survival after recurrence with positive patients having a 5-year survival of 88% versus 46% (p = .006). The extent of the recurrent disease provides prognostic information. In the present series patients with a single nodule had a 5-year survival of 77% compared to 44% when there were multiple nodules (p = .O 1). Others have noted that a single nodule on the chest wall is associated with a 5-year survival of 40-60% compared to a 5-year survival ranging from lo-26% when there are multiple nodules present (6, 17, 18, 27, 3 1, 39). The size of the largest recurrence is also significant. Smaller lesions are associated with better survival than larger recurrences. Lesions I 2 cm had a 70% 5-year survival versus a 39% survival for larger recurrences (p = .02). Bedwinik et al. (6) reported a 5 1% 5-year survival for recurrence I 1 cm versus 18% for lesions measuring l-3 cm. Similarly, Mango et al. (3 1) reported improved survival in patients with lesions < 1 cm compared to larger recurrences. The influence of the site of recurrence on survival is unclear. Patients with chest wall only, nodal only, or chest wall plus nodal disease had a 5-year survival of 63%, 40%, and 21%, respectively (p = .20). Thus we found no statistically significant differences in survival for single sites of involvement versus multiple sites. However, some authors report reduced survival for nodal recurrences (15, 25,4 1) while others report no difference (4,6, 10, 17, 18). When multiple sites of involvement occur, several authors have reported reduced survival compared to patients with single sites of disease (1, 14, 25). Chemotherapy is commonly used at the time of a local regional recurrence because the majority of patients ultimately develop distant metastasis (1, 6, 37, 39, 42). In this series chemotherapy given at the time of recurrence did not appear to influence survival for the entire group. Patients receiving chemotherapy had a 5-year survival of 54% compared to 47% in patients not treated with chemotherapy (p = .57). However, in excised patients chemotherapy improved the overall and relapse-free survival. Patients receiving chemotherapy had an overall and relapse-free survival of 72% and 53% compared to 48% and 19% in patients not receiving chemotherapy. Although the differences did not reach statistical significance, the magnitude of the difference suggests that the improvement in overall and relapse-free survival could be clinically important. To date, there are no prospective randomized trials which have evaluated the effects of the combination of chemotherapy and radiation. Results of retrospective reviews are variable. Janjan et al. (27) noted no significant difference in 5-year survival for chemotherapy and radiation therapy vs radiation alone. Toonkel et al. (41) reported a 5-year survival of 35%
July I99 I, Volume 2 I, Number 2
for patients who received radiation alone versus 38% for the radiation plus chemotherapy group. There are less data on the impact of hormones given at recurrence. Deutsch et al. ( 14) found no difference in systemically treated patients (chemotherapy +- hormones). This series found that patients with gross disease appeared to have improved survival with the use of hormones at recurrence. Beck et al. (4) compared the results of patients treated with systemic therapy only to those receiving radiation only and reported a 5-year disease free survival of 25% for the systemically treated group versus 6% for the group receiving radiation only. Thus, the role of systemic therapy at local regional recurrences requires further investigation. In this series ultimate local regional control was associated with survival. Controlled patients had a 5-year survival of 64% versus 35% when uncontrolled (p = .OOl). Although factors that impact on local regional control might also contribute to survival, obtaining local regional control was significant in a multivariate analysis. In addition, the improved survival cannot be due to lead time bias because patients having the event (local regional recurrence) had a shorter survival than controlled patients. Several authors have previously noted that controlled patients have 5-year survivals ranging from 26-63% vs O37% for uncontrolled patients (9, 10, 12, 25, 27, 31, 39). The impact of excision on survival is difficult to assess since