- Email: [email protected]
Treatment of locoregionally advanced breast cancer with surgery, radiotherapy, and combination chemoimmunotherapy
IN. 1. Radialion Oncolo~ Bml. Phys.. Vol Printed in the U.S.A. All rights raerwd
03sCr3016/S3/050643~S03.00/0 Copyqht Q 1983 t’ergamm Press Ltd.
9. pp. 643-650
??Original Contribution
TREATMENT OF LOCOREGIONALLY ADVANCED BREAST CANCER WITH SURGERY, RADIOTHERAPY, AND COMBINATION CHEMOIMMUNOTHERAPY G. N. HORTOBAGYI,M.D.,* W. SPANOS, M.D.?E.D. MONTAGUE, M.D.,?A. BUZDAR, M.D.,* H.-Y.YAP, M.D.* AND G. R. BLUMENSCHEIN,M.D.*
U.
The University of Texas M. D. Anderson Hospital and 6723 Bertner Avenue,
Tumor Institute at Houston, Houston, TX 77030
Fifty-two patients witb locally advanced primary breast cancer (T,, T,, N2, N,) but MI evidenceof distant metastases were treated witb tbree cycles of combination cbemotberapy. The regimen consisted of !Muorouracil, Adriamycin, cyclophosplmmide,and Bacillus Calmette-Cuerin (FAC-BCG), followed by local tberapy (simple mastectomy and/or radiofberapy to tk breast/chest wall and tbe regioaal lymphatic system) and adjuvant chemotherapy for two full years. The rem&s were compared with hose in an bistorkal control group of 52 patients ma&bed for initial stage of disease wk were treated by a simple mastectomy and postoperative radiotbempy only. Forty-nine (94% ) of 52 FAGtreated patients and 48 (92 %) of tbe control patients became free of clinically detectabie disease. At tbe median follow-up time of 56 months, 37.5 % of the FAGtreated patients and 19.5 W of tbe control patients bad remained free of disease. FAGtreated patients wbo completed 2 years of tberapy and in whom ?? djuvant cbemotbempy was started promptly after bcal treatment bad a 48% disease-free sunlivnl rate of 4 years. In tbose in whom tbe initial nmaifestathm was supmcbvkuhr idvolvement,tbe estimated S-year disease-free survival rate was 42% for patieuts treated with FAC amd 9% for coatroi patients. There were local recurrences in 25% of FAGtreated patiatts and 23 W of control patients (not sigttiftcant). LXstant metastasesdevelopedin 50 % of FAGtreated patients and 77 % of coatrol patients (p c 0.01). Tbe medirn dii free interval was 2S months in tbe FAGtreated group and 11 months in the control group (p - 0.025). Tbe greatest improvement in prognosis was in patients witb supmchvkul8r involvement;tbe median diifree survival was 26 months in FAC-treated patients and 6 months in tbe control group (p = 0.007). This multimodal approach et&&rely readers tbe majority of patients ivitb locoregionally advanced breast cancer free of disease and prolongs the d&ease-free survival period. Breast cancer, Combined modality treatment, 7’,-T, breast cancer, Cbemotberapy, Surgery, Radiotbempy.
INTROdUCTION In the majority of patients with primary breast cancer, no evidence of distant metastases is discernible and most have tumors that are potentially curable by surgery. A small proportion of patients, however, have large primary tumors involving the skin or deeper structures and, in some cases, large regional metastases are present (i.e., matted and fixed axillary nodes or subciavicuiar or supraciavicuiar metastases). These tumors conform to the T,, T4, IV,, IV3subgroups of the American Joint Commission’s TNM classification. Most women with T,, TI tumors but without sizabie axiiiary or supraciavicuiar metastases technically could be suitable candidates for mastectomy; however, the experience of most investigators suggests that radical excision of these tumors provides no iongterm benefits and is soon followed by multiple early
recurrences.” Patients with N, tumors are not, of course, surgical candidates. For these reasons, the mainstay of therapy for this subgroup of patients had been radiotheraPY.3.4.‘3While improvements in radiotherapy techniques have enhanced local control, many patients still die of distant metastases. In most patients the disease-free interval is brief, and the doses of radiotherapy required for local control produce significant side effects. Combination chemotherapy in breast cancer has become increasingly effective in recent years, a fact affirmed by the response to treatment in 50 to 80% of patients. 5*6Since 1974 we have treated patients in the T,, T,-IV3 subgroups with combination chemotherapy prior to definitive local treatment in an attempt to decrease the need for radical surgery or the radical radiotherapy doses necessary for adequate control. The results of this muitimodal program are defined in this paper and are com-
Presented in part at the American Radium Society Meeting, Philadelphia, Pennsylvania, April 26-May 2, 1980. *Medical Breast Service, Department of Internal Medicine. tDepartment of Radiotherapy.
Reprint requests to: G. N. Hortobagyi, M.D. Acknowfedgmenf-The authors would like to thank Joan Trammell for her assistance in the preparation of this manuscript. Accepted for publication IO December 1982. 643
Radiation Oncology 0 Biology0 Physics
644
pared to our historical experience with local treatment alone.
May 1983, Volume 9. Number 5
Table 2. Advanced primary breast cancer: Treatment program Surgery FACx3 1
METHODS AND MATERIALS Eligibility criteria
Treatment plan
The initial treatment regimen consisted of three cycles of chemotherapy with 5-fluorouracii, Adriamycin, and cyclophosphamide along with nonspecific immunotherapy with Bacillus-Calmette-Guerin (FAC-BCG) at the doses and schedule shown in Table 2. The patients then were reevaluated and, depending on their response status after chemotherapy, local treatment was planned. A tumor board consisting of a medical oncologist, surgical oncologist, and radiotherapist made the decisions about local therapy. An attempt was made to schedule surgery approximately
4to 14 weeks
9 weeks
Breast cancer patients with locally advanced primary tumors classified as T,, T,, iV,, iV3 were included in this trial. All tumors were graded according to the American Joint Commission modification of the TNM classification. Patients with advanced primary tumors amenable to complete surgical resection and patients with inflammatory carcinoma were not entered in this study. Patients with overt congestive heart failure dr uncontrolled hypertension or those otherwise ineligible for chemotherapy were not included. Patients were not excluded for reasons of age or performance status. Patients with distant metastases at the time of diagnosis were excluded. Fifty-nine eligible patients were entered in this program between 1974 and 1976. Seven patients were considered inevaluable because they did not receive local therapy as dictated by protocol. One achieved a complete remission after 8 cycles of FAC, had no local therapy, and remained free of disease 20 months after diagnosis; four had local therapy only after 6 to 11 courses of chemotherapy (2 of them after evidence of local progression), and all of them died 19 to 24 months after diagnosis; and 2 were lost to follow-up after the first dose of chemotherapy. Fifty-two patients completed at least ihe first three courses of chemotherapy and local therapy with surgery, radiotherapy, or both. The distribution of these patients and thiir matched controls by TNM subcategories are shown in Table 1.
