- Email: [email protected]
Premenopausal breast cancer patients treated with conservative surgery, radiotherapy and adjuvant chemotherapy have a low risk of local failure
Inl. J. Radiatm Oncdogy Bid. Phys., Vol. Printed in the U.S.A. All rights reserved.
17,pp.7I l-7
0360-3016/89 $3.00 + .OO Copyright 0 1989 Pergamon Pres plc
I7
??Original Contribution
PREMENOPAUSAL BREAST CANCER PATIENTS TREATED WITH CONSERVATIVE SURGERY, RADIOTHERAPY +ND ADJUVANT CHEMOTHERAPY HAVE A LOW RISK OF LOCAL FAILURE MARY ANN ROSE,M.D.,’ I. CRAIG HENDERSON,M.D.,2 REBECCAGELMAN, PH.D.,~ JOHN BOYAGES,M.D.,’ STACEYM. GORE, M.D.,4 STEVENCOME, M.D.,4 BARBARASILVER,B.A.,’ ABRAM RECHT,M.D.,’ JAMESL. CONNOLLY, M.D.,5 STUARTJ. SCHNITT,M.D.,5 C. NORMAN COLEMAN,M.D.’ AND JAY R. HARRIS,M.D.’ ‘Joint Center for Radiation Therapy and Department of Radiation Therapy, Harvard Medical School; *Division of Medical Oncology, Dana-Farber Cancer Institute and Harvard Medical School; 3Division of Biostatistics and Epidemiology, Dana-Farber Cancer Institute and Harvard School of Public Health; 4Division of Hematology-Oncology, Department of Medicine, Beth Israel Hospital and Harvard Medical School; and ‘Department of Pathology, Beth Israel Hospital and Harvard Medical School, Boston, MA The use of adjuvant chemotherapy in premenopausal breast cancer patients with positive nodes is now routine, but the optimal local treatment of these patients is uncertain. To determine the effect of adjuvant chemotherapy on the likelihood of local recurrence as the first site of failure in premenopausal patients treated with conservative surgery (CS) and radiotherapy (RT), we examined the outcome of 74 patients treated with CS, RT, and adjuvant chemotherapy and compared it to the outcome in 192 patients treated with CS and RT alone. Adjuvant chemotherapy consisted of four or more cycles of either a doxorubicin-containing regimen or cyclophosphamide, methotrexate, and 5fluorouracil. All patients were less than 50 years old, had UICC-AJCC Stage I or II breast cancer treated between 1968 and 1981, had gross excision of the primary tumor, and had a total radiation dose to the primary tumor bed of 26000 cCy. Factors predicting for local recurrence, such as extensive intraductal carcinoma and age less than 35, were equivalent in the two groups. Women treated with adjuvantchemotherapyhad significantlyworse T- and N-stages than women treated with conservative surgery and radiotherapy alone: 61% versus 36% had T2 tumors (p = 0.0003), 34% versus 6% had clinically positive nodes (p < O.OOOl),and 97% versus 4% had pathologically positive nodes (p < 0.0001). Despite the poorer prognosis of patients treated with adjuvant chemotherapy, within 5 years of diagnosis, 4% of patients who received adjuvant chemotherapy had their initial relapse in the breast and 24% had initial failure elsewhere, compared with 15% local failure first and 14% failure elsewhere first for those treated without chemotherapy (p = 0.01). We conclude that premenopausal patients with positive nodes treated with combined modality therapy (conservative surgery, radiation therapy, and adjuvant chemotherapy) have a low risk of local recurrence as a first site of failure. These results suggest a possible interaction between radiation therapy and chemotherapy in their effects on local tumor control. Breast cancer, Radiotherapy,
Local recurrence, Adjuvant chemotherapy.
INTRODUCTION
found to have pathologically positive lymph nodes are
receiving adjuvant chemotherapy. Thus, premenopausal patients with positive nodes are being considered for treatment with conservative surgery, radiation therapy, and adjuvant chemotherapy. This approach has raised several issues including the influence of chemotherapy on the cosmetic outcome, the possibility of drug dose reductions necessitated by radiation therapy, and the optimal sequencing of these two modalities. Given these concerns and a perception that mastectomy may provide better local tumor control, some clinicians have recommended mastectomy in premenopausal patients with positive
Presented in part at the Annual Meeting of the American Society for Therapeutic Radiology and Oncology, New Orleans, October 1988.
Reprint requests to: Mary Ann Rose, M.D., Joint Center for Radiation Therapy, 50 Binney Street, Boston, MA 02 115. Accepted for publication 13 April 1989.
In recent years, two major trends have evolved in the management of patients with early stage breast cancer. First, increasing numbers of women are being treated with conservative surgery (CS) and radiation therapy as an alternative to mastectomy. This is based on the results of prospective randomized trials comparing CS and radiation with mastectomy which have demonstrated equivalent rates of survival (8, 24). Second, premenopausal patients
711
712
I. J. Radiation
Oncology
0 Biology 0 Physics
nodes. However, the results from the National Surgical Adjuvant Breast Project (NSABP) B-06 trial suggest that the use of CS, radiotherapy, and adjuvant chemotherapy provides a high level of local tumor control in node-positive patients ( 10). To study this issue further, we examined the likelihood of breast recurrence among node-positive premenopausal women treated with CS, radiotherapy, and adjuvant chemotherapy, and compared it to the outcome in premenopausal patients treated with CS and radiation therapy alone. Our results indicate that premenopausal patients with positive nodes treated with CS, radiotherapy, and adjuvant chemotherapy have a low risk of local recurrence as a first site of failure, suggesting an interaction between radiotherapy and chemotherapy in achieving local tumor control.
METHODS
AND MATERIALS
Three hundred and thirty-two premenopausal women with UICC-AJCC clinical Stage I or II breast cancer were treated at the Joint Center for Radiation Therapy (JCRT) between July 1968 and December 1981 (1). (Note that staging was done according to the system adopted in 1983.) For the purpose of this analysis, premenopausal was defined as age less than 50 years. To eliminate the confounding effect of possibly inadequate treatment, seven patients who had less than excisional biopsy (i.e., needle biopsy or incisional biopsy only) prior to radiotherapy and 46 patients who received a total dose of less than 6000 cGy to the regional of the primary tumor were excluded from the analysis. In addition, 12 patients who received either single agent chemotherapy or regimens which did not include cyclophosphamide, methotrexate, and 5-fluorouracil (CMF) or doxorubicin (Adriamycin), and one patient who received only two cycles of CMF were excluded from the analysis. The study population thus contained 266 patients. Patient accrual was not uniform throughout the study period; 1 1 l/ 192 patients (58%) who did not receive adjuvant chemotherapy and 41174 patients (55%) who did receive adjuvant chemotherapy were treated in the last 2 years of the study (1980-8 1). Follow-up was obtained as of April 1987 with a median follow-up of 75 months (range, 37-l 50 months). All patients were treated with conservative surgery and radiation therapy. Surgery consisted of an excisional biopsy, defined as gross total removal of the primary tumor without histologic confirmation of negative margins. In addition, 239 patients (90%) also underwent axillary nodal dissection which was usually confined to Level I and II nodes. Details of the radiation treatment have been described elsewhere ( 19). All patients received whole breast irradiation, typically to a total dose of 4500-5000 cGy over 5 weeks. A supplemental “boost” dose to the primary tumor site was given to all patients. This was accomplished through Iridium- 192 implantation in 254 patients and by electron beam or photon irradiation in 12 patients. The
October
1989. Volume
17, Number
4
total dose to the region of the primary tumor ranged from 6000 to 7800 cGy (median, 6800 cGy). Radiation doses were comparable for the chemotherapy and nonchemotherapy groups: the ranges were 6000-7250 (median, 6800) and 6000-7800 (median, 6780) respectively. Fortyfive patients received radiation to the breast only. The remaining patients received radiation to the breast and to draining regional lymph nodes. Seventy-four patients received a minimum of four cycles of adjuvant CMF or an Adriamycin-containing regimen. Patients received their chemotherapy either sequentially, with chemotherapy commencing after completion of radiation (41 patients); concurrently, with chemotherapy commencing at the same time as radiation (24 patients); or in a “sandwich” fashion, with at least one cycle of chemotherapy administered prior to radiation (nine patients). The details of the chemotherapy administered in these patients have been previously described ( 12). Only one patient received adjuvant hormonal therapy (Tamoxifen) in addition to chemotherapy. Clinico-pathologic features examined in this study included primary tumor size, clinical nodal status, pathologic nodal status, lymphatic invasion, and the presence and extent of intraductal carcinoma within the infiltrating primary tumor and in adjacent grossly-normal breast tissue. We have previously shown that the most significant predictor of breast recurrence among our patients is an “extensive intraductal component” (EIC) in the primary surgical specimen (2 1). This factor is defined as the presence of intraductal carcinoma both within the primary tumor (comprising at least 25% of the tumor area) and in adjacent grossly-normal breast tissue. Microscopic evaluation of margins was not available in most patients. Local recurrence was defined as the detection of cancer in the treated breast tissue or overlying skin occurring before the development of distant metastases. Patients were censored from the calculation of breast recurrence at the time of last follow-up or the discovery of regional or metastatic disease. Among the 74 patients who received chemotherapy, one was noted to have a recurrence in the skin overlying the treated breast 14 months after distant failure. Among the 192 patients who did not receive chemotherapy, one patient was noted to have a skin recurrence 1 I months after distant failure. These failures are not counted in the analysis. One patient who received chemotherapy had an infraclavicular node failure and three patients who did not receive chemotherapy had axillary nodal failures. These are counted as distant failures for the purpose of analysis. Local failures were classified by location in the breast relative to the primary tumor site as true recurrences (TR, within the area of the boost), marginal misses (MM, adjacent to the boost area), recurrences elsewhere (E, in other quadrants of the breast), and skin recurrences (S). The risk of local recurrence was evaluated by the use of actuarial calculations. Actuarial curves were calculated by the Kaplan-Meier method and comparisons between
Chemotherapy radiotherapy effect in breast cancer 0 M. A. ROSE et al.
