management of stage iii primary breast cancer with primary chemotherapy, surgery, and radiation...

Management of Stage Ill Primary Breast Cancer With Primary Chemotherapy, Surgery, and Radiation Therapy

GABRIEL N. HORTOBAGYI, MD, F. C. AMES, MD,t A. U. BUZDAR, MD,' S. W. KAU, RN,' M. D. McNEESE, MD,* D. PAULUS, MD,§ V. HUG, MD,' F. A. HOLMES. MD,' M. M. ROMSDAHL, MD,t

G. FRASCHINI, MD,' C. M. MCBRIDE, MD,t R. G. MARTIN, MD,t AND E. MONTAGUE, MD*

One hundred seventy-four evaluable patients with noninflammatory Stage I11 (both operable and inoperable) breast cancer were treated with a combined modality strategy between 1974 and 1985. All patients received combination chemotherapy with 5-fluorouracil, Adriamycin (doxorubicin), and cyclophosphamide (FAC) as their initial form of therapy. After three cycles of chemotherapy, local treatment in the form of a total mastectomy with axillary dissection, or radiotherapy, or both, was completed. Subsequently, adjuvant chemotherapy was continued. There were 48 patients with Stage IIIA, and 126 patients with Stage IIIB disease. A complete remission was achieved in 16.7% of the patients, and 70.7% achieved a partial remission after the initial three cycles of FAC. The complete response rate was higher for patients with Stage IIIA, than for patients with Stage IIIB disease. All but six of the 174 patients treated were rendered disease- free after induction chemotherapy and local treatment. The median follow-up of this group of patients is 59 months. The 5-year disease-free survival rates were 84% for patients with Stage IIIA, and 33% for patients with Stage IIIB disease. The 5-year survival rate for, patients with Stage IIIA was 84%, and for patients with Stage IIIB 44%. At 10 years, 56% of patients with Stage IIIA and 26% of patients with Stage IIIB disease are projected to be alive. Younger patients, and those with estrogen receptor-positive tumors, had a trend for better survival than older patients and those with estrogen receptor-negative tumors. The quality of response to induction chemotherapy correlated prominently with prognosis, as did compliance with treatment. Twenty-six patients (15.3%) had locoregional recurrence. This multidisciplinary approach to locally advanced breast cancer rendered most patients disease-free and produced an excellent local control rate. Modifications of this treatment strategy may result in further improvement of survival rates.

Cancer 62:2507-2516, 1988.

TAGE 111 BREAST CANCER, as defined by the latest S Tumor, Nodes and Metastases (TNM) Classifica- tion,' includes patients with locally advanced primary breast cancer with various degrees of local and regional extension, and markedly different prognoses after local therapies. It is subdivided into Stage IIIA and Stage IIIB. The former comprises patients with a large primary tumor, or matted, or fixed axillary lymph nodes, but without involvement of the skin or fixation to the chest wall. This

Presented in part at the 23rd Annual Meeting of the American Society of Clinical Oncology, May 17-19, 1987, Atlanta, Georgia.

From the Departments of *Medical Oncology, ?General Surgery, $Radiotherapy, and §Diagnostic Radiology, The University of Texas M. D. Anderson Hospital and Tumor Institute at Houston, Texas.

The authors thank Rose Mary Zuniga for assistance in the preparation of the manuscript and Diane F. Bush for editorial review.

Address for reprints: Gabriel N. Hortobagyi, MD, Department of Medical Oncology, The University of Texas M. D. Anderson Hospital and Tumor Institute at Houston, 1515 Holcombe Boulevard, Houston, TX 77030.

Accepted for publication June 22, 1988.

subgroup of patients is technically considered operable. Stage IIIB includes all those patients whose tumors directly involve the skin of the breast, or whose tumors are fixed to the chest wall, or those patients who present with supraclavicular or subclavicular node involvement. Al- though some patients with localized skin involvement might be technically resectable,' it is generally accepted that patients who belong to Stage IIIB are not considered candidates for surgical re~ect ion.~.~

Until recently, the treatment of choice for Stage IIIA breast cancer was a radical mastectomy with, or without, postoperative radiation therapy.' The 5-year survival rate for patients with Stage IIIA disease varies, depending on the T and N composition of the group studied, between 30% and 45%, re~pectively~,~; the 10-year survival rate for this group of patients is 10% to 3 1% after local therapies, respe~tively.'.~ For Stage IIIB disease the outlook is much worse, with 5-year survival rates varying between 10% and 28%6*7 and 10-year survival rates around 10% or less.'**

For the last several decades the treatment of choice for

2507

2508 CANCER December 15 1988 Vol. 62

patients with Stage IIIB breast cancer has been radiation therapy, either preceding or after a total maste~tomy.~- '~ Most patients in both Stage IIIA and IIIB groups die as a consequence of distant metastases, and in a significant fraction of patients the local recurrence rate after regional therapy is also unacceptably high.4

In 1974 we initiated a combined modality treatment which included primary combination chemotherapy followed by surgery, or radiation, or both, to increase local and systemic control of locally advanced breast cancer. We describe here the results of treatment in patients with Stage 111, noninflammatory, primary breast cancer. A preliminary report that described the experience with the first 52 patients was published earlier.I3