three
weeks following
the third
Radiotherapy -
FACx6-CMk
81 to91 weeks
2 years
F>C S-fluorouracil Adriamycin Cyclophosphamide Immunotherapy with BCG
> 500 mg/m* IV D I + 8 50 mg/m* IV D I 500 mg/m* IV D I 6 x IO8 units by scarification on days 9, 13, and I7
CMF
Cyclophosphamide Methotrexate S-fluorouracil Immunotherapy as with FAC
500 mg/m* PO D 2 I + 8 30mg/m’IMD 500 mg/m* PO D I + 8
Duration of cycles, FAC or CMF: 21 days.
chemotherapy and to start radiotherapy two to three weeks after surgery or three weeks after the last dose of chemotherapy. Chemotherapy and radiotherapy were not administered simultaneously. A tumor dose of 5,000 rad was administered with a Cobalt 60 machine to the chest wall or breast and the internal mammary, supraclavicular. and axillary areas. The area of residual tumor was given an additional 1,500 to 2,000 rad to a restricted field in the manner described by Brown ef al.’ In patients with bulky disease after initial chemotherapy and who received radiotherapy only, fractionation was employed twice daily to a dose of 5,500 rad to the breast over five weeks with a boost to areas of residual disease. Whenever possible, the electron beam was used for treatment of internal mammary nodes to lessen the risk of cardiotoxicity. Twenty-five patients were given local treatment consisting of surgery or radiotherapy (Table 3), and 27 patients were treated with both modalities. Seven patients did not receive radiotherapy. Of these seven patients, the mastectomy specimen in two contained no evidence of Table 3. Local therapy
cycle of No. free
Table 1. Distribution of FAC-treated TNM subgroups
patients by
Rsponse to initial FAC
T*
0 2 21 25
4 6 23 31
10 8 27 45
14 16 71 100
Note: Controls were matched by TNM subgroup. Numbers represent percentage of patients in each subcategory.
treatment
No. of patients
Surgery Radiotherapy (RT) Surgery + RT
2 5
(35 patients)
Surgery RT Surgery + RT
5 12 18
Stable disease (9 patients)
RT Surgery + RT
I 8
Complete remission
T, T, Total
Type of local
(8 patients) Partial remission
I
di&e 1
3
Treatment of locorepionally
advanced breast cancer 0 G. N. HORTOBAGYIet al.
tumor; in one, radiation necrosis of the thin skin Rap was being avoided; in another one mastectomy was on the same side as the only lung (agenesis); and one patient with concurrent stomach cancer became lost to follow-up after mastectomy. Two other patients with advanced bilateral primary breast cancer were staged on each side, but in this report they are considered as one patient each, counted by the more advanced of the two cancers. Three weeks after completion of radiotherapy or two weeks after surgery, chemotherapy with FAC-BCG was reinstituted until a total cumulative dose of 450 mg/m2 of Adriamycin was reached. Adriamycin was then discontinued and maintenance chemotherapy with cyclophosphamide, methotrexate, and Huorouracil (CMF) plus BCG was started at the doses and schedule shown in Table 1. The planned duration of therapy, systemic and local, was 24 months. All patients initially were staged with a complete history, physical examination, hemoglobin, hematocrit, white blood cell count with differential and platelet counts, SMA-12/100, carcinoembryonic antigen (CEA), chest Xray, metastatic bone survey and/or bone scan, and liver scan. Bilateral xeromammograms also were performed. Peripheral hematologic values were obtained weekly during chemotherapy and radiotherapy, and the doses of chemotherapy were adjusted to the degree of myeiosuppression as described in an earlier report.6 Tumor measurements were monitored every three weeks until a disease-free status was reached. A complete physical examination and metastatic survey were done at three to four month intervals during therapy and every four to six months thereafter. Before entering a patient in the trial, histologic proof of malignancy and recurrent disease was obtained if possible. A complete remission was defined as disappearance of ail subjective and objective evidence of tumor. A partial remission represented a ~50% reduction of the product of the largest perpendicular diameters of measurable lesions without the development of new lesions. Patients with ~50% decrease or ~25% increase in tumor measurement were considered to have stable disease. Ail other patients had progressive disease. Both palpation and mammographic measurements were used to determine responses.
Control group An historical control group of patients was selected from the files of the Department of Radiotherapy by matching initial stage of disease for each treated patient. Three of them had bilateral breast cancer and were treated in the same manner as the two in the chemotherapy group with this involvement. The diagnosis of cancer had been made in two patients in the control group (both NJ in 1959 and in 13 between 1960 and 1970 (9 with IV,). The rest had been diagnosed and treated between 1970 and 1975. Staging workup for these latter patients was similar to that employed in the chemotherapy-treated
645
group. Radionuclide scans were not performed in the patients seen prior to 1970. A simple mastectomy, with resection of gross axillary disease, was performed in all patients in the control group followed by radiotherapy to the chest wall and peripheral lymphatic nodes, as was done in those in the chemotherapy group with this type of involvement. Treatment after recurrence or the appearance of distant metastases in the chemotherapy-treated group consisted of various hormonal manipulations, chemotherapy, or local salvage treatment (surgery or radiotherapy). Patients in the control group received mainly single-agent chemotherapy or hormonal therapy. Only 12 of these latter patients were treated with an Adriamycin-containing combination, while two others were given cyciophosphamide, methotrexate, and 5-fluorouracil. Disease-free survival time was calculated from the date of mastectomy or the date of complete remission. Survival was calculated from the date of diagnosis for both groups of patients. The Kaplan and Meier method was used to calculate and to plot disease-free intervals and survival curves,’ and a generalized Wilcoxon test (two-tailed analysis) was used to test differences between curves.” RESULTS After three courses of chemotherapy, 8 patients achieved complete remission (CR), 35 patients partial remission (PR), and 9 patients remained stable. Progressive disease was not detected in any of the 52 patients during this treatment period. After the initial three cycles of chemotherapy and local treatment, 49 of 52 patients (94%) were rendered free of disease. Of the three who did not respond, two had large supraclavicular nodes and one had fixed axillary nodes. Forty-eight of 52 control patients (92%) were rendered free of clinical disease after simple mastectomy and radiotherapy. The initial finding in ail four who did not respond was supraclavicular nodes. In general, response to chemotherapy was prompt. The median time to achieve a partial remission was six weeks. The pretreatment status of both groups of patients is shown in Table 4. The age and menopausal status distribution were similar in both groups. There was a higher proportion of blacks and Latin Americans in the group with locally advanced breast cancer than is the case in the general breast cancer population referred to our institution. Similarly, in the few patients for whom estrogen receptor results were available, a predominance of estrogen receptor-positive tumors was apparent. Recurrences
The median follow-up time for the chemotherapytreated patients was 56 months (range, 11 to 90) and for the control group 68 months (range, 55 to 124). In 32 of the 52 (62%) chemotherapy-treated patients disease reappeared; local recurrence occurred in 6, distant metastases in 19, and both types in 7. Local recurrence deveioped in 2 of 7 patients treated with FAC and surgery, 4 of
Radiation Oncology
646
0 Biology 0 Physics
May 1983, Volume 9, Number
5
Dbeue-FraoWvtvd Table 4. Characteristics at diagnosis Chemotherapy group Age (median (range)
Control group
Total Poll 82 82
(2Qj-578)
Menopausal status Premenopausal Postmenopausal
21 31
Estrogen receptor + Unknown
12 4 36
Race White Black Latin American
32 IO 10
32 . FAG 42 . ComfOl
I D~~uo-F~~
(29?9)
0-0.026
14 38 Not available Not available 52
(14)
32 14 6
18 treated with FAC and radiotherapy, and 5 of 21 patients treated with FAC, surgery, and radiotherapy (Table 5). Four in whom disease again recurred achieved a disease-free state following surgery. On the other hand, locoregional recurrences developed in 12 of 52 (23%) control patients during their clinical course. Seven of 52 control patients (13%) and none of the FAC-treated patients had recurrences in the supraclavicular area. Distant metastases developed in 26 (50%) FACtreated patients and.in 40 (77%) control patients. There was no difference in the distribution of metastatic sites between the FAC-treated and control groups except for a lower incidence of pulmonary metastases and supraclavicuiar recurrences in the former group. The incidence of second primary breast cancer was similar in both groups. The median disease-free interval was 25 months for FAGtreated patients and 11 months for control patients (p = 0.025) (Fig. 1). The prognostic factors used to analyze the rate of failure are listed in Table 6. Twenty of 31 (64%) postmenopausal patients and 11 of 21 (52%) premenopausal patients in the FAC-treated group had progression of disease. The respective figures in the control group were 76% and 71%. The differences between premenopausal
000’ 0
48
24
72
Fig. I. Disease-free survival of FAGtreated control groups.