curves were made using the two-tailed “naive” log-rank test. The Kaplan-Meier estimates, however, can only be shown to be unbiased and maximum likelihood estimates if time to distant and time to local failure are statistically independent. This assumption can be questioned in this data set. Such non-independent censoring could cause the log-rank test to be biased in either direction (too significant or not significant enough). To minimize the problem of competing risks, distant recurrence was defined as the detection of cancer outside of the treated breast or death from intercurrent disease (not breast cancer) before the development of local recurrence. Thus three patients, one who received chemotherapy and two who did not, died without evidence of recurrence and were counted as distant failures. Patients were censored from the calculation of distant failure at the time of last follow-up or at the time of local failure. Three patients (one who received chemotherapy and two who did not) had simultaneous local and distant failure. These patients with simultaneous failures were counted as distant failures because distant failure determined their subsequent therapy. The distribution of the site of first failure (local, distant, none) within 5 years of treatment was compared for the chemotherapy and nonchemotherapy groups using the multinomial test (Fisher exact test) (11). The multinomial test allows an unbiased test of whether the distribution of causes of failure (distant or local) differed in the two treatment groups. Differences in patient characteristics between the two groups were analyzed using the Fisher exact test and the Wilcoxon test (for age, follow-up, and radiation doses). Two-sided p-values of 0.05 or less were considered statistically significant.
RESULTS A comparison of characteristics between the 74 premenopausal patients who were treated with CS, radiotherapy. and adjuvant chemotherapy (combined modality treatment) and the 192 premenopausal patients who were
Table 1. Patient
treated with CS and radiotherapy alone is presented in Table 1. The age distribution and follow-up distribution were comparable in the two groups. Patients who received combined modality treatment generally presented in a more advanced stage than patients who did not receive chemotherapy. The percentage of T2 tumors in the group receiving adjuvant chemotherapy was significantly higher than in the group not receiving drugs: 6 1% versus 36%, p = 0.0003. Only 66% of patients who received chemotherapy were clinically staged NO/N 1A,compared to 94% of patients who did not receive drugs (p < 0.000 1). Similar percentages of patients in each group underwent axillary nodal dissection. However, histologically confirmed lymph node metastases were present in 97% of patients who received chemotherapy compared to only 4% of patients who did not receive chemotherapy (p < 0.0001). The percentage of patients with an extensive intraductal component (among evaluable patients with infiltrating ductal histology) was similar for the group that received chemotherapy and for those who did not: 38% versus 35%, p = NS. The percentage of patients with respectively, lymphatic invasion was higher (36%) among patients who received chemotherapy than among patients who did not receive chemotherapy (17%) (p = 0.003). Overall, patients treated with conservative surgery, radiotherapy, and adjuvant chemotherapy had significantly worse clinical and pathologic stages than patients treated with conservative surgery and radiotherapy alone. Figure 1 shows the actuarial calculations of time to breast failure as the first site of recurrence for the 74 patients treated with CS, radiotherapy, and adjuvant chemotherapy compared to the 192 patients who were treated with CS and radiotherapy alone. The 5-year actuarial rate of local failure was 5% for patients who received chemotherapy and 17% for those who did not receive chemotherapy (p = 0.03, naive log-rank test). Within 5 years of diagnosis, three of 74 patients given chemotherapy failed first locally (crude rate of 4%) compared with 28 of 192 patients not given chemotherapy (crude rate of 15%). The local failure types for the chemotherapy group were one TR, one MM, and one S. The failure types for the
characteristics
CS+RT No. pts. % Tl % NO/IA % axillary dissection % nodes pathologically positive % with lymphatic invasion % extensive intraductal component Median age (range) % age < 35 Median follow-up (range) Median cGY to breast (range) CS = Conservative
surgery; RT = Radiation
CS+RT+CT 74 39 66 95 97 36 38 40 (29-49) 15 76 mos (37-121) 6800 (6000-7250)
192 64 94 88 4 17 :: (25-49) 16 75 mos (49-150) 6780 (6000-7800) therapy;
713
CT = Adjuvant
chemotherapy.
p-value
0.0003
I. J. Radiation Oncology 0 Biology 0 Physics
0 No Chemotherapy (192) ?? Chemotherapy (74)
T/ME fyeofs/ Fig. 1. Actuarial risk of breast failure as the first site of recurrence for the 74 patients treated with adjuvant chemotherapy compared with the 192 patients who did not receive adjuvant chemotherapy (p = 0.03, naive log-rank test). The numbers in parentheses indicate patients at risk at 5 years.
that did not receive chemotherapy were 15 TRs, seven MMs, five Es, and one S. The sequencing of radiation and chemotherapy did not appear to have a significant effect on local recurrence, but the number of failures in the group that received chemotherapy was quite small. The 5-year actuarial rate of distant failure was 25% for patients who received chemotherapy compared to 14% for those who did not receive chemotherapy (a = 0.05, naive log-rank test; Fig. 2). When distant failure as the first site of failure within 5 years was examined, 18 of 74 patients given adjuvant chemotherapy failed distantly (crude rate of 24%) compared with 27 of 192 patients not given chemotherapy (crude rate of 14%). The Fisher exact test (for differences in distribution of type of failure) was associated with a p-value of 0.0 1. group
I””
0 No Chemotheropy (192) Chemotheropy (74)
??
kl
80-
p = 0.05
3 3
60-
$ g
40
6 s
20
0
T/ME fyemsl Fig. 2. Actuarial risk of distant failure as the first site of recurrence for the 74 patients treated with adjuvant chemotherapy compared with the 192 patients who did not receive adjuvant chemotherapy (p = 0.05, naive log-rank test). The numbers in parentheses at 5 years indicate patients at risk.
October 1989. Volume 17, Number 4
We examined the risk of breast recurrence and the risk of distant failure for various subsets of the two groups (Table 2). In many cases, the number of patients in each subset was small. Nevertheless, the use of conservative surgery, radiotherapy, and adjuvant chemotherapy was associated with a low risk of local recurrence as a first site of failure in all subsets. Although only seven patients with positive axillary nodes did not receive adjuvant chemotherapy, at 5 years their crude rate of breast recurrence was 43% and their crude rate of distant failure was 14%, compared with 4% and 25%, respectively, for the 68 nodepositive patients who did receive chemotherapy (p = 0.0 1).