the size and extent of disease determine resectability and are also prognostic factors. Nevertheless, excision is associated with improved local regional control (I, 4). This series found that excised patients had a survival of 54% versus 40% in those not excised (p = .002). Similarly, Andry et al. (3) reported an improved 5-year survival for excised patients, 49% versus 29% when unexcised. An attempt was made to determine independent predictors of overall and relapse-free survival. Analyzing only the disease-free interval, excisional biopsy, and local regional control, all remained significant. Thus, the importance of the other factors as independent predictors remains uncertain. Similarly, for relapse-free survival, the disease-free interval, excisional biopsy, and pathologic status of the axilla remained significant. In conclusion, the optimal treatment of a local regional recurrence includes excisional biopsy followed by comprehensive radiation therapy to a volume that includes the entire chest wall and probably the supraclavicular fossa. Routine inclusion of the axilla and internal mammary region does not appear necessary and their inclusion may increase complications. Doses of 4500 cGy are adequate for electively treated sites. For excised regions a minimum dose of 5000 cGy is required. For gross disease higher doses are required. Lesions 5 4 cm require a minimum dose of 6000 cGy. No dose response is evident for larger lesions with up to a 50% failure rate despite doses of 7000 cGy (25). Despite the use of comprehensive irradiation and adequate doses, recurrence after radiation therapy occurs in 23-57% of patients. Efforts to improve the local regional control with the use of hyperthermia are in progress and remain investigational in previously
RT for local regional recurrence 0 F. SCHWAIBOLD
untreated patients. Although the majority of patients develop distant metastases, the role of chemotherapy at the time of local regional recurrence requires further investigation. In the present series, excised patients had an improvement in local regional control, overall survival, and relapse-free survival. Although these differences did not reach statistical significance, the magnitude of the difference suggests that the improvement may be clinically important for the decision-making process.
et ul.
309
The 5-year survival after a local regional recurrence ranges from 1O-50% with a disease free survival of 2 l26%. However, in a multivariate analysis, favorable factors for local regional control, overall and relapse-free survival have been identified. A favorable subgroup of patients comprising 18% of the entire group and having a 5-year survival of 6 1% and a relapse free survival of 59% is identified. Thus, a fatalistic approach to patients with isolated local regional recurrence is not justified.
REFERENCES 1. Aberzik, W. J.; Silver, B.; Henderson, 1. C.; Kady, B.; Harris, J. R. The use of radiotherapy in patients with isolated locoregional recurrence of breast carcinoma after mastectomy. Cancer 58:1214; 1986. 2. Allegra, J. C.; Barlock, A.; Huff, K. K.; Lippman, M. E. Changes in multiple or sequential estrogen receptor determinations in breast cancer. Cancer 45:792; 1980. 3. Andry, G.; Suciu, S.; Vito, P.; Faverly, D.; t’Hooft, A.; Verhesst, A.; Nogaret, J. M.; Mattheie, M. W. Locoregional recurrences after 649 modified radical mastectomies: incidence and significance. Eur. J. Surg. Oncol. 15:476; 1989. 4. Beck, T. M.; Hart, N. E.; Woodard, D. A.; Smith, C. E. Local or regionally recurrent carcinoma ofthe breast: Results of therapy in 121 patients. J. Clin. Oncol. 1:400; 1983. 5. Bedwinik, J. M.; Fineberg, B.; Lee, J.; Ocwieza, M. Analysis of failures following local treatment of isolated locoregional recurrence of breast cancer. Int. J. Radiat. Oncol. Biol. Phys. 7581; 1981. 6. Bedwinik, J. M.; Lee, J.; Fineberg, B.; Ocwieza, M. Prognostic indicators in patients with isolated local regional recurrence of breast cancer. Cancer 47:22; 198 1. 7. Bonadonna, G.; Valagussa, P.; Rossi, A.; Tancini, G.; Brambilla, C.; Zambetti, M.; Veronesi, U. Ten year experience with CMF-based adjuvant chemotherapy in resectable breast cancer. Breast Cancer Res. Treat. 5:95; 1985. 