Patients and Methods

All patients with histologic evidence of breast cancer, clinically staged as having a T3 or T4 primary tumor (excluding inflammatory breast cancer) or those with regional lymph node involvement falling in the N2 or N3' categories, and without evidence of distant metastases were eligible for this trial. There was no age limitation, and patients with bilateral tumors also were accepted. All patients were initially examined by a surgeon, radiotherapist, and a medical oncologist. In addition to a thorough physical exam and assessment of clinical stage, initial evaluation included a complete blood count (CBC) with differential and platelet counts, a biochemical survey (SMA- 12), chest x-ray, bilateral xeromammogram, bone scan, and liver imaging (radionuclide liver scan from 1974 through 1978, and hepatic ultrasound or computerized tomography from 1978 to 1985). Patients with pain, with an abnormal bone scan, or elevated alkaline phosphatase levels also had pertinent bone radiographs to rule out distant metastases. During treatment, patients were evaluated with a weekly CBC with differential and platelet count, an SMA-12, and urinalysis before each cycle of chemotherapy, a repeat mammogram of the involved breast after three cycles, chest x-ray, bone scan, and liver imaging every 3 to 4 months for the first 2 years. Subsequently, patient follow-up was done every 4 to 6 months for the next 3 years and yearly thereafter.

Initial chemotherapy consisted of 5-fluorouracil (500 mg/m2 intravenously [IV] on days 1 and 8), Adriamycin (doxorubicin) 50 mg/m2 IV on day 1, and cyclophosphamide 500 mg/m2 IV on day 1 of each 2 1-day cycle (FAC); in addition, bacillus Calmette-Guerin (BCG) was administered by scarification alternating the upper and lower extremities on days 9, 13, and 17 of each cycle. After three cycles of chemotherapy, all patients were reevaluated by the team, and local therapy was recommended. In general, patients who had achieved a marked tumor reduction, with minimal or no residual disease, were treated

with radiotherapy, whereas those with moderate residual disease had a total mastectomy and axillary dissection followed by radiation therapy. Those patients with no response, or progressive disease, often were treated with preoperative radiation therapy followed (if possible) by surgical resection. Initially, patients with large breasts had a mastectomy, even if minimal or no residual disease was present for radiotherapy-related technical reasons. Be- tween 1974 and 1978, patients treated with both local modalities had surgery first, followed immediately by radiation therapy. From 1978 until 1985, surgery was performed first, followed by adjuvant chemotherapy, and radiation therapy was delayed until completion of all systemic treatment.

After the initial locoregional therapy, FAC-BCG was continued until completing 450 to 500 mg/m2 of doxorubicin. Subsequently, methotrexate was substituted for doxorubicin, and a CMF-type combination was continued until 2 years of therapy were completed.13 The administration of BCG continued during CMF. After 1978, however, BCG was deleted from the entire treatment program. In 1980, the schedule of administration of doxorubicin was changed to a 48-hour continuous infusion, in view of our own institutional data showing that the continuous infusion schedule lowered the risk of doxorubicin-related cardiac toxicity without altering the therapeutic effi- ~ a c y . ' ~ . ' ~ Also, in 1980 we recognized that compliance with the 2-year treatment was poorI3 and it was also be- lieved that the maximum impact of chemotherapy was achieved during the initial 6 months of chemotherapy. Therefore, CMF maintenance therapy was discontinued, and the total duration of treatment was reduced from 2 years to approximately 9 months.

Surgery consisted of a modified radical mastectomy or an extended simple mastectomy (total mastectomy plus axillary lymph node dissection).

Radiotherapy after a total mastectomy consisted of 5000 rad to the chest wall plus a boost of 1000 rad to the scar, and 5000 rad to the internal mammary and supraclavicular lymph node chains. Whenever possible, the electron beam was used for treatment of internal mammary nodes to lessen the risk of cardiac toxicity. In patients treated with radiotherapy alone, 5000 rad were administered to the breast and the area of residual tumor was given an additional 1500 to 2000 rad to a restricted field. In patients with bulky residual disease after initial chemotherapy, twice daily fractionation was employed to a dose of 5500 rad to the breast over 5 weeks with a boost to areas of residual disease. The internal mammary and supraclavicular areas were treated to a total tumor dose of 5000 rad. The axilla was not irradiated after an axillary dissection.

Response to induction chemotherapy was evaluated by physical examination before each cycle of therapy and

No. 12 STAGE I11 BREAST CANCER MANAGEMENT * Hortobagyi et al. 2509

more recently, by a combination of physical examination and repeat xeromammograms. Standard criteria of tumor regression were utilized as described elsewhere. l 6 Disease- free survival was measured from the time of mastectomy or date of complete remission after chemotherapy or radiotherapy for the 168 patients rendered disease-free. Survival was measured from initiation of chemotherapy for all 174 patients treated. The Kaplan and Meier method was used to calculate and plot survival curves'7 and the generalized Wilcoxon test was used to test differences between them. ' *

Results

One hundred ninety-one eligible patients were entered on this prospective trial between March 1974 and March 1985. Of the 19 1 patients, 17 could not be evaluated: 14 had major protocol violations, mostly related to using different combinations of cytotoxic agents or deviations in the timing or appropriateness of local therapy. Two patients were lost to follow-up after the first cycle of induction chemotherapy, and one patient was treated with hormonal therapy alone. Among the 174 eligible and evaluable patients, 48 had Stage IIIA disease and 126 Stage IIIB primary breast cancer (Table 1). Four patients presented with synchronous bilateral breast cancer. Only the more advanced stage of the two tumors is included in Table 1 and the rest of the analysis. There were only 13 patients (7%) with clinically negative nodes and most patients had very advanced and neglected tumors. The pretreatment patient characteristics are detailed in Table 2. The median age was 5 1 years (range, 24-78 years). Fifty- four (43%) of the patients with Stage IIIB and 29 (60%) of those with Stage IIIA were 4 0 years old. Estrogen receptor assay by the dextran-coated charcoal method be- came routine practice at this institution in 1978, and pro- gesterone receptor in 198 1. Values > 10 fmol/mg protein were considered positive. None of the patients treated before 1978 had hormonal receptor assays performed on their tumor.