DISCUSSION In this study, premenopausal breast cancer patients treated with conservative surgery (CS), radiation therapy, and adjuvant chemotherapy had a very low risk of breast recurrence as the first site of failure, despite their overall poorer prognosis. When these patients were compared to patients treated with CS and radiotherapy alone, they had a lower rate of breast recurrence at 5 years even though factors predicting for local recurrence, such as age and an extensive intraductal component in the tumor (EIC), were equivalent in the two groups. This suggests a possible interaction between radiotherapy and chemotherapy in achieving local tumor control. The effect of combined modality treatment on local tumor control in node-positive patients has been examined in three previous studies. Note that the results given in these three studies apply to the combined outcome in postmenopausal and premenopausal patients, and no separate analysis by menopausal status is available. The results of NSABP trial B-06, like those of the present study, provide evidence that combined modality treatment is associated with a high level of local tumor control (8). In that trial, 1843 women with clinical Stage I or II breast cancer were randomized to receive either total mastectomy, segmental mastectomy alone, or segmental mastectomy followed by breast irradiation. All patients had axillary node dissections and patients with positive nodes received adjuvant chemotherapy consisting of melphalan and 5-fluorouracil. At 5 years, only 4% of the I92 patients with positive nodes treated with conservative surgery, radiation therapy, and adjuvant chemotherapy had a local recurrence as the first site of failure compared to 9% of the 378 node-negative patients treated without chemotherapy (10). Note that the 5-year percentages of breast recurrence as the first site of failure of 4% observed in the NSABP trial and 5% observed in our series are lower than the local recurrence rates for node-positive patients treated with mastectomy and chemotherapy, which range from 9 to 18% at 5 years (13.20). Similar results have also been observed in the most recent Milan study, in which a 3% local failure rate was seen in 127 patients after quadrantectomy, radiation, and chemotherapy compared with a 7% rate seen in 172 patients treated with mastectomy and
Chemotherapy radiotherapy effect in breast cancer 0 M. A.
715
ROSE et al.
Table 2. 5-year crude percentage of local and distant failure in premenopausal patients treated with conservative surgery and radiotherapy with and without adjuvant chemotherapy (CT) Adjuvant CT
No adjuvant CT % local failure
% distant failure
192
15
14
74
50 93 49
30 6 14
6 18 14
23 37 14
30 112 50 31 161
13 15 14 26 12
30 12 10 10 15
24 24 26 11 63
7
43
14
68
No. pts. All patients EIC present EIC absent EIC unevaluable Lymphatic invasion No lymphatic invasion Unknown lymphatic invasion Age < 35 Age 35-49 Nodes histoloticallv
nositive
* p-value refers to the comparison of the distribution EIC-= extensive intraductal component.
No. pts
% distant failure
p-value*
4
24
0.01
13 0 0
17 27 29
0.13 0.20 0.20
8 0 4 9 3
25 29 19 36 22
0.78 0.01 0.31 0.15 0.06
4
25
0.01
% local failure
of site of first failure (local, distant, neither) within 5 years (Fisher exact test)
chemotherapy (3). Finally, in a study from the Hospital of the University of Pennsylvania examining the risk of local failure in 552 breasts treated with CS and RT, 3% of node-positive patients treated with CS, RT, and CT failed locally versus 8% of patients who did not receive chemotherapy (p = NS) (23). One interpretation of these results is that radiation therapy and chemotherapy are interactive in improving local tumor control following CS. Note that an interaction between radiotherapy and adjuvant chemotherapy in improving local tumor control is best evaluated in a prospective clinical trial in which patients are randomized to conservative surgery and radiotherapy or to conservative surgery, radiotherapy, and adjuvant chemotherapy. We are unaware of any such study and, given the benefit of adjuvant chemotherapy observed among premenopausal patients treated with mastectomy, it is unlikely that such a study will now be performed. As a result, this issue can only be addressed by retrospective analysis, either by indirect comparisons within a randomized trial designed to address another issue (such as in the NSABP or Milan studies) or in a defined study population such as this. Also note that the method for assessing local tumor control is not straightforward. The most frequently used method has been Kaplan-Meier actuarial calculations, as used in both the NSABP B-06 study and the present study. Such calculations are, however, not strictly valid because of the problems of non-independent competing risks of local and distant failures. By convention, local recurrence is censored by distant failure. It is common practice not to attempt to detect or score a local recurrence in a patient with metastatic disease. If, in node-positive patients, the time to distant failure is simply shorter than the time to local failure, it would appear that these patients have a high level of local tumor control because local failure would occur after distant failure and would not be scored. Although a high distant failure rate may possibly ex-
plain the low local failure rate seen in patients who received combined modality treatment, we believe it is less likely than a true interaction between the modalities for two reasons. First, although only seven patients with pathologically positive nodes in our study population did not receive chemotherapy, three (43%) failed first locally within 5 years compared with only three of 68 patients (4%) with pathologically positive nodes treated with CS, RT, and CT (a = 0.01). Secondly, several studies have shown that node-positive patients have a greater, not lesser, risk of local recurrence as a first site of failure than node-negative patients when treated with CS and radiotherapy without chemotherapy (2, 17). Also of note in the NSABP B-06 study, among patients randomized to CS without radiation. node-positive patients treated with adjuvant chemotherapy had a greater risk of local recurrence than node-negative patients treated without chemotherapy (39% vs 29%). These results indicate that when patients with positive nodes are treated with CS and either radiotherapy or chemotherapy (but not both), there is a high rate of local recurrence as a first site of failure, suggesting that it is the combination of chemotherapy and radiation therapy which is critical in assuring local tumor control in these patients. The effectiveness of adjuvant chemotherapy in reducing the risk of local failure has also been evaluated among node-positive patients treated with mastectomy. However, it is important to note that a chest wall failure postmastectomy is not analogous to a breast failure after conservative surgery and radiotherapy. The results from several prospective randomized trials comparing patients who received chemotherapy after mastectomy to a control group that did not suggest that the use of chemotherapy is associated with a small reduction in the local recurrence rate. The lo-year results from the Milan trial showed an 11.6% crude locoregional first failure rate in 207 patients who received CMF compared with a 14.5% crude rate in
716
I. J. Radiation Oncology 0 Biology 0 Physics
179 patients who received no chemotherapy (4). The NSABP trial B-05, comparing patients treated with Lphenylalanine mustard (L-PAM) to a control group, found the cumulative incidence of locoregional recurrence as the site of first failure at 10 years to be 14% in 179 patients who received L-PAM and 24% in 170 patients who received a placebo (9). Five-year results of two separate trials from Manchester and Guy’s Hospital in Great Britain have been reported (13, 20). In the first, in which nodepositive patients received either melphalan or no treatment, first failure occurred in locoregional sites by 5 years in 18% of 187 patients who received chemotherapy compared with 27% of 183 controls (p = NS) (20). In the second trial, in which node-positive patients were randomized to CMF versus no treatment, the crude locoregional failure rate was 9% in the CMF group and 2 1% in the control group (p = 0.007) (13). Other trials with longterm follow-up, notably those from Switzerland (22). Japan ( 14) and the British West Midlands ( 16), have failed to confirm the reduction of local recurrences postmastectomy by chemotherapy. Overall, the results from these studies suggest that adjuvant chemotherapy by itself given after mastectomy has only a modest effect on the frequency of local recurrence as a first site of failure. This observation also supports the notion that it is the combination of chemotherapy and radiotherapy which is important for local tumor control. The mechanism by which adjuvant chemotherapy might improve local tumor control in patients treated with conservative surgery and radiation therapy is not clear. One explanation is that the cytoreductive effects of radiation and chemotherapy are simply additive in eliminating the remaining tumor cells in the breast. In addition, since certain chemotherapeutic agents, including 5-fluorouracil,
October 1989, Volume 17. Number 4
are known to be radiosensitizing agents, the use of adjuvant chemotherapy may be associated with enhanced tumor kill by radiation. One or both of these mechanisms may apply in this setting. An alternative hypothesis for the improvement in local tumor control is that adjuvant chemotherapy may alter the hormonal milieu in premenopausal patients. Recent results from our institution suggest that premenopausal patients are at greater risk of breast recurrence than postmenopausal patients (18). It is possible that alteration in the circulating estrogen levels caused by the use of chemotherapy may affect the likelihood of breast recurrence in premenopausal patients. The findings presented here have implications for the clinical management of premenopausal patients with positive nodes, for whom adjuvant chemotherapy is now considered standard therapy (6). Although randomized trials comparing mastectomy to conservative surgery and radiation therapy have demonstrated equivalent survival rates (8,24), some clinicians have favored the use of mastectomy in these patients because this operation is perceived as being more aggressive than conservative surgery and radiation therapy, and may interfere less with the use of adjuvant chemotherapy. However, several studies have suggested that the prior administration of radiation therapy does not significantly compromise the ability to deliver full-dose chemotherapy (3, 5, 7, 15). Our finding that the risk of local recurrence as the first site of failure in premenopausal patients treated with conservative surgery, radiation therapy, and adjuvant chemotherapy is low provides an additional reason to consider this treatment approach. Based on our results and the other currently available data, we believe that this combined modality approach is the preferred treatment option for premenopausal patients with positive axillary nodes.