and historical
and postmenopausal groups were not significant. When the data on patients in both the FAC-treated and control groups were analyzed by age groups, the relapse rates were found to be similarly distributed. Race had no influence on relapse rate. Estrogen receptor information was available in only 16 FAC-treated patients and none of the controls. All four patients who had estrogen receptornegative tumors suffered a relapse. At five years the estimated relapse rate was 91% in control patients who initially were found to have a supraclavicular node (NJ and 7 1% in patients without supraclavicular involvement (Figs. 2 and 3). Conversely, in the group that received chemotherapy, the five-year estimated relapse rate was 57% for patients who had supraclavicular involvement and 77% for those who did not. These results represent a
Table 5. Treatment failures Control group No. of relapses Local and regional
Local and regional only
Chest wall Axilla Supraclavicular
7 2 7
10 (19%)
Distant only
30 (58%)
Second primary breast cancer
Total no. patients with relapse I l/52 (21%)
Chemotherapy group No. of relapses 13
Total no. patients with relapse 13/52 (25%)
0 6 (12%) 7 (13%) 19 (37%)
2 (3.8%)
Local and distant
w
l/52 (1.9%)
2/52 (3.8%)
pue
(IN) I#!M sluayed
II!aaueqdwo:,
Ioword
pazkleue
aM ‘dnoA? [oiluo~ aql u! lolDe3 ysou8old ~ueliodru! lsoul aql aq 01 paieadde luaruaA[oau! lepou leyu! amg .aseas!p 30 ama.mmai paaua!ladxa (go9) sluaged oz asaql 30 ZI pue ‘Ldelaql pauueld +aql ala[dt_um lou p!p oqm swayed 103 sqluow 11 seM po!lad IeAglns
aaq-aseas!p
uegxmx aqL ‘(y&5)
asdelal e paiagns
ruaql
Jo 61 pue SqluouJ sz SF?M swayed asaql 10~ Iehlalu! aalj -asrras!p ue!paru aqL .asdelar [gun lualulearl JO AdeJaql [mo[ pue c+ualsdS 30 sJe!arComi palaldum sluaged ZE .+x0 *(aum.Ju%!s Alleyls!~els lou) SoE 30 ~e+.ms JeaL -in03 e peq iah sm Aelap aqi woqm u! asoqL ‘%gt 30 [e!yms leakno e peq luaruleall IeaoI laye &dluold palms SzM f>38-3VJ 30 aSin qllnOJ aql LUOqM u! sluaged *ICdeJaqlsyal& 30 uoydrunsaJ aql papazald sqaam 62 01 L uaamlaq &lap ale!idolddeu! sluaged lsou~ III ‘luawleail leaoo( ial3e 6lldruoid pawnsal Lde -JaqlouIaq3 SBM sluaged pal~!all-f)3$~-3~.~ p 1 QO UI ~sqlu0t.u z 1 pue L le la3um lseaiq a+sal%rd 30 pa!p Adwaql [emI pua /(deJaqlouraqD 30 sasmoa aalql alaldum e aAa!qae IOU leyu! aql %u!Mo~]o~ uo!ssyal p!p oqm swayed aaiql aql30 0mL ‘(SN = d) r(lar\!lmdsal ‘sqwow 81 pue ‘0~ ‘sz aJaM asuodsa3 afqels JO ‘leg -led ‘ala[dum e paAa!qae oqm swayed lo3 spo!lad [e+ -ins say-ase?s!p ueyam aqL .pawnleAa sv~ 0sIe aseas!p 30 aal %u!u!ewaJ 30 hyqeqold aql uo AdelaqlouxaqD 30 sapki aalqi le!l!u! aql 01 asuodsar 30 aauanyu! aqL .iCla%ins01 lo!ld (cg& sJoum1 E-f~ ql!m paluasaid oqm asoqi u! ueql s!sou%wp 30 awl aql ie Jouml “J e ql!m
.uoyenieAa leyu! jo auy aql 11?sap011leIns!Aeplndns lnoql!m sluayid Io1luoa pue palea4l-~y~ Jo leA!hlns aaij-aseas!a 'E f&d
-
0
‘Ilop
-en[eAa re!l!u!30 ayi aqlle sapou JE[na!At?13eidns ql!M sluayed [OnUO:, pue paviaJi-yf~ Jo pzA!Alns aay-aseas!a '2 +!d WuW
@C
1_
‘TTY”’
sluaged u! (o~>zE)lawd seM a3uammal leaol ‘Jalramoq ‘dnol% palearl bdeiaqlomaqa aql UI ‘(9 alqel) dnol% loilum aql u! awl aauaiinaai 1mo1 qly alelaiJoa lou p!p lseaiq aql30 luawar\loy ioum dleuyd 30 iuaixa aqL .( E %J) (9L.o - d) luvX@!s lou se* (sqluoru L 1 ‘umpaur) dnol% low03 agztdsal I!aqt purz (sqwom EZ ‘ue!mu) Ed lnoql!m swayed palEaJl-ddEJaqiOwaq3 aql30 ~e++.ms aai3-aseas!p aqi uaamlaq aaua3aJ!p aqL ‘(~ofj*o - d fsqluolu 9 ‘ue!paur) uogelndod jorlum aql ql!m paieduroa (sqluow sz ‘uegm~.~) dnoi% f~ paleaq-/(delaqlousaqD aql Jo3 am!1 le+uns aaJ3-aseas!p u! aseamy alqeJap!suoa
BZE %I!