Table 3 shows the objective response status after the initial three cycles of combination chemotherapy. The complete response rate was slightly higher for patients with Stage IIIA disease than for patients with Stage IIIB. In addition, the combination of clinical and mammographic criteria for evaluation of response were more stringent and resulted in lower response rates than when responses were evaluated by clinical measurements alone. There were no patients whose tumor progressed during initial chemotherapy among the group of patients with Stage IIIA disease, whereas three patients among those with Stage IIIB developed progressive disease. The overall response rate to the initial three cycles of FAC, assessed by clinical measurements, was 87.4%. Of these, 16.7%

TABLE 1. Distribution of Evaluable Patients According to T and N Categories*

Regional nodal involvement

NO N l N2 N3 Total

Primary tumor TI-2 - - 9 i 8 17 T3 4 18 17 5 44 T4 9 35 37 32 113

Total 13 53 63 45 174 ~ ~~

* American Joint Committee on Cancer: Breast. In: Beahrs OH, Myers MH, eds. Manual for Staging of Cancer, ed. 2. Philadelphia: JB Lippincott 1983; 127-133.'

t No. of patients.

represented complete remissions and 70.7% represented partial remissions. An additional 4.6% of patients achieved a minor response.

One hundred sixty-eight of the 174 patients (96.5%) were rendered disease-free after the three cycles of chemotherapy and local treatment. Six patients, all with an initial diagnosis of Stage IIIB breast cancer, had residual disease. Three of these six patients experienced progressive disease during primary chemotherapy, whereas the other three achieved only a partial remission. Three of these

TABLE 2. Pretreatment Patient Characteristics

Stage

IIIA IIIB Total ~

No. of oatients (%)

Study group Age (yr)

<40 40-49 50-59 60-69 2 7 0

White Black Hispanic Other

Menopausal status Premenopausal Postmenopausal

Positive

Race

Estrogen receptor status

Age <50 Age 2 5 0

Age 4 0 Age 250

Primary tumor size

Negative

Unknown

<6 cm 6.1-9 cm >9 cm

48 (27.5)

13 (27.0) 16 (33.3) 7 (14.6) 8 (16.7) 4 (8.3)

29 (60.4) 12 (25.0) 4 (8.3) 3 (6.3)

30 (62.5) 18 (37.5)

6 (12.5) 9 (18.7)

18 (37.5) 7 (14.6) 8 (17.0)

12 (25.0) 32 (66.7) 4 (8.3)

126 (72.5)

24 (19.0) 30 (23.8) 39 (30.9) 27 (2 1.4) 6 (4.7)

80 (63.5) 26 (20.6) 20 ( I 5.9)

0

53 (42.1) 73 (57.9)

17 (13.5) 19 (15.0)

15 (11.9) 24 (19.0) 51 (40.5)

47 (37.3) 44 (34.9) 35 (27.0)

174 (100)

37 (21.2) 46 (26.4) 46 (26.4) 35 (20.1) 10 (5.7)

109 (62.6) 38 (21.8) 24 (13.8)

3 (1.7)

83 (47.7) 91 (52.3)

23 (1 3.2) 28 (16.1)

33 ( 1 8.9) 31 (17.8) 59 (33.9)

59 (33.9) 76 (43.7) 39 (22.4)

25 10 CANCER December 15 1988 Vol. 62

TABLE 3. Response Status After the Initial Three Cycles of 5-Fluorouraci1, Adnamycin (Doxorubicin), and

Cyclophosphamide Chemotherapy

Stage IIIA Stage IIIB

Method of evaluation Method of evaluation

Clinical Clinical Clinical + Clinical +

Response only mammography only mammography ~

Complete

Partial

Minor

remission 14 (29)* 8 (17) 15 (12) 10 (8)

remission 29 (60) 36 (75) 94 (75) 99 (79)

response 4 (8) 3 (6) 3 (2) 3 (2) No change 1 (2) 1 (2) 11 (9) 1 1 (9) Progression - - 3 (2) 3 (2)

* No. of patients (%).

patients were treated with radiotherapy alone and still had residual disease that was unresectable; the remaining three developed metastatic disease and were, therefore, not candidates for locoregional therapy.

Table 4 shows the regional therapy utilized in this multimodal protocol. Twenty-eight patients were treated with radiotherapy alone, and 103 were treated with both modalities of local treatment. Although surgery alone was not initially contemplated as adequate locoregional management, 40 patients were so treated. One of these 40 patients had a medical contraindication to radiotherapy; two died of intercurrent disease before radiotherapy; eight patients had no, or only microscopic residual disease at mastectomy and the attending physician opted for no additional local therapy; 12 patients developed distant metastases before scheduled radiotherapy; eight patients re- fused radiotherapy; and nine patients received no radiotherapy because of physician noncompliance.

We analyzed the correlation between clinical and radiographic complete remission and the absence of residual disease at the time of mastectomy. For this analysis, there were 12 1 patients who had achieved a complete or partial remission by clinical and mammographic criteria and who were treated with a mastectomy after the initial three cycles of FAC. Nine of these had a complete remission by

TABLE 4. Locoreeional Theraov Utilized

Stage IllA Stage IIIB Total

Radiotherapy only 3 (6)* 25 (20) 28 (16) Surgery and

radiotherapy 31 (64) 72 (57) 103 (59) Surgery only 14 (29) 26 (21) 40 (23) No local therapy - 3 (2) 3 (2)

* No. of patients (%).