REFERENCES 1. American
Joint Committee for Cancer Staging and EndResults Reporting. Manual for staging of cancer. 2nd edition, 1983. 2. Atkins, H.; Hayward, J. L.; Klugman, D. J.; Wayte, A. B. Treatment of early breast cancer: a report after ten years of a clinical trial. Br. Med. J. 2:423-429; 1972. 3. Bonadonna, G.; Valagussa, P.; Moliterni, A.; Buzzoni. R.; Zambetti, M.; Zucali, R.: Banfi, A.; Veronesi, U. Adjuvant chemotherapy following quadrantectomy plus radiotherapy or mastectomy (Abstr.). Proc. ASCO. 6:60; 1987. 4. Bonadonna, G.; Valagussa, P.; Rossi. A.; Tancini, G.; Brambella, C.; Zambetti, M.: Veronesi, U. Ten-year experience with CMF-based adjuvant chemotherapy in resectable breast cancer. Breast Cancer Res. Treat. 5:95-l 15; 1985. 5. Botnick, L. E.: Come, S.; Rose, C. M.; Goldstein, M.; Lange, R.; Tishler, S.; Schnipper, L. Primary breast irradiation and concomitant adjuvant chemotherapy. In: Harris. J. R., Hellman, S., Silen, W., eds. Conservative management of breast cancer. Philadelphia, PA: J. B. Lippincott; 1983:32 l328. 6. Consensus Development Conference Report. Adjuvant chemotherapy of breast cancer. JAMA. 254:346 l-3463: 1985.
7. Danoff, B. F.; Haller. D. G.; Glick, J. H.; Goodman, R. L. Conservative surgery and irradiation in the treatment of early breast cancer. Ann. Int. Med. 102:634-642; 1985. 8. Fisher, B.; Bauer, M.: Margolese, R.; Poisson, R.; Pilch, Y.; Redmond, C.: Fisher, E.; Wolmark, N.; Deutsch. M.; Montague, E.; Saffer, E.; Wickerham, L. Five-year results of a randomized clinical trial comparing total mastectomy and segmental mastectomy with or without radiation in the treatment of breast cancer. N. Engl. J. Med. 312:665-673; 1985. 9. Fisher, B.; Fisher, E. R.: Redmond, C. Ten year results from the NSABP clinical trial evaluating the use of L-phenylalanine mustard (L-PAM) in the management of primary breast cancer. J. Clin. Oncol. 4:929-941: 1986. 10. Fisher, B.; Wolmark, N. Conservative surgery: the American experience. Sem. Oncol. 13:425-433; 1986. 1 I. Fisher, R. A. Statistical methods for research workers. Edinburgh: Oliver and Boyd; 1934. 12. Gore, S. M.; Come, S. E.; Griem, K.; Rose, M. A.; Recht, A.; Botnick, L. E.; Rose, C.; Beadle, G.; Silver. B.; Henderson, I. C.; Harris, J. R. Influence of the sequencing of chemotherapy and radiation therapy in node-positive breast cancer patients treated by conservative surgery and radiation
Chemotherapy radiotherapy effect in breast cancer 0 M. A. ROSE et
13.
14.
15.
16.
17.
18.
therapy. In: Jones, S. E., Salmon, S. E., eds. Adjuvant therapy of cancer V. Orlando, FL: Grune and Stratton; 1987: 365-373. Howell, A.; George, W. D.; Crowther, D.; Rubens, R. D.; Bulbrook, R. D.; Bush, H.; Howat, J. M. T.; Sellwood, R. A.; Hayward, J. L.; Fentiman, I. S.; Chaudary, M. Controlled trial of adjuvant chemotherapy with cyclophosphamide, methotrexate and fluorouracil for breast cancer. Lancet II:307-311; 1984. Koyama, H.; Wada, T.; Takahashi, Y.; Nishizawa, Y.; Aoki, Y; Terasawa, T.; Kosaki, G.; Kajita, A.; Wada, A. Surgical adjuvant chemotherapy with mitomycin-C and cyclophosphamide in Japanese patients with breast cancer. Cancer 46:2373-2379; 1980. Lippman, M. E.: Edwards, B. K.; Findlay, P.; Danforth, D. W.; MacDonald, H.; D’Angelo, T.; Gorrell, C. Influence of definitive radiation therapy for primary breast cancer on ability to deliver adjuvant chemotherapy. NC1 Monog. I: 99-104; 1986. Morrison, J. M.; Howell, A.; Grieve, R. J. The West Midlands Oncology Association trials of adjuvant chemotherapy for operable breast cancer. In: Salmon, S. E., Jones, S. E., eds. Adjuvant therapy of cancer IV. Orlando, FL: Grune and Stratton; 1984:253-259. Osborne, M. P.; Ormison, N.; Harmer, C. L.; McKinna, A.; Baker, J.; Greening, W. P. Breast conservation in the treatment of early breast cancer. Cancer 53:349-355: 1984. Recht, A.; Connolly. J. L.; Schnitt, S. J.; Silver, B.; Rose, M. A.; Love, S.; Harris, J. R. The effect of young age on tumor recurrence in the treated breast after conservative
19.
20.
2 1.
22.
23.
24.
al.
717
surgery and radiotherapy. Int. J. Radiat. Oncol. Biol. Phys. 14(1):3-10; 1988. Recht, A.; Silver, B.; Schnitt, S.; Connolly, J.; Hellman, S.; Harris, J. R. Breast relapse following primary radiation therapy for early breast cancer. I. Classification, frequency and salvage. Int. J. Radiat. Oncol. Biol. Phys. 11:1271-1276; 1985. Rubens, R. D.; Knight, R. K.; Fentiman, I. S.; Howell, A.; Crowther, D.; George, W. D.; Hayward, J. L.; Bulbrook, R. D.; Chaudary, M.; Bush, H.; Sellwood, S. A.; Howat, J. M. T. Controlled trial of adjuvant chemotherapy with melphalan for breast cancer. Lancet 1:839-843; 1983. Schnitt. S. J.; Connolly, J. L.; Harris, J. R.; Hellman, S.; Cohen, R. B. Pathologic predictors of early local recurrence in stage I and II breast cancer treated by primary radiation therapy. Cancer 53: 1049- 1057; 1984. Senn, H. J.; Jung, W. F.; Amgwerd, R.; Hochuli, E.; Ammen, J.; Engelhert, G.; Heinz, C.; Wick, A.; Enderlin, F.; Creux, G. Adjuvant chemoimmunotherapy with LMF and BCG in node-negative and node-positive breast cancer: 8-year results. In: Senn, H., ed. Recent results in cancer research. Adjuvant chemotherapy of breast cancer. New York, NY: Springer-Verlag; 1984:90- 10 1. Solin, L. J.; Fowble, B.; Martz, K. L.; Goodman, R. L. Definitive irradiation for early stage breast cancer: the University of Pennsylvania experience. Int. J. Radiat. Oncol. Biol. Phys. 14:235-242; 1988. Veronesi, U.; Zucali, R.; Luini, A. Local control and survival in early breast cancer: the Milan trial. Int. J. Radiat. Oncol. Biol. Phys. 12:7 17-720; 1986.