8. Campbell, F. C.; Elston, C. W.; Blarney, R. W.; Morris, A. H.; Nicholson, R. I.; Griffiths, K.; Haybittle, J. L. Quantitative oestradiol receptor values in primary breast cancer and response of metastases to endocrine therapy. Lancet 2: 1317; 1981. 9. Chen, K. K. Y.; Montague, E. D.; Oswald, M. J. Results of irradiation in treatment of locoregional breast cancer recurrence. Cancer 56: 1269; 1985. 10. Chu, F. C. H.; Lin, F. J.; Kim, J. H.; Huh, S. H.; Garmatis, C. .I. Locally recurrent carcinoma of the breast: results of radiation therapy. Cancer 37:2677; 1976. 1 1. Danckers. U. F.; Hamann, A.; Savage, J. L. Postoperative recurrence of breast cancer after 32 years: a case report and review of the literature. Surgery 47:656; 1960. 12. Danoff, B. F.; Coia, L. R.; Cantor, R. I.; Pajak, T. F.; Kramer, S. Locally recurrent breast carcinoma: the effect of adjuvant chemotherapy on prognosis. Radiology 1147:849; 1983. 13. Demaree, E. W. Local recurrence following surgery for cancer of the breast. Ann. Surg. 134:863; I95 1. 14. Deutsch, M.; Parsons, J. A.; Mittal, B. B. Radiation therapy for local regional recurrent breast carcinoma. Int. J. Radiat. Oncol. Biol. Phys. 12:2061; 1986. 15. Di Pietro, S.; Bertario, L.; Piva, L. Prognosis and treatment of loco-regional breast cancer recurrences: critical considerations on 120 cases. Tumori 66:33 1; 1980. 16. Donegan, W. L.; Perez-Mesa, C. M.; Watson, F. R. A biostatistical study of locally recurrent breast cancer. Surg. Gynecol. Obstet. 122:529; 1966. 17. Fentiman, I. S.; Matthews. P. N.; Davison, 0. W.; Millis,
18.
19.
20.
2 I. 22.
23.
24.
25.
26.
27.
28. 29.
30. 3 1.
32.
33.
R. R.; Hayward, J. L. Survival following local skin recurrence after mastectomy. Br. J. Surg. 72: 14; 1985. Fentiman, I. S.; Lavelle, M. A.; Caplan, D.; Milla, N.; Millis, R. R.; Hayward, J. L. The significance of supraclavicular fossa node recurrence after radical mastectomy. Cancer 57: 908; 1986. Fisher, B.; Fisher, E. R.; Redmond, C. Ten year results from the National Surgical Adjuvant Breast and Bowel Project (NSABP). Clinical trials evaluating the use of L-phenylalanine mustard (L-Pam) in the management of primary breast cancer. J. Clin. One. 4:929: 1986. Fletcher, G. H. Local results of irradiation in the primary management of localized breast cancer. Cancer 29:545; 1972. Fletcher, G. H.; Montague, E. D. Radical irradiation of advanced breast cancer. Am. J. Roentgenol. 93:573; 1965. Fowble, B.; Gray, R.; Gilchrist, K.; Goodman, R. L.: Taylor, S.: Tormey, D. C. Identification of a subgroup of patients with breast cancer and histologically positive nodes receiving adjuvant chemotherapy who may benefit from postoperative radiotherapy. J. Clin. Oncol. 6: 1107: 1988. Gilliand, M. D.; Barton, R. M.; Copeland, E. M. The implications of local recurrence of breast cancer as the first site of therapeutic failure. Ann. Surg. 197:284; 1983. Goldhirsch. A.: Gelba, R. Adjuvant treatment for early breast cancer: The Ludwig breast cancer studies in NC1 monograph. Proceedings of the NIH Consensus Development Conference on Adjuvant Chemotherapy and Endocrine Therapy for Breast Cancer Institute 55; 1986. Halverson, K. J.; Perez, C. A.: Kuske, R. R.: Garcia. D. M.: Simpson, J. R.: Fineberg, B. Isolated local-regional recurrence of breast cancer following mastectomy: radiotherapeutic management. Int. J. Radiat. Oncol. Biol. Phys. 19: 851; 1991. Jackson, S. M. Carcinoma of the breast-the significance of supraclavicular lymph node metastases. Clin. Radiol. 17: 107; 1966. Janjan, N. A.; McNeese, M. D.; Buzdar, A. U.; Montague, E. D.; Osward, M. J. Management of local regional recurrent breast cancer. Cancer 58: 1552; 1986. Kaplan, E. L.; Meier, P. Nonparametric estimation from incomplete observations. J. Am. Stat. Assoc. 53:457: 1958. Karabali-Dalomaga, S.; Souhami, R. L.; O’Higgins. N. J.; Soumilas, A.; Clark, C. G. Natural history and prognosis of recurrent breast cancer. Br. Med. J. 2:730; 1978. Lawless, J. F. Statistical models and methods for lifetime data. New York: John Wiley and Sons; 1982. Mango, L.; Bignardi, M.; Micheletti, E.; Bardelli. D.; Plebani, F. Analysis of prognostic factors in patients with isolated chest wall recurrence of breast cancer. Cancer 60:240; 1987. Mantel, N. Evaluation of survival data and two new rank order statistics arising in its consideration. Cancer Chemother. Rep. 50: 163; 1966. Oliver, D. R.; Sugarbaker, E. D. The significance of skin
310
34.