TABLE 5. Correlation of Initial Clinical Nodal Status With Pathologic Nodal Status After Induction Chemotherapy

in 136 of 143 Patients Treated With Surgery*

No. of patients in subgroup of Clinical pathologic nodal status nodal No. of status patients 0 1-3 4-10 210

NO 10 7 1 1 1 N1 51 13 19 1 1 8 N2 53 9 18 17 9 N3 22 5 5 6 6

Total 136 34 43 35 24

* Seven patients did not have an axillary dissection.

the combined stringent criteria of response. All nine were found to have no residual disease in the mastectomy spec- imen. Of the 1 12 patients who reached a partial remission, one had no residual tumor at mastectomy, and 19 had minimal microscopic residual disease. The rest had substantial residual tumor. Patients who did not achieve at least a 50% tumor regression, or those who did not have surgery as their first modality of local therapy, were not included in this correlative analysis.

The correlation between initial clinical nodal status and residual nodal involvement at the time of surgical resection is shown on Table 5. Although most patients with clinical NO had no residual tumor at mastectomy, the distribution of patients with various degrees of surgical/ pathologic nodal involvement was evenly distributed among the other clinical nodal subgroups suggesting that downstaging after chemotherapy occurred in all subgroups.

As of September 1987, the minimum follow-up on this study was 30 months and the maximum, 156; the median follow-up was 59 months, being 65 for patients with Stage IIIB disease and 4 1 for patients with Stage IIIA. The disease-free survival is shown on Figure 1A for patients with Stage IIIA breast cancer and Figure 1B for patients with Stage IIIB breast cancer. Although the median disease- free survival for patients with Stage IIIA has not been reached, it was 30 months for patients with Stage IIIB. At 5 years, 7 1 % of the former and 33% of the latter remained disease-free. At 10 years, it is projected that 30% of patients with Stage IIIB breast cancer will remain disease-free. The overall survival of the 174 patients is shown in Figures 2A and 2B. The median survival for patients with Stage IIIA breast cancer has not been reached, and 84% and 56% of the patients were calculated to remain alive at 5 and 10 years, respectively (Fig. 2A). The median survival of the 126 patients with Stage IIIB disease was 48 months. Forty-four percent of these patients were projected to be alive at 5 years and 26% are projected to be alive at 10 years (Fig. 2B). Fifty-four patients (3 1%) were registered

No. 12 STAGE 111 BREAST CANCER MANAGEMENT Hortobugyi et ul.

1 .oo

g 0.80 t % g 0.60 6

FJ

C 0

Q

'g 0.40

2 a t Total Failed

- 4a 11 Oa20 t t I 1 I I 1 1 1 1 1 I I

1 .oo

0.80

0.60

0.40

0.20

2511

Total Failed

120 77 - I I I 1 1 I 1 1 I I J

24 48 72 96 120 24 48 72 96 120 Months B0 Months A0

FIGS. 1A AND 1B. (A) Disease-free survival of 48 patients with Stage IIIA breast cancer. (B) Disease-free survival of 120 patients with Stage IIIB breast cancer.

on this program before September 1977. Twelve of them (22%) remain alive; another six (1 1%) died of intercurrent disease without recurrent breast cancer and 36 (67%) died of recurrent or metastatic breast cancer.

Although age did not influence outcome among patients with Stage IIIA disease, patients younger than 50 years old seemed to do better among the group with Stage IIIB disease. This difference was statistically significant

1 .oo

0.80

Q) > 0.60

C 0 .-

0.40 l i

Total Failed

for disease-free survival only in this group (Tables 6 and 7). There was a trend for improved disease-free survival among patients with estrogen receptor-positive tumors, and there was a statistically significant survival benefit for this group of patients among those with Stage IIIB, but not for those with Stage IIIA. Patients with unknown estrogen receptor status had a prognosis similar to those who had estrogen receptor-positive tumors. There was an

1 .oo

0.80

0 > 2 0.60 c 0 ?z .-

0.40 2 a

O.*O I - 4 8 7 Total Failed 126 75 -

[1111111111 24 48 72 96 120 24 48 72 96 120

11111111111 Months B0 Months A0

FIGS. 2A AND 2B. (A) Overall survival of 48 patients with Stage IIIA breast cancer. (B) Overall survival of 126 patients with Stage IIIB breast cancer.

2512 CANCER December 15 1988 Vol. 62

TABLE 6. Five-Year Disease-Free Survival Rates According to Prognostic Factors

Stage

IIIA P IIIB P

Age (yr) <50 250

Estrogen receptor 2 10 fmol/mg <I0 fmol/mg Unknown

Tumor size 5 6 cm 6.1-9 cm >9 cm

Tumor stage T4 T3 T1,2,x

Nodal stage NO NI N2 N3

Response to induction

Complete remission Partial remission No response

With BCG Without BCG

FAC chemotherapy

Treatment

79 ] 0.67 63

0.244 100

0.101 49

0.834 72

” 63 ] 0.456

NA

0.405 NR

8o } 0.881 69

49 ] 0.005 21

0.099 41

0.796 36

0.005 89

41

0.014 14

26 ] 0.447 36

NA: not applicable: NR: not reached; FAC 5-fluorouracil, Adriamycin (doxorubicin), cyclophosphamide: BCG: bacillus Calmette-Guerin.

inverse correlation between the size of the primary tumor and 5-year disease-free, and overall survival. However, this was only a trend and did not reach statistical signif- icance.

There were 65 patients treated with BCG and 109 treated without immunotherapy. The response rate in both groups was the same. In addition both the disease- free and overall survival curves for the patients treated with and without BCG were superimposable, even after correcting for initial clinical stage. The 5-year disease-free and overall survival figures are given on Tables 6 and 7, respectively.