%PZ o%OZ
‘L t-1
L
1ualxalotunL
C-I
auop 1ON auop ION
N rN
lUaWaA[OAU! [EpN aA!le%aN aA!l!SOd
loldaaal ua%oiisg lesnedouaumod [esnedouatuald dnoL3 ideJaqlouIaq3
dnoh? [OJIUO~ SJOl3EJ3!lSoU%Old ‘9 a1qe.J.
LV9
648
Radiation Oncology 0 Biology 0 Physics
without (N,_,) initial supraclavicular involvement. Twenty of 24 (83%) N, patients completed two years of therapy or received therapy until relapse, while only 12 of 29 (42%) N,_, patients did so. Furthermore, 10 of 24 (42%) N, patients were restarted on chemotherapy promptly after local treatment compared with only 4 of 28 (14%) N,_* patients who completed their therapy according to the planned program. Poor patient compliance was unexpectedly frequent in this group of patients in contrast with a drop-out rate of 5% or less in our experience with patients with metastatic breast cancer. Compliance with local treatment was excellent. Survival
Twenty-eight of 52 (54%) chemotherapy-treated patients have died, 24 from progressive breast cancer and 4 from intercurrent disease. If patients who died with intercurrent disease were censored at the time of death, the overall three- and five-year survival rates were 65 and 55% respectively for the FAC-treated group. Median survival time is estimated to be 66 months. Three- and five-year survival rates for the control group were 47 and 27% respectively, with a median of 30 months (p - 0.0 1) (Fig. 4). Tolerance to treatment was good in both groups of patients. Although nausea, vomiting, and alopecia were present in virtually all FAC-treated patients (Table 7), the dose-limiting toxicity was myelosuppression. Neutropenia was predictable, the lowest granulocyte count occurred between days 10 and 14, and recovery was
42 . Control pnO.0 1
I
I
I
24
48
72
I 96
Montha Fig. 4. Survival control patients.
from time of diagnosis of FAC-treated
and
May 1983, Volume 9, Number
5
Table 7. Nonhematologic Nausea, vomiting Alopecia Stomatitis Diarrhea Adriamycin infiltration Radiation recall Congestive heart failure Excessive BCC reactions Fever unknown origin Urinary tract infection Pneumonia
toxicity all all 7/52 3152 l/52 1152 4152 4152 5152 i/52 2152
prompt. Thrombocytopenia was of no clinical significance. Myelosuppressive toxicity was similar to that described in an earlier FAC-BCG program.6 Following completion of radiotherapy, a moderate increase in hematologic toxicity occurred in all patients, and a 20 to 40% dose reduction was necessary for subsequent courses of chemotherapy. Infectious complications included pneumonia in two patients, urinary tract infection in one, and fever of undetermined origin in five others during episodes of neutropenia. No deaths related to infection occurred. Four patients suffered congestive heart failure during or shortly after treatment with the Adriamycin-containing combination. Digitalis and diuretics readily controlled the clinical manifestations in three patients, but one died of uncontrolled congestive heart failure. Side effects related to BCG were mild and consisted of local soreness, pruritis, low grade fever, and an ill-defined flulike syndrome for 12 to 24 hours following each scarification. Postradiation fibrosis and apical lung changes frequently were seen on chest X rays following local treatment. Additive, or synergistic, toxicity of chemotherapy and radiotherapy developed in one patient in whom radiation recall appeared over the chest wall when chemotherapy was restarted. DISCUSSION Combination chemotherapy was effective in producing objective tumor regression in this group of patients with locally and regionally advanced breast cancer. Despite the large tumor volume to be treated, 84% of our patients achieved complete or partial response after three courses of chemotherapy and 94% were rendered free of disease with subsequent local therapy. Because of this favorable response to systemic therapy, sufficient tumor reduction was achieved in 43 patients for them to be technically operable. In fact, no residual tumor was found in two mastectomy specimens. This degree of tumor reduction in most cases permitted less radical forms of surgery, lower doses of radiotherapy, and in 35% of patients no mastectomy at all was necessary and only moderate doses of radiotherapy given. This multimodal approach significantly prolonged disease-free survival time, especially in the group of patients with nodal stage N3. These results
Treatment
of locorcgionally
advanced breast cancer 0 G. N. HORTOBAGYIet al.
are especially interesting in view of the poor complicance with systemic therapy. Analysis of our results suggested several means to improve therapy and, therefore, the prognosis for this group of patients. Patients in whom systemic therapy was resumed soon after conclusion of local treatment remained disease-free longer than those who started after a longer delay. Furthermore, the disease-free survival period was better for both groups of patients than for those who refused systemic therapy after local treatment, suggesting that early adjuvant therapy may prolong disease-free survival. Age, race, and menopausal status had no significant influence on outcome. The predominance of hormonedependent tumors would explain why primary tumors grow to an advanced stage without clinical evidence of distant metastasis, but the unusual distribution needs to be confirmed in a larger patient population. However, this initial lead suggests that the addition of hormonal therapy at some stage of multimodal treatment might be an appropriate strategy. The best combination and sequence of therapies to use at this stage of the disease still is not known and further studies are needed. Only 14 of the 52 patients entered in this study completed the two-year treatment program as planned. Unnecessary delays were mainly because of poor patient compliance. in retrospect, this should not be surprising since these patients had ignored the initial diagnosis and many months passed before they sought medical care for their primary breast cancer. Awareness of this personality trait should lead to closer collaboration between the various specialists and the establishment of strong reinforcement techniques to better motivate these patients. Careful and timely scheduling of surgery and radiotherapy might prevent the delays disclosed in our study. Since this series is being compared to historical controls, it should be pointed out that staging for some of the control patients did not include routine bone and liver scanning; therefore, some patients with occult metastatic disease may have been included in this group. This could have conferred a poorer prognosis on the control group. Systemic treatment after relapse was different in the control and FAC-treated gropus, making the comparison of overall survivals difficult. Serrou ef al.” reported on the prolongation of disease-
649
free and overall survival of patients with T,T, breast cancer treated with radiotherapy followed by adjuvant chemotherapy. Morris ef aL9 recently reported three patients with locally advanced breast cancer who were treated with combination chemotherapy followed by cytoreductive or “debulking” surgery. in all three patients, systemic treatment clearly made local therapy easier and less radical. Sponzo et ~1.‘~achieved complete remission in 11 of 12 such patients treated with radiotherapy and combination chemotherapy consisting of cyclophosphamide, 5-fluorouracil. and prednisone. Our own experience with inflammatory breast cancer treated with a similar regimen shows a modest prolongation of disease-free and overall survival. More important, an increase in long-term disease-free survival was observed in postmenopausal patients.*
in a multimodal program almost identical to the one described in this paper, De Lena et af.* found that following the initial cycles of chemotherapy and local treatment, 83% of patients had achieved complete remission. in those who received no maintenance therapy, the median duration of such remission was 11 months, while in those who did receive therapy this interval was 19 months. This difference was highly significant. Whether maintenance chemotherapy reduces the incidence of metastases or only delays their appearance has not been established. The reportsby Serrou et al.” and by De Lena ef al.’ as well as our findings clearly point to the usefuiness of systemic maintenance therapy. A longer follow-up of these patients will be necessary to determine whether the length of survival has increased. Clearly, local therapy alone is inadequate, and advanced primary breast cancers remain a challenge to the oncologists. Improvement in local control may be achieved by combination surgery and radiotherapy,‘-’ combination radiotherapy with hyperthermia or hypoxic sensitizers, or by improving the efficacy of systemic therapy. The incorporation of hormonal therapy into multimodality treatment may be a valuable, relatively nontoxic addition. While optimal local control continues to be a major goal of treatment, methods of controlling metastatic disease should have the highest priority. Only carefully designed clinical trials will define the optimal sequence of the various modalities of treatment in the management of patients with advanced breast cancer.