17,pp.7I l-7
0360-3016/89 $3.00 + .OO Copyright 0 1989 Pergamon Pres plc
I7
??Original Contribution
PREMENOPAUSAL BREAST CANCER PATIENTS TREATED WITH CONSERVATIVE SURGERY, RADIOTHERAPY +ND ADJUVANT CHEMOTHERAPY HAVE A LOW RISK OF LOCAL FAILURE MARY ANN ROSE,M.D.,’ I. CRAIG HENDERSON,M.D.,2 REBECCAGELMAN, PH.D.,~ JOHN BOYAGES,M.D.,’ STACEYM. GORE, M.D.,4 STEVENCOME, M.D.,4 BARBARASILVER,B.A.,’ ABRAM RECHT,M.D.,’ JAMESL. CONNOLLY, M.D.,5 STUARTJ. SCHNITT,M.D.,5 C. NORMAN COLEMAN,M.D.’ AND JAY R. HARRIS,M.D.’ ‘Joint Center for Radiation Therapy and Department of Radiation Therapy, Harvard Medical School; *Division of Medical Oncology, Dana-Farber Cancer Institute and Harvard Medical School; 3Division of Biostatistics and Epidemiology, Dana-Farber Cancer Institute and Harvard School of Public Health; 4Division of Hematology-Oncology, Department of Medicine, Beth Israel Hospital and Harvard Medical School; and ‘Department of Pathology, Beth Israel Hospital and Harvard Medical School, Boston, MA The use of adjuvant chemotherapy in premenopausal breast cancer patients with positive nodes is now routine, but the optimal local treatment of these patients is uncertain. To determine the effect of adjuvant chemotherapy on the likelihood of local recurrence as the first site of failure in premenopausal patients treated with conservative surgery (CS) and radiotherapy (RT), we examined the outcome of 74 patients treated with CS, RT, and adjuvant chemotherapy and compared it to the outcome in 192 patients treated with CS and RT alone. Adjuvant chemotherapy consisted of four or more cycles of either a doxorubicin-containing regimen or cyclophosphamide, methotrexate, and 5fluorouracil. All patients were less than 50 years old, had UICC-AJCC Stage I or II breast cancer treated between 1968 and 1981, had gross excision of the primary tumor, and had a total radiation dose to the primary tumor bed of 26000 cCy. Factors predicting for local recurrence, such as extensive intraductal carcinoma and age less than 35, were equivalent in the two groups. Women treated with adjuvantchemotherapyhad significantlyworse T- and N-stages than women treated with conservative surgery and radiotherapy alone: 61% versus 36% had T2 tumors (p = 0.0003), 34% versus 6% had clinically positive nodes (p < O.OOOl),and 97% versus 4% had pathologically positive nodes (p < 0.0001). Despite the poorer prognosis of patients treated with adjuvant chemotherapy, within 5 years of diagnosis, 4% of patients who received adjuvant chemotherapy had their initial relapse in the breast and 24% had initial failure elsewhere, compared with 15% local failure first and 14% failure elsewhere first for those treated without chemotherapy (p = 0.01). We conclude that premenopausal patients with positive nodes treated with combined modality therapy (conservative surgery, radiation therapy, and adjuvant chemotherapy) have a low risk of local recurrence as a first site of failure. These results suggest a possible interaction between radiation therapy and chemotherapy in their effects on local tumor control. Breast cancer, Radiotherapy,
Local recurrence, Adjuvant chemotherapy.
INTRODUCTION
found to have pathologically positive lymph nodes are
receiving adjuvant chemotherapy. Thus, premenopausal patients with positive nodes are being considered for treatment with conservative surgery, radiation therapy, and adjuvant chemotherapy. This approach has raised several issues including the influence of chemotherapy on the cosmetic outcome, the possibility of drug dose reductions necessitated by radiation therapy, and the optimal sequencing of these two modalities. Given these concerns and a perception that mastectomy may provide better local tumor control, some clinicians have recommended mastectomy in premenopausal patients with positive
Presented in part at the Annual Meeting of the American Society for Therapeutic Radiology and Oncology, New Orleans, October 1988.
Reprint requests to: Mary Ann Rose, M.D., Joint Center for Radiation Therapy, 50 Binney Street, Boston, MA 02 115. Accepted for publication 13 April 1989.
In recent years, two major trends have evolved in the management of patients with early stage breast cancer. First, increasing numbers of women are being treated with conservative surgery (CS) and radiation therapy as an alternative to mastectomy. This is based on the results of prospective randomized trials comparing CS and radiation with mastectomy which have demonstrated equivalent rates of survival (8, 24). Second, premenopausal patients
711
712
I. J. Radiation
Oncology
0 Biology 0 Physics
nodes. However, the results from the National Surgical Adjuvant Breast Project (NSABP) B-06 trial suggest that the use of CS, radiotherapy, and adjuvant chemotherapy provides a high level of local tumor control in node-positive patients ( 10). To study this issue further, we examined the likelihood of breast recurrence among node-positive premenopausal women treated with CS, radiotherapy, and adjuvant chemotherapy, and compared it to the outcome in premenopausal patients treated with CS and radiation therapy alone. Our results indicate that premenopausal patients with positive nodes treated with CS, radiotherapy, and adjuvant chemotherapy have a low risk of local recurrence as a first site of failure, suggesting an interaction between radiotherapy and chemotherapy in achieving local tumor control.
METHODS
AND MATERIALS
Three hundred and thirty-two premenopausal women with UICC-AJCC clinical Stage I or II breast cancer were treated at the Joint Center for Radiation Therapy (JCRT) between July 1968 and December 1981 (1). (Note that staging was done according to the system adopted in 1983.) For the purpose of this analysis, premenopausal was defined as age less than 50 years. To eliminate the confounding effect of possibly inadequate treatment, seven patients who had less than excisional biopsy (i.e., needle biopsy or incisional biopsy only) prior to radiotherapy and 46 patients who received a total dose of less than 6000 cGy to the regional of the primary tumor were excluded from the analysis. In addition, 12 patients who received either single agent chemotherapy or regimens which did not include cyclophosphamide, methotrexate, and 5-fluorouracil (CMF) or doxorubicin (Adriamycin), and one patient who received only two cycles of CMF were excluded from the analysis. The study population thus contained 266 patients. Patient accrual was not uniform throughout the study period; 1 1 l/ 192 patients (58%) who did not receive adjuvant chemotherapy and 41174 patients (55%) who did receive adjuvant chemotherapy were treated in the last 2 years of the study (1980-8 1). Follow-up was obtained as of April 1987 with a median follow-up of 75 months (range, 37-l 50 months). All patients were treated with conservative surgery and radiation therapy. Surgery consisted of an excisional biopsy, defined as gross total removal of the primary tumor without histologic confirmation of negative margins. In addition, 239 patients (90%) also underwent axillary nodal dissection which was usually confined to Level I and II nodes. Details of the radiation treatment have been described elsewhere ( 19). All patients received whole breast irradiation, typically to a total dose of 4500-5000 cGy over 5 weeks. A supplemental “boost” dose to the primary tumor site was given to all patients. This was accomplished through Iridium- 192 implantation in 254 patients and by electron beam or photon irradiation in 12 patients. The
October
1989. Volume
17, Number
4
total dose to the region of the primary tumor ranged from 6000 to 7800 cGy (median, 6800 cGy). Radiation doses were comparable for the chemotherapy and nonchemotherapy groups: the ranges were 6000-7250 (median, 6800) and 6000-7800 (median, 6780) respectively. Fortyfive patients received radiation to the breast only. The remaining patients received radiation to the breast and to draining regional lymph nodes. Seventy-four patients received a minimum of four cycles of adjuvant CMF or an Adriamycin-containing regimen. Patients received their chemotherapy either sequentially, with chemotherapy commencing after completion of radiation (41 patients); concurrently, with chemotherapy commencing at the same time as radiation (24 patients); or in a “sandwich” fashion, with at least one cycle of chemotherapy administered prior to radiation (nine patients). The details of the chemotherapy administered in these patients have been previously described ( 12). Only one patient received adjuvant hormonal therapy (Tamoxifen) in addition to chemotherapy. Clinico-pathologic features examined in this study included primary tumor size, clinical nodal status, pathologic nodal status, lymphatic invasion, and the presence and extent of intraductal carcinoma within the infiltrating primary tumor and in adjacent grossly-normal breast tissue. We have previously shown that the most significant predictor of breast recurrence among our patients is an “extensive intraductal component” (EIC) in the primary surgical specimen (2 1). This factor is defined as the presence of intraductal carcinoma both within the primary tumor (comprising at least 25% of the tumor area) and in adjacent grossly-normal breast tissue. Microscopic evaluation of margins was not available in most patients. Local recurrence was defined as the detection of cancer in the treated breast tissue or overlying skin occurring before the development of distant metastases. Patients were censored from the calculation of breast recurrence at the time of last follow-up or the discovery of regional or metastatic disease. Among the 74 patients who received chemotherapy, one was noted to have a recurrence in the skin overlying the treated breast 14 months after distant failure. Among the 192 patients who did not receive chemotherapy, one patient was noted to have a skin recurrence 1 I months after distant failure. These failures are not counted in the analysis. One patient who received chemotherapy had an infraclavicular node failure and three patients who did not receive chemotherapy had axillary nodal failures. These are counted as distant failures for the purpose of analysis. Local failures were classified by location in the breast relative to the primary tumor site as true recurrences (TR, within the area of the boost), marginal misses (MM, adjacent to the boost area), recurrences elsewhere (E, in other quadrants of the breast), and skin recurrences (S). The risk of local recurrence was evaluated by the use of actuarial calculations. Actuarial curves were calculated by the Kaplan-Meier method and comparisons between
Chemotherapy radiotherapy effect in breast cancer 0 M. A. ROSE et al.