Sixty-five patients were treated with the 2-year program; of these six (9%) presented with Stage IIIA breast cancer and 59 (9 1%) with Stage IIIB disease. One hundred nine (42 with Stage IIIA, and 67 with Stage IIIB) were treated with the abbreviated program. The 5-year survival rate for the patients treated with the 2-year and the 9-month programs were 50% and 54%, respectively ( P = 0.1 1). The slight trend in favor of the group treated without maintenance therapy or BCG can be explained by the larger proportion of patients with Stage IIIA breast cancer in this group (38.5% versus 9%).

Response to induction chemotherapy correlated prominently with prognosis. Those patients who achieved a complete remission after the initial three cycles of FAC had a 5-year disease-free survival rate of 75% and an overall survival rate of 88%. Patients who only achieved a partial remission had an intermediate prognosis, and those with less than a partial remission or no response, had a very poor 5-year disease-free and overall survival. These differences were highly statistically significant for patients with Stage IIIB disease; for patients with Stage IIIA there was a trend in favor of complete responders but the differences were not statistically significant (P = 0.4) (Figs. 3A and 3B). Table 8 shows the 5-year disease-free and overall survivals for the various TN combinations.

Those patients who completed all treatment as planned (or those who completed all treatment as planned until recurrence) had a disease-free and overall survival substantially superior to those who either did not complete treatment, or had erratic therapy (Figs. 4A and 4B). More than 90% of scheduled doses were delivered to 92% of patients during the first cycle, 84.6% of patients during the third cycle, and 73.5% of them during the sixth cycle of chemotherapy. No patient received less than 60% of scheduled doses for any cycle of therapy.

TABLE 7. Five-Year Survival Rates According to Prognostic Factors

Stage

IIIA P IIIB P

Age (yr) 4 0 250

Estrogen receptor 210 fmol/mg i 10 fmol/mg Unknown

Tumor size 1 6 cm 6.1-9 cm 2 9 cm

Tumor stage T4 T3 TI ,2,x

Nodal stage NO N1 N2 N3

Response to induction

Complete remission Partial remission No response

With BCG Without BCG

FAC chemotherapy

Treatment

84 ] 0.772 82

0.308 100

;i ] 0.54 50

0.589 65

I!! ] 0.173

NA

0.596 NR

84 85 ] 0.711

ii ] 0.162

0.03 I 56

0.956 34

0.046 67 ’’ 29 ] 0.218

38

0.0 10 24

47 ] 0.610 41

NA: not applicable; NR: not reached; FAC: 5-fluorouracil, Adriamycin (doxorubicin), cyclophosphamide; BCG: bacillus Calmette-Guerin.

No. 12 STAGE 111 BREAST CANCER MANAGEMENT * Hortobagyi et al. 2513

Treatment Failure

Eighty-five patients (1 1 with Stage IIIA disease and 74 with Stage IIIB) have developed recurrent or metastatic breast cancer. Twenty-one patients (12.3%) had a chest wall recurrence and five (3%), a regional lymph node recurrence as first manifestation of disease. All 26 had Stage IIIB disease initially. None of the patients with Stage IIIA breast cancer developed locoregional recurrence. There was only one (5.5%) locoregional recurrence among the 18 patients who achieved a clinical/mammographic complete remission after induction chemotherapy; there were 22 among the 135 who achieved a partial response (16.3%) and three among the 18 with minor response or no change (16.7%). Two additional patients developed chest wall recurrence as secondary site of treatment failure. Fifty-nine patients developed distant metastases as first evidence of treatment failure. Among these, 20 patients had bone metastases; one, a second primary breast cancer; three, central nervous system metastases; and the rest, other visceral sites of disease.

Toxicity

Acute toxicity was similar to that observed with FAC chemotherapy for metastatic breast cancer. 19320 Almost

1 .oo

a 0.80 Q)

a u)

u)

C

r 0

0

4 g 0.60 5 .o 0.40

a h 0.20

.- a C 0

Q

a

- 'j ... .- 5 - ..

i j 2 ..... ." ... I-.. I ....____ " .__ .......................................

I .': i i _ .............

C 0

Q

.- 5

............ - ........................

TABLE 8. Five-Year Results According to TNM Substage

No. of Disease free Alive TNM substage patients ("I.) ("I.)

T 1 -2N2 9 12 65 T3NO 4 15 61 T3N 1 18 84 100 T3N2 17 51 82 T2N3 I 86 86 T3N3 4 0 25 T4NO 9 40 39 T4N 1 35 35 55 T4N2 31 19 31 T4N3 32 29 42

all patients developed total alopecia, and the majority developed some degree of nausea and vomiting. The fre- quency and severity of nausea and vomiting diminished after 1980, when doxorubicin started to be administered by a slow, continuous infu~ion. '~

Myelosuppression was severe, with the median lowest granulocyte count around 900. There were ten episodes of fever of unknown origin, three episodes of pneumonia, and six episodes of sepsis, three of them related to a central venous catheter. Two patients experienced thrombotic events, also related to a central venous catheter. Nine patients (5.1%) (ages 30, 41, 54, 54, 63, 66, 69, 69, and 78

1 .oo

0.80

0.60

0.40

0.20

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

1 I I I I 1 I I 1 I I 24 40 72 96 120 0 24 48 72 96 120

Months B Months Total Failed Total Failed Definition

A 0

- 10 2 Complete Remission - -. ............. 96 62 Partial Remission 14 13 NoChange .............

10 2 Complete Remission

17 15 NoResponse

- .- 99 58 Partial Remission ..............