REFERENCES Brown, G.R., Horiot, J-C, Fletcher, G.H., White, E.C., Ange, D.W.: Simple mastectomy and radiationtherapy for locally advanced breast cancers technically suitable for radical mastectomy. Am. J. Roentgenol. 120: 67-73, 1974. De Lena, M., Zucali, R., Viganotti, G., Valagussa, P., and Bonadonna, G.: Combined chemotherapy-radiotherapy approach in locally advanced ( TJh-T,) breast cancer. Cuncer Chemofher. Pharmacol. 1: 53-59, 1978. Fletcher, G.H., Montague, E.D.: Radical irradiation of advanced breast cancer. Am. J. Roentgenol. 93: 573-584. 1965.
Haagensen, C.D.: The clinical classification of carcinoma and the choice of treatment. In Discuses ofthe Breosf, 2nd edition, C.D. Haagenson (Ed.). Philadelphia, ders Co. I97 1, pp. 617-668.
W.B. Saun-
Henderson, l.c., Canellos, G.P.: Cancer of the breast: the past decade. N. Engl. J. Med. 302: 17-30, 1980. Hortobagyi, G.N., Gutterman, J.U., Blumenschein, G.R., Tashima, C.K., Burgess, M.A., Einhorn, L., Buzdar, A.U., Richman, S.P., Hersh, E.M.: Combination chemoimmunotherapy and radiation therapy for locally advanced breast
650
7.
8.
9.
10.
11.
Radiation Oncology 0 Biology??Physics
cancers technically suitable for radical mastectomy. Cancer 43: 1225-1233, 1979. Kaplan, E.L., Meier. P.: Nonparametric estimation from incomplete observations. J. Amer. Stat. Assoc. 53: 457481, 1958. Krutchik, A.N., Buzdar. A.U., Blumenschein, G.R., Hortobagyi, G.N., Tashima, C.K., Gutterman, J.U., Yap, H-Y, Hersh, E.M.: Combined chemoimmunotherapy and radiation therapy of inflammatory breast carcinoma. J. Surg. Oncoi. I I: 325-332.1979. Morris, D., Aisner, J., Elias, E.G., Wiernik, P.H.: Mastectomy as an adjunct to combination chemotherapy. Arch. Surg. 113: 282-284, 1978. Rubens, R.D., Armitage, P., Winter, P.J., Tong, D.. Hayward, J.L.: Prognosis in inoperable stage III carcinoma of the breast. Europ. J. Cancer 13: 805-811, 1977. Serrou, B., Sancho-Gamier, H., Cappelaere, P., Plagne, R.,
May 1983, Volume 9, Number 5
Metz, R., Schneider, M... Chollet, P., Namer, H., Pujol. H., Gary-Bobo, J., Meyer, G., Mathe, G.: Results of a randomized trial of prophylactic chemotherapy in T,-T, breast cancer patients previously treated by radiotherapy. In Adjuvant Therapies and Markers of Post-surgical Minimal Residual Disease Il. Adjuvant Therapies of the Various Primary Tumors, G. Bonadonna, G. Math&, and S.E. Salmon (Eds.). New York, Springer-Verlag. 1979, pp. 105-108. 12. Sponzo. R.W., Cunningham, T.J., Caradonna, R.R.: Management of non-resectable (Stage III) breast cancer. ht. J. Radiat. Oncol. Biol. Phys. 5: 1415-1478, 1979. 13. Zucali. R., Uslenghi, C., Kenda. R., Bonadonna, G.: Natural history and survival of inoperable breast cancer treated with radiotherapy and radiotherapy followed by radical mastectomy. Cancer 31: 1422-l 43 1, 1976.
03sCr3016/S3/050643~S03.00/0 Copyqht Q 1983 t’ergamm Press Ltd.
9. pp. 643-650
??Original Contribution
TREATMENT OF LOCOREGIONALLY ADVANCED BREAST CANCER WITH SURGERY, RADIOTHERAPY, AND COMBINATION CHEMOIMMUNOTHERAPY G. N. HORTOBAGYI,M.D.,* W. SPANOS, M.D.?E.D. MONTAGUE, M.D.,?A. BUZDAR, M.D.,* H.-Y.YAP, M.D.* AND G. R. BLUMENSCHEIN,M.D.*
U.
The University of Texas M. D. Anderson Hospital and 6723 Bertner Avenue,
Tumor Institute at Houston, Houston, TX 77030
Fifty-two patients witb locally advanced primary breast cancer (T,, T,, N2, N,) but MI evidenceof distant metastases were treated witb tbree cycles of combination cbemotberapy. The regimen consisted of !Muorouracil, Adriamycin, cyclophosplmmide,and Bacillus Calmette-Cuerin (FAC-BCG), followed by local tberapy (simple mastectomy and/or radiofberapy to tk breast/chest wall and tbe regioaal lymphatic system) and adjuvant chemotherapy for two full years. The rem&s were compared with hose in an bistorkal control group of 52 patients ma&bed for initial stage of disease wk were treated by a simple mastectomy and postoperative radiotbempy only. Forty-nine (94% ) of 52 FAGtreated patients and 48 (92 %) of tbe control patients became free of clinically detectabie disease. At tbe median follow-up time of 56 months, 37.5 % of the FAGtreated patients and 19.5 W of tbe control patients bad remained free of disease. FAGtreated patients wbo completed 2 years of tberapy and in whom ?? djuvant cbemotbempy was started promptly after bcal treatment bad a 48% disease-free sunlivnl rate of 4 years. In tbose in whom tbe initial nmaifestathm was supmcbvkuhr idvolvement,tbe estimated S-year disease-free survival rate was 42% for patieuts treated with FAC amd 9% for coatroi patients. There were local recurrences in 25% of FAGtreated patiatts and 23 W of control patients (not sigttiftcant). LXstant metastasesdevelopedin 50 % of FAGtreated patients and 77 % of coatrol patients (p c 0.01). Tbe medirn dii free interval was 2S months in tbe FAGtreated group and 11 months in the control group (p - 0.025). Tbe greatest improvement in prognosis was in patients witb supmchvkul8r involvement;tbe median diifree survival was 26 months in FAC-treated patients and 6 months in tbe control group (p = 0.007). This multimodal approach et&&rely readers tbe majority of patients ivitb locoregionally advanced breast cancer free of disease and prolongs the d&ease-free survival period. Breast cancer, Combined modality treatment, 7’,-T, breast cancer, Cbemotberapy, Surgery, Radiotbempy.