curves were made using the two-tailed “naive” log-rank test. The Kaplan-Meier estimates, however, can only be shown to be unbiased and maximum likelihood estimates if time to distant and time to local failure are statistically independent. This assumption can be questioned in this data set. Such non-independent censoring could cause the log-rank test to be biased in either direction (too significant or not significant enough). To minimize the problem of competing risks, distant recurrence was defined as the detection of cancer outside of the treated breast or death from intercurrent disease (not breast cancer) before the development of local recurrence. Thus three patients, one who received chemotherapy and two who did not, died without evidence of recurrence and were counted as distant failures. Patients were censored from the calculation of distant failure at the time of last follow-up or at the time of local failure. Three patients (one who received chemotherapy and two who did not) had simultaneous local and distant failure. These patients with simultaneous failures were counted as distant failures because distant failure determined their subsequent therapy. The distribution of the site of first failure (local, distant, none) within 5 years of treatment was compared for the chemotherapy and nonchemotherapy groups using the multinomial test (Fisher exact test) (11). The multinomial test allows an unbiased test of whether the distribution of causes of failure (distant or local) differed in the two treatment groups. Differences in patient characteristics between the two groups were analyzed using the Fisher exact test and the Wilcoxon test (for age, follow-up, and radiation doses). Two-sided p-values of 0.05 or less were considered statistically significant.
RESULTS A comparison of characteristics between the 74 premenopausal patients who were treated with CS, radiotherapy. and adjuvant chemotherapy (combined modality treatment) and the 192 premenopausal patients who were
Table 1. Patient
treated with CS and radiotherapy alone is presented in Table 1. The age distribution and follow-up distribution were comparable in the two groups. Patients who received combined modality treatment generally presented in a more advanced stage than patients who did not receive chemotherapy. The percentage of T2 tumors in the group receiving adjuvant chemotherapy was significantly higher than in the group not receiving drugs: 6 1% versus 36%, p = 0.0003. Only 66% of patients who received chemotherapy were clinically staged NO/N 1A,compared to 94% of patients who did not receive drugs (p < 0.000 1). Similar percentages of patients in each group underwent axillary nodal dissection. However, histologically confirmed lymph node metastases were present in 97% of patients who received chemotherapy compared to only 4% of patients who did not receive chemotherapy (p < 0.0001). The percentage of patients with an extensive intraductal component (among evaluable patients with infiltrating ductal histology) was similar for the group that received chemotherapy and for those who did not: 38% versus 35%, p = NS. The percentage of patients with respectively, lymphatic invasion was higher (36%) among patients who received chemotherapy than among patients who did not receive chemotherapy (17%) (p = 0.003). Overall, patients treated with conservative surgery, radiotherapy, and adjuvant chemotherapy had significantly worse clinical and pathologic stages than patients treated with conservative surgery and radiotherapy alone. Figure 1 shows the actuarial calculations of time to breast failure as the first site of recurrence for the 74 patients treated with CS, radiotherapy, and adjuvant chemotherapy compared to the 192 patients who were treated with CS and radiotherapy alone. The 5-year actuarial rate of local failure was 5% for patients who received chemotherapy and 17% for those who did not receive chemotherapy (p = 0.03, naive log-rank test). Within 5 years of diagnosis, three of 74 patients given chemotherapy failed first locally (crude rate of 4%) compared with 28 of 192 patients not given chemotherapy (crude rate of 15%). The local failure types for the chemotherapy group were one TR, one MM, and one S. The failure types for the
characteristics
CS+RT No. pts. % Tl % NO/IA % axillary dissection % nodes pathologically positive % with lymphatic invasion % extensive intraductal component Median age (range) % age < 35 Median follow-up (range) Median cGY to breast (range) CS = Conservative
surgery; RT = Radiation
CS+RT+CT 74 39 66 95 97 36 38 40 (29-49) 15 76 mos (37-121) 6800 (6000-7250)
192 64 94 88 4 17 :: (25-49) 16 75 mos (49-150) 6780 (6000-7800) therapy;
713
CT = Adjuvant
chemotherapy.
p-value
0.0003
I. J. Radiation Oncology 0 Biology 0 Physics
0 No Chemotherapy (192) ?? Chemotherapy (74)
T/ME fyeofs/ Fig. 1. Actuarial risk of breast failure as the first site of recurrence for the 74 patients treated with adjuvant chemotherapy compared with the 192 patients who did not receive adjuvant chemotherapy (p = 0.03, naive log-rank test). The numbers in parentheses indicate patients at risk at 5 years.
that did not receive chemotherapy were 15 TRs, seven MMs, five Es, and one S. The sequencing of radiation and chemotherapy did not appear to have a significant effect on local recurrence, but the number of failures in the group that received chemotherapy was quite small. The 5-year actuarial rate of distant failure was 25% for patients who received chemotherapy compared to 14% for those who did not receive chemotherapy (a = 0.05, naive log-rank test; Fig. 2). When distant failure as the first site of failure within 5 years was examined, 18 of 74 patients given adjuvant chemotherapy failed distantly (crude rate of 24%) compared with 27 of 192 patients not given chemotherapy (crude rate of 14%). The Fisher exact test (for differences in distribution of type of failure) was associated with a p-value of 0.0 1. group
I””
0 No Chemotheropy (192) Chemotheropy (74)
??
kl
80-
p = 0.05
3 3
60-
$ g
40
6 s
20
0
T/ME fyemsl Fig. 2. Actuarial risk of distant failure as the first site of recurrence for the 74 patients treated with adjuvant chemotherapy compared with the 192 patients who did not receive adjuvant chemotherapy (p = 0.05, naive log-rank test). The numbers in parentheses at 5 years indicate patients at risk.
October 1989. Volume 17, Number 4
We examined the risk of breast recurrence and the risk of distant failure for various subsets of the two groups (Table 2). In many cases, the number of patients in each subset was small. Nevertheless, the use of conservative surgery, radiotherapy, and adjuvant chemotherapy was associated with a low risk of local recurrence as a first site of failure in all subsets. Although only seven patients with positive axillary nodes did not receive adjuvant chemotherapy, at 5 years their crude rate of breast recurrence was 43% and their crude rate of distant failure was 14%, compared with 4% and 25%, respectively, for the 68 nodepositive patients who did receive chemotherapy (p = 0.0 1).