FIGS. 3A AND 38. (A) Disease-free survival of 120 patients with Stage IIIB breast cancer according to response status after three cycles of induction chemotherapy (P = 0.014 [Gehan-Breslow test], 0.006 [log-rank test]). (B) Overall survival of 126 patients with Stage IIIB breast cancer according to response status after three cycles of induction chemotherapy (P = 0.01 [Gehan-Breslow test], 0.014 [log-rank test]).

25 14 CANCER December 15 1988 Vol. 62

1 .oo

g 0.80 4 .- 3 Q)

0.60

n

$ 0.40 2 a

C

0 a

. . . . . . . . . . . . . . . . . .

o.20 0 ILL-- 24 48 72 96 120

A Months

1 .oo

0.80

Q) > 0.60

C 0 .- 5 0.40

a

0.20

0 24 48 72 96 120 B Months

Total Failed Definition 5-yr DFS Rate Total Failed Definition 5-yr Survival Rate

- 132 63 Excellent Compliance 48% 132 52 Excellent Compliance 58% 13 12 Delayed Therapy 0 ................... 13 11 DelayedTherapy 20%

- 14 9 No Postoperative Therapy 28% .............. 14 9 No Postoperative Treatment 40% 15 10 Early Termination of Therapy 34% . . . . . . . . . 15 10 Early Termination of Therapy 42%

....

..... ....

. . . . . . .

FIGS. 4A AND 4B. (A) Disease-free survival o f 168 patients wi th Stage I11 breast cancer (Fig. 4A) according to the degree of compliance wi th treatment (P = 0.021 [Gehan-Breslow test], 0.002 [log-rank test]). (B) Survival of 174 patients wi th Stage 111 breast cancer (Fig. 4B) according to the degree of compliance wi th treatment.

years, respectively) developed clinical congestive heart failure; eight of these nine had received mediastinal radiation in addition to chemotherapy (six of these eight with cobalt 60 [60Co]). Six of the nine had received cu- mulative doses of doxorubicin in excess of 400 mg/m2 and the other three, between 300 and 400 mg/m2. Seven of these nine patients were treated before 1980 and received doxorubicin administered by bolus. During this trial we did not perform baseline or serial cardiac function tests (e.g., multigated cardiac [MUGA] scans).

Two of the nine patients died free of metastatic disease, in congestive heart failure. The other seven had their cardiac decompensation well controlled by standard medical therapy. One patient died without recurrent or metastatic breast cancer after developing an unresectable gastric carcinoma 24 months after the diagnosis of breast cancer. A second patient died without evidence of recurrent breast cancer of malignant melanoma diagnosed five years after the diagnosis of breast cancer. Two patients died of acute nonlymphocytic leukemia 22 and 91 months after the diagnosis of breast cancer, respectively. Their ages were 43 and 63 years, respectively; both had received radiotherapy to the chest wall and regional lymphatics and combination chemotherapy for 9 and 24 months, respectively. One patient with leukemia had also developed congestive heart failure (63 years old, included above).

We did not notice an increase in surgical complications

during the performance of this multimodality treatment. Radiation recall was observed in one patient who had chemotherapy restarted soon after completing radiation therapy. We did not notice an increase in the incidence or severity of radiation pneumonitis or chronic radiation changes on the chest wall or breast.

Discussion

We have presented a large series of patients with locally advanced breast cancer treated with a combined modality treatment program and maximum follow-up of 13.5 years. The results confirmed our earlier assessment that combination chemotherapy, surgery, and radiation therapy can be utilized optimally without increasing the toxicity of either form of therapy.13 We have not, however, administered chemotherapy and radiation therapy simul- taneously. All but six patients were rendered disease-free without resorting to either radical surgery or radical radiotherapy dose scheduling. In addition, this combined modality regimen resulted in an excellent local control rate superior to that recorded with surgery alone, radiotherapy alone, or the combination of both. A small, but substantial fraction of patients received appropriate treatment without removal of their breast.

We did not observe a difference in the efficacy of the treatment when we deleted BCG from the regimen or

No. 12 STAGE I11 BREAST CANCER MANAGEMENT * Hortobagyi et al. 2515

when we shortened the total duration of therapy from 2 years to 9 months. This confirms the clinical impression that most of the benefit of chemotherapy is obtained within the first 6 to 12 months of therapy.2’ Additional treatment with the same regimen is superfluous and adds unnecessary toxicity. This prospective trial did not address, and cannot answer, the question related to optimal local therapy .

Two other prospective trials have shown that similar local control rates can be obtained with surgery or radiation therapy.22 DeLena et a1.22 administered three cycles of doxorubicin and vincristine to patients with Stage I11 breast cancer (excluding those with supraclavicular involvement, or N3) and then randomized them to receive local treatment with surgery or radiotherapy. The incidence of locoregional recurrence was 29.6% for surgery and 3 1.1 % for radiotherapy.

Lesnick et al. recently reported the preliminary results of a prospective randomized trial.23 Patients with Stage I11 breast cancer (excluding those with supraclavicular nodes “31) were treated with multiagent induction chemotherapy (three cycles) and then randomly assigned to mastectomy or radiotherapy. Locoregional relapses occurred in 39% and 52%, respectively. These two trials suggest that equivalent local control was produced by surgery or radiotherapy; however, recurrence rate was still unacceptably high.

suggested that the combination of both surgery and radiation therapy produced better local control rate than either treatment modality by itself. If one had to choose a uniform local treatment approach we would agree that the combination of surgery and radiation therapy would produce the best results in a given patient population. However, based on our results and those of others who have utilized similar multimodal it appears that those patients who have a good response and tumor regression after induction chemotherapy could achieve optimal local control with radiotherapy alone or subtotal surgical resection and radiation therapy. On the other hand, patients who do not respond well (or at all) to induction chemotherapy would probably require both modalities of local treatment for optimal local control.26 De- pending on whether the tumor is technically resectable or not, surgery would be followed or preceded by radiation therapy.