INTROdUCTION In the majority of patients with primary breast cancer, no evidence of distant metastases is discernible and most have tumors that are potentially curable by surgery. A small proportion of patients, however, have large primary tumors involving the skin or deeper structures and, in some cases, large regional metastases are present (i.e., matted and fixed axillary nodes or subciavicuiar or supraciavicuiar metastases). These tumors conform to the T,, T4, IV,, IV3subgroups of the American Joint Commission’s TNM classification. Most women with T,, TI tumors but without sizabie axiiiary or supraciavicuiar metastases technically could be suitable candidates for mastectomy; however, the experience of most investigators suggests that radical excision of these tumors provides no iongterm benefits and is soon followed by multiple early
recurrences.” Patients with N, tumors are not, of course, surgical candidates. For these reasons, the mainstay of therapy for this subgroup of patients had been radiotheraPY.3.4.‘3While improvements in radiotherapy techniques have enhanced local control, many patients still die of distant metastases. In most patients the disease-free interval is brief, and the doses of radiotherapy required for local control produce significant side effects. Combination chemotherapy in breast cancer has become increasingly effective in recent years, a fact affirmed by the response to treatment in 50 to 80% of patients. 5*6Since 1974 we have treated patients in the T,, T,-IV3 subgroups with combination chemotherapy prior to definitive local treatment in an attempt to decrease the need for radical surgery or the radical radiotherapy doses necessary for adequate control. The results of this muitimodal program are defined in this paper and are com-
Presented in part at the American Radium Society Meeting, Philadelphia, Pennsylvania, April 26-May 2, 1980. *Medical Breast Service, Department of Internal Medicine. tDepartment of Radiotherapy.
Reprint requests to: G. N. Hortobagyi, M.D. Acknowfedgmenf-The authors would like to thank Joan Trammell for her assistance in the preparation of this manuscript. Accepted for publication IO December 1982. 643
Radiation Oncology 0 Biology0 Physics
644
pared to our historical experience with local treatment alone.
May 1983, Volume 9. Number 5
Table 2. Advanced primary breast cancer: Treatment program Surgery FACx3 1
METHODS AND MATERIALS Eligibility criteria
Treatment plan
The initial treatment regimen consisted of three cycles of chemotherapy with 5-fluorouracii, Adriamycin, and cyclophosphamide along with nonspecific immunotherapy with Bacillus-Calmette-Guerin (FAC-BCG) at the doses and schedule shown in Table 2. The patients then were reevaluated and, depending on their response status after chemotherapy, local treatment was planned. A tumor board consisting of a medical oncologist, surgical oncologist, and radiotherapist made the decisions about local therapy. An attempt was made to schedule surgery approximately
4to 14 weeks
9 weeks
Breast cancer patients with locally advanced primary tumors classified as T,, T,, iV,, iV3 were included in this trial. All tumors were graded according to the American Joint Commission modification of the TNM classification. Patients with advanced primary tumors amenable to complete surgical resection and patients with inflammatory carcinoma were not entered in this study. Patients with overt congestive heart failure dr uncontrolled hypertension or those otherwise ineligible for chemotherapy were not included. Patients were not excluded for reasons of age or performance status. Patients with distant metastases at the time of diagnosis were excluded. Fifty-nine eligible patients were entered in this program between 1974 and 1976. Seven patients were considered inevaluable because they did not receive local therapy as dictated by protocol. One achieved a complete remission after 8 cycles of FAC, had no local therapy, and remained free of disease 20 months after diagnosis; four had local therapy only after 6 to 11 courses of chemotherapy (2 of them after evidence of local progression), and all of them died 19 to 24 months after diagnosis; and 2 were lost to follow-up after the first dose of chemotherapy. Fifty-two patients completed at least ihe first three courses of chemotherapy and local therapy with surgery, radiotherapy, or both. The distribution of these patients and thiir matched controls by TNM subcategories are shown in Table 1.
three
weeks following
the third
Radiotherapy -
FACx6-CMk
81 to91 weeks
2 years
F>C S-fluorouracil Adriamycin Cyclophosphamide Immunotherapy with BCG
> 500 mg/m* IV D I + 8 50 mg/m* IV D I 500 mg/m* IV D I 6 x IO8 units by scarification on days 9, 13, and I7
CMF
Cyclophosphamide Methotrexate S-fluorouracil Immunotherapy as with FAC
500 mg/m* PO D 2 I + 8 30mg/m’IMD 500 mg/m* PO D I + 8
Duration of cycles, FAC or CMF: 21 days.
chemotherapy and to start radiotherapy two to three weeks after surgery or three weeks after the last dose of chemotherapy. Chemotherapy and radiotherapy were not administered simultaneously. A tumor dose of 5,000 rad was administered with a Cobalt 60 machine to the chest wall or breast and the internal mammary, supraclavicular. and axillary areas. The area of residual tumor was given an additional 1,500 to 2,000 rad to a restricted field in the manner described by Brown ef al.’ In patients with bulky disease after initial chemotherapy and who received radiotherapy only, fractionation was employed twice daily to a dose of 5,500 rad to the breast over five weeks with a boost to areas of residual disease. Whenever possible, the electron beam was used for treatment of internal mammary nodes to lessen the risk of cardiotoxicity. Twenty-five patients were given local treatment consisting of surgery or radiotherapy (Table 3), and 27 patients were treated with both modalities. Seven patients did not receive radiotherapy. Of these seven patients, the mastectomy specimen in two contained no evidence of Table 3. Local therapy
cycle of No. free
Table 1. Distribution of FAC-treated TNM subgroups
patients by
Rsponse to initial FAC
T*
0 2 21 25
4 6 23 31
10 8 27 45
14 16 71 100
Note: Controls were matched by TNM subgroup. Numbers represent percentage of patients in each subcategory.
treatment
No. of patients
Surgery Radiotherapy (RT) Surgery + RT
2 5
(35 patients)
Surgery RT Surgery + RT
5 12 18
Stable disease (9 patients)
RT Surgery + RT
I 8
Complete remission
T, T, Total
Type of local
(8 patients) Partial remission
I
di&e 1
3
Treatment of locorepionally
advanced breast cancer 0 G. N. HORTOBAGYIet al.
tumor; in one, radiation necrosis of the thin skin Rap was being avoided; in another one mastectomy was on the same side as the only lung (agenesis); and one patient with concurrent stomach cancer became lost to follow-up after mastectomy. Two other patients with advanced bilateral primary breast cancer were staged on each side, but in this report they are considered as one patient each, counted by the more advanced of the two cancers. Three weeks after completion of radiotherapy or two weeks after surgery, chemotherapy with FAC-BCG was reinstituted until a total cumulative dose of 450 mg/m2 of Adriamycin was reached. Adriamycin was then discontinued and maintenance chemotherapy with cyclophosphamide, methotrexate, and Huorouracil (CMF) plus BCG was started at the doses and schedule shown in Table 1. The planned duration of therapy, systemic and local, was 24 months. All patients initially were staged with a complete history, physical examination, hemoglobin, hematocrit, white blood cell count with differential and platelet counts, SMA-12/100, carcinoembryonic antigen (CEA), chest Xray, metastatic bone survey and/or bone scan, and liver scan. Bilateral xeromammograms also were performed. Peripheral hematologic values were obtained weekly during chemotherapy and radiotherapy, and the doses of chemotherapy were adjusted to the degree of myeiosuppression as described in an earlier report.6 Tumor measurements were monitored every three weeks until a disease-free status was reached. A complete physical examination and metastatic survey were done at three to four month intervals during therapy and every four to six months thereafter. Before entering a patient in the trial, histologic proof of malignancy and recurrent disease was obtained if possible. A complete remission was defined as disappearance of ail subjective and objective evidence of tumor. A partial remission represented a ~50% reduction of the product of the largest perpendicular diameters of measurable lesions without the development of new lesions. Patients with ~50% decrease or ~25% increase in tumor measurement were considered to have stable disease. Ail other patients had progressive disease. Both palpation and mammographic measurements were used to determine responses.