DISCUSSION In this study, premenopausal breast cancer patients treated with conservative surgery (CS), radiation therapy, and adjuvant chemotherapy had a very low risk of breast recurrence as the first site of failure, despite their overall poorer prognosis. When these patients were compared to patients treated with CS and radiotherapy alone, they had a lower rate of breast recurrence at 5 years even though factors predicting for local recurrence, such as age and an extensive intraductal component in the tumor (EIC), were equivalent in the two groups. This suggests a possible interaction between radiotherapy and chemotherapy in achieving local tumor control. The effect of combined modality treatment on local tumor control in node-positive patients has been examined in three previous studies. Note that the results given in these three studies apply to the combined outcome in postmenopausal and premenopausal patients, and no separate analysis by menopausal status is available. The results of NSABP trial B-06, like those of the present study, provide evidence that combined modality treatment is associated with a high level of local tumor control (8). In that trial, 1843 women with clinical Stage I or II breast cancer were randomized to receive either total mastectomy, segmental mastectomy alone, or segmental mastectomy followed by breast irradiation. All patients had axillary node dissections and patients with positive nodes received adjuvant chemotherapy consisting of melphalan and 5-fluorouracil. At 5 years, only 4% of the I92 patients with positive nodes treated with conservative surgery, radiation therapy, and adjuvant chemotherapy had a local recurrence as the first site of failure compared to 9% of the 378 node-negative patients treated without chemotherapy (10). Note that the 5-year percentages of breast recurrence as the first site of failure of 4% observed in the NSABP trial and 5% observed in our series are lower than the local recurrence rates for node-positive patients treated with mastectomy and chemotherapy, which range from 9 to 18% at 5 years (13.20). Similar results have also been observed in the most recent Milan study, in which a 3% local failure rate was seen in 127 patients after quadrantectomy, radiation, and chemotherapy compared with a 7% rate seen in 172 patients treated with mastectomy and
Chemotherapy radiotherapy effect in breast cancer 0 M. A.
715
ROSE et al.
Table 2. 5-year crude percentage of local and distant failure in premenopausal patients treated with conservative surgery and radiotherapy with and without adjuvant chemotherapy (CT) Adjuvant CT
No adjuvant CT % local failure
% distant failure
192
15
14
74
50 93 49
30 6 14
6 18 14
23 37 14
30 112 50 31 161
13 15 14 26 12
30 12 10 10 15
24 24 26 11 63
7
43
14
68
No. pts. All patients EIC present EIC absent EIC unevaluable Lymphatic invasion No lymphatic invasion Unknown lymphatic invasion Age < 35 Age 35-49 Nodes histoloticallv
nositive
* p-value refers to the comparison of the distribution EIC-= extensive intraductal component.
No. pts
% distant failure
p-value*
4
24
0.01
13 0 0
17 27 29
0.13 0.20 0.20
8 0 4 9 3
25 29 19 36 22
0.78 0.01 0.31 0.15 0.06
4
25
0.01
% local failure
of site of first failure (local, distant, neither) within 5 years (Fisher exact test)
chemotherapy (3). Finally, in a study from the Hospital of the University of Pennsylvania examining the risk of local failure in 552 breasts treated with CS and RT, 3% of node-positive patients treated with CS, RT, and CT failed locally versus 8% of patients who did not receive chemotherapy (p = NS) (23). One interpretation of these results is that radiation therapy and chemotherapy are interactive in improving local tumor control following CS. Note that an interaction between radiotherapy and adjuvant chemotherapy in improving local tumor control is best evaluated in a prospective clinical trial in which patients are randomized to conservative surgery and radiotherapy or to conservative surgery, radiotherapy, and adjuvant chemotherapy. We are unaware of any such study and, given the benefit of adjuvant chemotherapy observed among premenopausal patients treated with mastectomy, it is unlikely that such a study will now be performed. As a result, this issue can only be addressed by retrospective analysis, either by indirect comparisons within a randomized trial designed to address another issue (such as in the NSABP or Milan studies) or in a defined study population such as this. Also note that the method for assessing local tumor control is not straightforward. The most frequently used method has been Kaplan-Meier actuarial calculations, as used in both the NSABP B-06 study and the present study. Such calculations are, however, not strictly valid because of the problems of non-independent competing risks of local and distant failures. By convention, local recurrence is censored by distant failure. It is common practice not to attempt to detect or score a local recurrence in a patient with metastatic disease. If, in node-positive patients, the time to distant failure is simply shorter than the time to local failure, it would appear that these patients have a high level of local tumor control because local failure would occur after distant failure and would not be scored. Although a high distant failure rate may possibly ex-
plain the low local failure rate seen in patients who received combined modality treatment, we believe it is less likely than a true interaction between the modalities for two reasons. First, although only seven patients with pathologically positive nodes in our study population did not receive chemotherapy, three (43%) failed first locally within 5 years compared with only three of 68 patients (4%) with pathologically positive nodes treated with CS, RT, and CT (a = 0.01). Secondly, several studies have shown that node-positive patients have a greater, not lesser, risk of local recurrence as a first site of failure than node-negative patients when treated with CS and radiotherapy without chemotherapy (2, 17). Also of note in the NSABP B-06 study, among patients randomized to CS without radiation. node-positive patients treated with adjuvant chemotherapy had a greater risk of local recurrence than node-negative patients treated without chemotherapy (39% vs 29%). These results indicate that when patients with positive nodes are treated with CS and either radiotherapy or chemotherapy (but not both), there is a high rate of local recurrence as a first site of failure, suggesting that it is the combination of chemotherapy and radiation therapy which is critical in assuring local tumor control in these patients. The effectiveness of adjuvant chemotherapy in reducing the risk of local failure has also been evaluated among node-positive patients treated with mastectomy. However, it is important to note that a chest wall failure postmastectomy is not analogous to a breast failure after conservative surgery and radiotherapy. The results from several prospective randomized trials comparing patients who received chemotherapy after mastectomy to a control group that did not suggest that the use of chemotherapy is associated with a small reduction in the local recurrence rate. The lo-year results from the Milan trial showed an 11.6% crude locoregional first failure rate in 207 patients who received CMF compared with a 14.5% crude rate in
716
I. J. Radiation Oncology 0 Biology 0 Physics
179 patients who received no chemotherapy (4). The NSABP trial B-05, comparing patients treated with Lphenylalanine mustard (L-PAM) to a control group, found the cumulative incidence of locoregional recurrence as the site of first failure at 10 years to be 14% in 179 patients who received L-PAM and 24% in 170 patients who received a placebo (9). Five-year results of two separate trials from Manchester and Guy’s Hospital in Great Britain have been reported (13, 20). In the first, in which nodepositive patients received either melphalan or no treatment, first failure occurred in locoregional sites by 5 years in 18% of 187 patients who received chemotherapy compared with 27% of 183 controls (p = NS) (20). In the second trial, in which node-positive patients were randomized to CMF versus no treatment, the crude locoregional failure rate was 9% in the CMF group and 2 1% in the control group (p = 0.007) (13). Other trials with longterm follow-up, notably those from Switzerland (22). Japan ( 14) and the British West Midlands ( 16), have failed to confirm the reduction of local recurrences postmastectomy by chemotherapy. Overall, the results from these studies suggest that adjuvant chemotherapy by itself given after mastectomy has only a modest effect on the frequency of local recurrence as a first site of failure. This observation also supports the notion that it is the combination of chemotherapy and radiotherapy which is important for local tumor control. The mechanism by which adjuvant chemotherapy might improve local tumor control in patients treated with conservative surgery and radiation therapy is not clear. One explanation is that the cytoreductive effects of radiation and chemotherapy are simply additive in eliminating the remaining tumor cells in the breast. In addition, since certain chemotherapeutic agents, including 5-fluorouracil,
October 1989, Volume 17. Number 4
are known to be radiosensitizing agents, the use of adjuvant chemotherapy may be associated with enhanced tumor kill by radiation. One or both of these mechanisms may apply in this setting. An alternative hypothesis for the improvement in local tumor control is that adjuvant chemotherapy may alter the hormonal milieu in premenopausal patients. Recent results from our institution suggest that premenopausal patients are at greater risk of breast recurrence than postmenopausal patients (18). It is possible that alteration in the circulating estrogen levels caused by the use of chemotherapy may affect the likelihood of breast recurrence in premenopausal patients. The findings presented here have implications for the clinical management of premenopausal patients with positive nodes, for whom adjuvant chemotherapy is now considered standard therapy (6). Although randomized trials comparing mastectomy to conservative surgery and radiation therapy have demonstrated equivalent survival rates (8,24), some clinicians have favored the use of mastectomy in these patients because this operation is perceived as being more aggressive than conservative surgery and radiation therapy, and may interfere less with the use of adjuvant chemotherapy. However, several studies have suggested that the prior administration of radiation therapy does not significantly compromise the ability to deliver full-dose chemotherapy (3, 5, 7, 15). Our finding that the risk of local recurrence as the first site of failure in premenopausal patients treated with conservative surgery, radiation therapy, and adjuvant chemotherapy is low provides an additional reason to consider this treatment approach. Based on our results and the other currently available data, we believe that this combined modality approach is the preferred treatment option for premenopausal patients with positive axillary nodes.