The correlation of survival with response to therapy has been criticized on various ground^.^' We have adopted the landmark method in this analysis since all patients (regardless of response) completed at least two cycles of chemotherapy. Therefore, there were no exclusions for reasons of early death. In fact, the earliest death occurred 7 months after initiation of primary chemotherapy. Therefore, we are confident that the correlation of response and survival is valid.

Reports by Balawajder et and Bedwinek et

This study confirmed our earlier rep01-t’~ that most patients with “unresectable” and “incurable” locally advanced breast cancer could be rendered disease-free, and many of them remained free of treatment failure 5 and 10 years after the initiation of therapy. Although this was not a randomized trial, extensive historical experience suggested that this long-term disease-free survival was a departure from what could be achieved in this poor prognostic group of patients before systemic therapy was in- troduced in the overall treatment ~trategy.~-’*~’

However, since somewhat different staging classifica- tions have been used in published reports, intertrial com- parisons are quite difficult. Because most patients with Stage IIIB breast cancer have unresectable tumors to start with, a prospective randomized trial is not feasible in un- selected patients. However, in patients with Stage IIIA breast cancer, and those few patients with resectable T4 primary tumors, a prospective randomized trial to com- pare surgical resection followed by adjuvant systemic therapy with the multimodal treatment described here would be appropriate and necessary. We believe, however, that even if there is no survival advantage in such a comparison, the ability to preserve the breast by utilizing induction chemotherapy and radiation therapy in good responders, as well as the information obtained relating to the in vivo responsiveness of the tumor to the systemic therapy utilized, are advantages that favor this multimodal treatment strategy.

We have previously reported that clinical evaluation of response to chemotherapy in patients with locally advanced breast cancer is not as accurate as the pathologic evaluation of response, and that the latter has greater prognostic value.29 In our current analysis, when mammographic evaluation of response was combined with clinical assessment, the correlation with pathologic disease status improved substantially. Thus, residual disease can be detected accurately with noninvasive methods and, therefore, surgical resection can be obviated for those patients who are considered in clinical/radiographic complete remission.

The surgical-pathologic staging system is the most accurate and accepted today. By using primary or induction chemotherapy, an accurate initial staging is not obtained. After three or four cycles of chemotherapy, many patients’ tumors are substantially down-staged, and their surgical- pathological staging at that time does not correspond with their initial stage. Therefore, it is of the utmost importance to define initially the clinical stage as accurately as possible. This can be done by using careful clinical and mammographic measurements, and in patients in whom the mammogram is not helpful in defining the extent and size of primary and nodal involvement, computed tomography scan has been helpful. Fine-needle aspiration has become a very reliable technique in diagnosing the presence of malignant di~ease.~’,~’ It can be used success-

2516 CANCER December I5 1988 Vol. 62

fully to confirm the presence of metastatic disease in axillary or supraclavicular nodes.

Fine-needle aspiration also may provide material for additional studies of great importance in the determina- tion of prognosis. Thus, estrogen receptor content, flow cytometric analysis, and cytogenetic studies all can be performed on cytologic aspirates. We expect that primary or induction chemotherapy will find expanded use in the treatment of primary breast cancer. It is therefore of the utmost importance not only to improve our methods of initial noninvasive staging but also to develop prognostic criteria based on the new markers that can be obtained through fine-needle aspiration. In this manner, today's gold standard (surgical-pathological staging) may not be necessary in the future to determine prognosis and the most appropriate therapeutic approach.

Despite the apparent improved prognosis among patients with Stage IIIB primary breast cancer, the majority of these patients still develop distant metastases. The local control rate in this trial appeared satisfactory, but more effective systemic treatment regimens will be necessary to further improve the outlook of these patients. In this trial we did not use hormonal therapy. The role of hormonal treatment in Stage 111 breast cancer needs to be evaluated. In addition, those patients who manifest resistant disease to induction chemotherapy and whose prognosis is so poor would be excellent candidates to try alternative chemotherapy combinations, including high-dose chemotherapy with autologous bone marrow rescue.

Based on our results and those of others we believe that this multimodality regimen which includes primary chemotherapy is the treatment of choice for patients with Stage IIIB primary breast cancer. Although other alter- natives may offer similar results to patients with Stage IIIA breast cancer, we suggest that multimodality treat- ments like the one used in this study would also represent the optimal treatment strategy for operable breast cancer. Prospective trials to confirm this hypothesis will be necessary.

REFERENCES

1. American Joint Committee on Cancer. Breast. In: Beahrs OH, Myers MH, eds. Manual for Staging of Cancer, ed. 2. Philadelphia: JB Lippincott, 1983; 127-133.

2. Zucali R, Kenda R. Small size T4 breast cancer: Natural history and prognosis. Tumori 198 1; 67:225-230.

3. Haagensen CD. Clinical classification of the stage of advancement of breast carcinoma. In: Diseases of the Breast, ed. 3. Philadelphia: WB Saunders, 1986; 851-863.

4. Davila E, Vogel CL. Management of locally advanced breast cancer (Stage 111): A review. Int Adv Surg Oncol 1984; 7:197-327.

5. Fracchia AA, Evans JF, Einsenberg BL. Stage I11 carcinoma of the breast: A detailed analysis. Ann Surg 1980; 192:705-710.