Control group An historical control group of patients was selected from the files of the Department of Radiotherapy by matching initial stage of disease for each treated patient. Three of them had bilateral breast cancer and were treated in the same manner as the two in the chemotherapy group with this involvement. The diagnosis of cancer had been made in two patients in the control group (both NJ in 1959 and in 13 between 1960 and 1970 (9 with IV,). The rest had been diagnosed and treated between 1970 and 1975. Staging workup for these latter patients was similar to that employed in the chemotherapy-treated
645
group. Radionuclide scans were not performed in the patients seen prior to 1970. A simple mastectomy, with resection of gross axillary disease, was performed in all patients in the control group followed by radiotherapy to the chest wall and peripheral lymphatic nodes, as was done in those in the chemotherapy group with this type of involvement. Treatment after recurrence or the appearance of distant metastases in the chemotherapy-treated group consisted of various hormonal manipulations, chemotherapy, or local salvage treatment (surgery or radiotherapy). Patients in the control group received mainly single-agent chemotherapy or hormonal therapy. Only 12 of these latter patients were treated with an Adriamycin-containing combination, while two others were given cyciophosphamide, methotrexate, and 5-fluorouracil. Disease-free survival time was calculated from the date of mastectomy or the date of complete remission. Survival was calculated from the date of diagnosis for both groups of patients. The Kaplan and Meier method was used to calculate and to plot disease-free intervals and survival curves,’ and a generalized Wilcoxon test (two-tailed analysis) was used to test differences between curves.” RESULTS After three courses of chemotherapy, 8 patients achieved complete remission (CR), 35 patients partial remission (PR), and 9 patients remained stable. Progressive disease was not detected in any of the 52 patients during this treatment period. After the initial three cycles of chemotherapy and local treatment, 49 of 52 patients (94%) were rendered free of disease. Of the three who did not respond, two had large supraclavicular nodes and one had fixed axillary nodes. Forty-eight of 52 control patients (92%) were rendered free of clinical disease after simple mastectomy and radiotherapy. The initial finding in ail four who did not respond was supraclavicular nodes. In general, response to chemotherapy was prompt. The median time to achieve a partial remission was six weeks. The pretreatment status of both groups of patients is shown in Table 4. The age and menopausal status distribution were similar in both groups. There was a higher proportion of blacks and Latin Americans in the group with locally advanced breast cancer than is the case in the general breast cancer population referred to our institution. Similarly, in the few patients for whom estrogen receptor results were available, a predominance of estrogen receptor-positive tumors was apparent. Recurrences
The median follow-up time for the chemotherapytreated patients was 56 months (range, 11 to 90) and for the control group 68 months (range, 55 to 124). In 32 of the 52 (62%) chemotherapy-treated patients disease reappeared; local recurrence occurred in 6, distant metastases in 19, and both types in 7. Local recurrence deveioped in 2 of 7 patients treated with FAC and surgery, 4 of
Radiation Oncology
646
0 Biology 0 Physics
May 1983, Volume 9, Number
5
Dbeue-FraoWvtvd Table 4. Characteristics at diagnosis Chemotherapy group Age (median (range)
Control group
Total Poll 82 82
(2Qj-578)
Menopausal status Premenopausal Postmenopausal
21 31
Estrogen receptor + Unknown
12 4 36
Race White Black Latin American
32 IO 10
32 . FAG 42 . ComfOl
I D~~uo-F~~
(29?9)
0-0.026
14 38 Not available Not available 52
(14)
32 14 6
18 treated with FAC and radiotherapy, and 5 of 21 patients treated with FAC, surgery, and radiotherapy (Table 5). Four in whom disease again recurred achieved a disease-free state following surgery. On the other hand, locoregional recurrences developed in 12 of 52 (23%) control patients during their clinical course. Seven of 52 control patients (13%) and none of the FAC-treated patients had recurrences in the supraclavicular area. Distant metastases developed in 26 (50%) FACtreated patients and.in 40 (77%) control patients. There was no difference in the distribution of metastatic sites between the FAC-treated and control groups except for a lower incidence of pulmonary metastases and supraclavicuiar recurrences in the former group. The incidence of second primary breast cancer was similar in both groups. The median disease-free interval was 25 months for FAGtreated patients and 11 months for control patients (p = 0.025) (Fig. 1). The prognostic factors used to analyze the rate of failure are listed in Table 6. Twenty of 31 (64%) postmenopausal patients and 11 of 21 (52%) premenopausal patients in the FAC-treated group had progression of disease. The respective figures in the control group were 76% and 71%. The differences between premenopausal
000’ 0
48
24
72
Fig. I. Disease-free survival of FAGtreated control groups.