REFERENCES 1. American
Joint Committee for Cancer Staging and EndResults Reporting. Manual for staging of cancer. 2nd edition, 1983. 2. Atkins, H.; Hayward, J. L.; Klugman, D. J.; Wayte, A. B. Treatment of early breast cancer: a report after ten years of a clinical trial. Br. Med. J. 2:423-429; 1972. 3. Bonadonna, G.; Valagussa, P.; Moliterni, A.; Buzzoni. R.; Zambetti, M.; Zucali, R.: Banfi, A.; Veronesi, U. Adjuvant chemotherapy following quadrantectomy plus radiotherapy or mastectomy (Abstr.). Proc. ASCO. 6:60; 1987. 4. Bonadonna, G.; Valagussa, P.; Rossi. A.; Tancini, G.; Brambella, C.; Zambetti, M.: Veronesi, U. Ten-year experience with CMF-based adjuvant chemotherapy in resectable breast cancer. Breast Cancer Res. Treat. 5:95-l 15; 1985. 5. Botnick, L. E.: Come, S.; Rose, C. M.; Goldstein, M.; Lange, R.; Tishler, S.; Schnipper, L. Primary breast irradiation and concomitant adjuvant chemotherapy. In: Harris. J. R., Hellman, S., Silen, W., eds. Conservative management of breast cancer. Philadelphia, PA: J. B. Lippincott; 1983:32 l328. 6. Consensus Development Conference Report. Adjuvant chemotherapy of breast cancer. JAMA. 254:346 l-3463: 1985.
7. Danoff, B. F.; Haller. D. G.; Glick, J. H.; Goodman, R. L. Conservative surgery and irradiation in the treatment of early breast cancer. Ann. Int. Med. 102:634-642; 1985. 8. Fisher, B.; Bauer, M.: Margolese, R.; Poisson, R.; Pilch, Y.; Redmond, C.: Fisher, E.; Wolmark, N.; Deutsch. M.; Montague, E.; Saffer, E.; Wickerham, L. Five-year results of a randomized clinical trial comparing total mastectomy and segmental mastectomy with or without radiation in the treatment of breast cancer. N. Engl. J. Med. 312:665-673; 1985. 9. Fisher, B.; Fisher, E. R.: Redmond, C. Ten year results from the NSABP clinical trial evaluating the use of L-phenylalanine mustard (L-PAM) in the management of primary breast cancer. J. Clin. Oncol. 4:929-941: 1986. 10. Fisher, B.; Wolmark, N. Conservative surgery: the American experience. Sem. Oncol. 13:425-433; 1986. 1 I. Fisher, R. A. Statistical methods for research workers. Edinburgh: Oliver and Boyd; 1934. 12. Gore, S. M.; Come, S. E.; Griem, K.; Rose, M. A.; Recht, A.; Botnick, L. E.; Rose, C.; Beadle, G.; Silver. B.; Henderson, I. C.; Harris, J. R. Influence of the sequencing of chemotherapy and radiation therapy in node-positive breast cancer patients treated by conservative surgery and radiation
Chemotherapy radiotherapy effect in breast cancer 0 M. A. ROSE et
13.
14.
15.
16.
17.
18.
therapy. In: Jones, S. E., Salmon, S. E., eds. Adjuvant therapy of cancer V. Orlando, FL: Grune and Stratton; 1987: 365-373. Howell, A.; George, W. D.; Crowther, D.; Rubens, R. D.; Bulbrook, R. D.; Bush, H.; Howat, J. M. T.; Sellwood, R. A.; Hayward, J. L.; Fentiman, I. S.; Chaudary, M. Controlled trial of adjuvant chemotherapy with cyclophosphamide, methotrexate and fluorouracil for breast cancer. Lancet II:307-311; 1984. Koyama, H.; Wada, T.; Takahashi, Y.; Nishizawa, Y.; Aoki, Y; Terasawa, T.; Kosaki, G.; Kajita, A.; Wada, A. Surgical adjuvant chemotherapy with mitomycin-C and cyclophosphamide in Japanese patients with breast cancer. Cancer 46:2373-2379; 1980. Lippman, M. E.: Edwards, B. K.; Findlay, P.; Danforth, D. W.; MacDonald, H.; D’Angelo, T.; Gorrell, C. Influence of definitive radiation therapy for primary breast cancer on ability to deliver adjuvant chemotherapy. NC1 Monog. I: 99-104; 1986. Morrison, J. M.; Howell, A.; Grieve, R. J. The West Midlands Oncology Association trials of adjuvant chemotherapy for operable breast cancer. In: Salmon, S. E., Jones, S. E., eds. Adjuvant therapy of cancer IV. Orlando, FL: Grune and Stratton; 1984:253-259. Osborne, M. P.; Ormison, N.; Harmer, C. L.; McKinna, A.; Baker, J.; Greening, W. P. Breast conservation in the treatment of early breast cancer. Cancer 53:349-355: 1984. Recht, A.; Connolly. J. L.; Schnitt, S. J.; Silver, B.; Rose, M. A.; Love, S.; Harris, J. R. The effect of young age on tumor recurrence in the treated breast after conservative
19.
20.
2 1.
22.
23.
24.
al.
717
surgery and radiotherapy. Int. J. Radiat. Oncol. Biol. Phys. 14(1):3-10; 1988. Recht, A.; Silver, B.; Schnitt, S.; Connolly, J.; Hellman, S.; Harris, J. R. Breast relapse following primary radiation therapy for early breast cancer. I. Classification, frequency and salvage. Int. J. Radiat. Oncol. Biol. Phys. 11:1271-1276; 1985. Rubens, R. D.; Knight, R. K.; Fentiman, I. S.; Howell, A.; Crowther, D.; George, W. D.; Hayward, J. L.; Bulbrook, R. D.; Chaudary, M.; Bush, H.; Sellwood, S. A.; Howat, J. M. T. Controlled trial of adjuvant chemotherapy with melphalan for breast cancer. Lancet 1:839-843; 1983. Schnitt. S. J.; Connolly, J. L.; Harris, J. R.; Hellman, S.; Cohen, R. B. Pathologic predictors of early local recurrence in stage I and II breast cancer treated by primary radiation therapy. Cancer 53: 1049- 1057; 1984. Senn, H. J.; Jung, W. F.; Amgwerd, R.; Hochuli, E.; Ammen, J.; Engelhert, G.; Heinz, C.; Wick, A.; Enderlin, F.; Creux, G. Adjuvant chemoimmunotherapy with LMF and BCG in node-negative and node-positive breast cancer: 8-year results. In: Senn, H., ed. Recent results in cancer research. Adjuvant chemotherapy of breast cancer. New York, NY: Springer-Verlag; 1984:90- 10 1. Solin, L. J.; Fowble, B.; Martz, K. L.; Goodman, R. L. Definitive irradiation for early stage breast cancer: the University of Pennsylvania experience. Int. J. Radiat. Oncol. Biol. Phys. 14:235-242; 1988. Veronesi, U.; Zucali, R.; Luini, A. Local control and survival in early breast cancer: the Milan trial. Int. J. Radiat. Oncol. Biol. Phys. 12:7 17-720; 1986.