6. Sicher K, Waterhouse JAH. Evaluation of TNM classification of carcinoma of the breast. Br J Cancer 1973; 28380-588.

7. MacKay EN, Sellers AH. A prospective trial of the TNM classification of breast cancer by the Regional Cancer Treatment Centres in Ontario, 1960-1964. Int J Cancer 1970; 6517-528.

8. Garcia Vilanova A, Company Serrano RA, Fuster Diana E et a[.

Sistema TNM 1978 en cancer de mama: Propuesta de modificacion. Revisiones Espanol Oncol 1985; 32:95- 108.

9. Fletcher GH, Montague ED. Radical irradiation of advanced breast cancer. Am J Roenteml 1965; 93573-584.

10. Rubens RD, Armitage P, Winter PJ, Tong D, Hayward JL. Prog- nosis in inoperable Stage 111 carcinoma ofthe breast. Ear J Cancer 1977;

1 1. Zucali R, Uslenghi C, Kenda R et al. Natural history and survival of inoperable breast cancer treated with radiotherapy and radiotherapy followed by radical mastectomy. Cancer 1976; 37:1422-1431.

12. Bedwinek J, Rao DV, Perez C et al. Stage 111 and localized Stage IV breast cancer: Irradiation alone versus irradiation plus surgery. JRadiat Oncol Biol Phys 1982; 8:31-36.

13. Hortobagyi GN, Blumenschein GR, Spanos W et al. Multimodal treatment of locoregionally advanced breast cancer. Cancer 1983; 5 I :

14. Legha SS, Benjamin RS, Mackay B et al. Reduction ofdoxorubicin wdiotoxicity by prolonged continuous intravenous infusion. Ann Intern Med 1982; 96:133-139.

15. Hortobagyi GN, Frye D, Blumenschein GR et al. FAC with Adriamycin by continuous infusion for treatment of advanced breast cancer. Proceedings of the 19th Annual American Society of Clinical Oncology Meeting (Abstr C-408) May 22-24, 1983; 2:105.

16. Hayward JL, Carbone PP, Heuson JC et al. Assessment of response to therapy in advanced breast cancer: A project of the Programme on Clinical Oncology of the International Union Against Cancer, Geneva, Switzerland. Cancer 1977; 39:1289-1294.

17. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958; 53:457-481.

18. Gehan EA. A generalized Wilcoxon test for comparing arbitrarily singly-censored samples. Biornetrics 1965: 52:203-223.

19. Hortobagyi GN, Gutterman JU, Blumenschein GR et al. Com- bined chemoimmunotherapy for advanced breast cancer: A comparison of BCG and levamisole. Cancer 1979; 43: 1 1 12-1 122.

20. Hortobagyi GN, Gutterman JU, Blumenschein GR. Combination chemoimmunotherapy of metastatic breast cancer with 5-fluorouracil, Adriamycin, cyclophosphamide, and BCG. Cancer 1979; 43: 1225-1233.

21. Tancini G, Bonadonna G, Valagussa P et al. Adjuvant CMF in breast cancer: Comparative 5-year results of 12 versus 6 cycles. J Clin Oncol 1983; 1:2-10.

22. DeLena M, Varini M, Zucali R ef al. Multimodal treatment for locally advanced breast cancer: Results of chemotherapy-radiotherapy versus chemotherapy-surgery. Cancer Clin Trials I98 1 ; 4:229-236.

23. Lesnick G, Perloff M, Korzun A ef al. Combination chemotherapy prior to surgery or radiotherapy in locally advanced breast cancer. In: Jacquillat C, Weil M, Khayat D, eds. Paris: John Libbey Eurotext, 1986;

24. Balawajder I, Antich PP, Boland J. An analysis of the role of radiotherapy alone and in combination with chemotherapy and surgery in the management of advanced breast carcinoma. Cancer 1983; 51:

25. Jacquillat C, Weil M, Auclerc G et al. Neo-adjuvant chemotherapy in the conservative management of breast cancers: Study on 205 patients. In: Jacquillat C, Weil M, Khayat D, eds. Paris: John Lihbey Eurotext,

26. Ragaz J, Manji M, Olivotto R et al. Role of mastectomy (M) in preoperative (neoadjuvant) combined modality therapy (CMT) of locally advanced breast cancer (LABC). Proceedings of the 23rd Annual Amer- ican Society of Clinical Oncology Meeting (Abstr 212), May 17-19, 1987, Atlanta, Georgia, 655.

27. Anderson JR, Cain KC, Gelber RD. Analysis of survival by tumor response. J Clin Oncol 1983; 1:710-719.

28. Schottenfeld D, Nash AG, Robbins GF, Beattie EJ. Ten-year results of the treatment of primary operable breast carcinoma: A summary of 304 patients evaluated by the TNM system. Cancer 1976; 38:lOOl- 1007.

29. Feldman LD, Hortobagyi GN, Buzdar AU, Ames FC, Blumen- schein GR. Pathological assessment of response to induction chemotherapy in breast cancer. Cancer Res 1986; 46:2578-2581.

30. Strawbridge HTG, Bassett AA, Foldes I. Role of cytology in management of lesions of the breast. Surg Gynecol Obstet 1981; 152: 1-7.

3 1. Bell DA, Hajdu SI, Urban JA, Gaston JP. Role of aspiration cytology in the diagnosis and management of mammary lesions in office practice. Cancer 1983; 51:1182-1189.

1 3~805-8 1 1.

763-768.

207-2 1 1.

574-580.

1986; 197-206.

management of stage iii primary breast cancer with primary chemotherapy, surgery, and radiation...

Documents