35.
36. 37.
38. 39.
40.
I. J. Radiation Oncology 0 Biology 0 Physics recurrences following radical mastectomy. Surg. Gynecol. Obstet. 85360; 1947. Olson, C. E.; Ansfield, F. J.; Richards, M. J.; Ramirez, G.; Davis, H. L. Review of soft tissue recurrence of breast cancer irradiated with and without actinomycin D. Cancer 39: 198 1; 1977. Patanaphan, V.; Salazar, 0. M.; Poussin-Rosillo, H. Prognosticators in recurrent breast cancer: a 15 year experience with irradiation. Cancer 54:228; 1984. Pawlias, K. T.; Dockerty, M. B.; Ellis, F. H. Late local recurrent carcinoma ofthe breast. Ann. Surg. 148: 192; 1958. Shimkin, M. B.; Lucia, E. L.; Low-Beer, V. A.; Bell, H. G. Recurrent cancer of the breast: Analysis of frequency, distribution and mortality at the University of California Hospital, 1918 to 1947, inclusive. Cancer 7:29; 1954. Spratt, J. S. Locally recurrent cancer after radical mastectomy. Cancer 20: 105 1; 1967. Stadler, B.; Kogelink, D. Local control and outcome of patients irradiated for isolated chest wall recurrences of breast cancer. Radiother. Oncol. 8: 105; 1987. Stefanik, D.; Goldberg, R.; Byrne, P.; Smith, F.; Uneo, W.; Smith, L.; Hatter, K.; Bachenheimer, L.; Beiser, C.; Drit-
July 1991, Volume 21, Number 2 schilo, A. Local regional failure in patients treated with adjuvant chemotherapy for breast cancer. J. Clin. Oncol. 3: 660-665; 1985. 4 1. Toonkel, L. N.; Fix, I.; Jacobson, L. H.; Wallach, C. B. The significance of local recurrence of carcinoma of the breast. Int. J. Radiat. Oncol. Biol. Phys. 9:33; 1983. 42. Tormey, D. C.; Weinberg, V. E.; Holland, J. F.; Weiss, R. B.; Glidwell, 0. J.; Perloff, M.; Falkson, G.; Falkson, H. C.; Henry, P.; Leone, L. A.; Rafla, S.; Ginsberg, S. J.; Silver, R. T.; Blom, J.; Carey, R. W.; Schein, P. S.; Lesnick, G. J. A randomized trial of five and three drug chemotherapy and chemoimmunotherapy in women with operable node positive breast cancer. J. Chn. Oncol. 1: 138; 1983. 43. Tough, 1. C. K. The significance of recurrence in breast cancer. Br. J. Surg. 53:897; 1966. 44. Valagussa, P.; Bonadonna, G.; Veronesi, U. Patterns of relapse and survival following radical mastectomy: analysis of 7 16 consecutive patients. Cancer 4 1: 1170; 1978. 45. Zimmerman, K. W.; Montague, E. D.; Fletcher, G. H. Frequency anatomical distribution and management of local recurrences after definitive therapy for breast cancer. Cancer 19:67: 1966.