and historical
and postmenopausal groups were not significant. When the data on patients in both the FAC-treated and control groups were analyzed by age groups, the relapse rates were found to be similarly distributed. Race had no influence on relapse rate. Estrogen receptor information was available in only 16 FAC-treated patients and none of the controls. All four patients who had estrogen receptornegative tumors suffered a relapse. At five years the estimated relapse rate was 91% in control patients who initially were found to have a supraclavicular node (NJ and 7 1% in patients without supraclavicular involvement (Figs. 2 and 3). Conversely, in the group that received chemotherapy, the five-year estimated relapse rate was 57% for patients who had supraclavicular involvement and 77% for those who did not. These results represent a
Table 5. Treatment failures Control group No. of relapses Local and regional
Local and regional only
Chest wall Axilla Supraclavicular
7 2 7
10 (19%)
Distant only
30 (58%)
Second primary breast cancer
Total no. patients with relapse I l/52 (21%)
Chemotherapy group No. of relapses 13
Total no. patients with relapse 13/52 (25%)
0 6 (12%) 7 (13%) 19 (37%)
2 (3.8%)
Local and distant
w
l/52 (1.9%)
2/52 (3.8%)
pue
(IN) I#!M sluayed
II!aaueqdwo:,
Ioword
pazkleue
aM ‘dnoA? [oiluo~ aql u! lolDe3 ysou8old ~ueliodru! lsoul aql aq 01 paieadde luaruaA[oau! lepou leyu! amg .aseas!p 30 ama.mmai paaua!ladxa (go9) sluaged oz asaql 30 ZI pue ‘Ldelaql pauueld +aql ala[dt_um lou p!p oqm swayed 103 sqluow 11 seM po!lad IeAglns
aaq-aseas!p
uegxmx aqL ‘(y&5)
asdelal e paiagns
ruaql
Jo 61 pue SqluouJ sz SF?M swayed asaql 10~ Iehlalu! aalj -asrras!p ue!paru aqL .asdelar [gun lualulearl JO AdeJaql [mo[ pue c+ualsdS 30 sJe!arComi palaldum sluaged ZE .+x0 *(aum.Ju%!s Alleyls!~els lou) SoE 30 ~e+.ms JeaL -in03 e peq iah sm Aelap aqi woqm u! asoqL ‘%gt 30 [e!yms leakno e peq luaruleall IeaoI laye &dluold palms SzM f>38-3VJ 30 aSin qllnOJ aql LUOqM u! sluaged *ICdeJaqlsyal& 30 uoydrunsaJ aql papazald sqaam 62 01 L uaamlaq &lap ale!idolddeu! sluaged lsou~ III ‘luawleail leaoo( ial3e 6lldruoid pawnsal Lde -JaqlouIaq3 SBM sluaged pal~!all-f)3$~-3~.~ p 1 QO UI ~sqlu0t.u z 1 pue L le la3um lseaiq a+sal%rd 30 pa!p Adwaql [emI pua /(deJaqlouraqD 30 sasmoa aalql alaldum e aAa!qae IOU leyu! aql %u!Mo~]o~ uo!ssyal p!p oqm swayed aaiql aql30 0mL ‘(SN = d) r(lar\!lmdsal ‘sqwow 81 pue ‘0~ ‘sz aJaM asuodsa3 afqels JO ‘leg -led ‘ala[dum e paAa!qae oqm swayed lo3 spo!lad [e+ -ins say-ase?s!p ueyam aqL .pawnleAa sv~ 0sIe aseas!p 30 aal %u!u!ewaJ 30 hyqeqold aql uo AdelaqlouxaqD 30 sapki aalqi le!l!u! aql 01 asuodsar 30 aauanyu! aqL .iCla%ins01 lo!ld (cg& sJoum1 E-f~ ql!m paluasaid oqm asoqi u! ueql s!sou%wp 30 awl aql ie Jouml “J e ql!m
.uoyenieAa leyu! jo auy aql 11?sap011leIns!Aeplndns lnoql!m sluayid Io1luoa pue palea4l-~y~ Jo leA!hlns aaij-aseas!a 'E f&d
-
0
‘Ilop
-en[eAa re!l!u!30 ayi aqlle sapou JE[na!At?13eidns ql!M sluayed [OnUO:, pue paviaJi-yf~ Jo pzA!Alns aay-aseas!a '2 +!d WuW
@C
1_
‘TTY”’
sluaged u! (o~>zE)lawd seM a3uammal leaol ‘Jalramoq ‘dnol% palearl bdeiaqlomaqa aql UI ‘(9 alqel) dnol% loilum aql u! awl aauaiinaai 1mo1 qly alelaiJoa lou p!p lseaiq aql30 luawar\loy ioum dleuyd 30 iuaixa aqL .( E %J) (9L.o - d) luvX@!s lou se* (sqluoru L 1 ‘umpaur) dnol% low03 agztdsal I!aqt purz (sqwom EZ ‘ue!mu) Ed lnoql!m swayed palEaJl-ddEJaqiOwaq3 aql30 ~e++.ms aai3-aseas!p aqi uaamlaq aaua3aJ!p aqL ‘(~ofj*o - d fsqluolu 9 ‘ue!paur) uogelndod jorlum aql ql!m paieduroa (sqluow sz ‘uegm~.~) dnoi% f~ paleaq-/(delaqlousaqD aql Jo3 am!1 le+uns aaJ3-aseas!p u! aseamy alqeJap!suoa
BZE %I!
%PZ o%OZ
‘L t-1
L
1ualxalotunL
C-I
auop 1ON auop ION
N rN
lUaWaA[OAU! [EpN aA!le%aN aA!l!SOd
loldaaal ua%oiisg lesnedouaumod [esnedouatuald dnoL3 ideJaqlouIaq3
dnoh? [OJIUO~ SJOl3EJ3!lSoU%Old ‘9 a1qe.J.
LV9
648
Radiation Oncology 0 Biology 0 Physics
without (N,_,) initial supraclavicular involvement. Twenty of 24 (83%) N, patients completed two years of therapy or received therapy until relapse, while only 12 of 29 (42%) N,_, patients did so. Furthermore, 10 of 24 (42%) N, patients were restarted on chemotherapy promptly after local treatment compared with only 4 of 28 (14%) N,_* patients who completed their therapy according to the planned program. Poor patient compliance was unexpectedly frequent in this group of patients in contrast with a drop-out rate of 5% or less in our experience with patients with metastatic breast cancer. Compliance with local treatment was excellent. Survival
Twenty-eight of 52 (54%) chemotherapy-treated patients have died, 24 from progressive breast cancer and 4 from intercurrent disease. If patients who died with intercurrent disease were censored at the time of death, the overall three- and five-year survival rates were 65 and 55% respectively for the FAC-treated group. Median survival time is estimated to be 66 months. Three- and five-year survival rates for the control group were 47 and 27% respectively, with a median of 30 months (p - 0.0 1) (Fig. 4). Tolerance to treatment was good in both groups of patients. Although nausea, vomiting, and alopecia were present in virtually all FAC-treated patients (Table 7), the dose-limiting toxicity was myelosuppression. Neutropenia was predictable, the lowest granulocyte count occurred between days 10 and 14, and recovery was
42 . Control pnO.0 1
I
I
I
24
48
72
I 96
Montha Fig. 4. Survival control patients.
from time of diagnosis of FAC-treated
and
May 1983, Volume 9, Number
5
Table 7. Nonhematologic Nausea, vomiting Alopecia Stomatitis Diarrhea Adriamycin infiltration Radiation recall Congestive heart failure Excessive BCC reactions Fever unknown origin Urinary tract infection Pneumonia
toxicity all all 7/52 3152 l/52 1152 4152 4152 5152 i/52 2152
prompt. Thrombocytopenia was of no clinical significance. Myelosuppressive toxicity was similar to that described in an earlier FAC-BCG program.6 Following completion of radiotherapy, a moderate increase in hematologic toxicity occurred in all patients, and a 20 to 40% dose reduction was necessary for subsequent courses of chemotherapy. Infectious complications included pneumonia in two patients, urinary tract infection in one, and fever of undetermined origin in five others during episodes of neutropenia. No deaths related to infection occurred. Four patients suffered congestive heart failure during or shortly after treatment with the Adriamycin-containing combination. Digitalis and diuretics readily controlled the clinical manifestations in three patients, but one died of uncontrolled congestive heart failure. Side effects related to BCG were mild and consisted of local soreness, pruritis, low grade fever, and an ill-defined flulike syndrome for 12 to 24 hours following each scarification. Postradiation fibrosis and apical lung changes frequently were seen on chest X rays following local treatment. Additive, or synergistic, toxicity of chemotherapy and radiotherapy developed in one patient in whom radiation recall appeared over the chest wall when chemotherapy was restarted. DISCUSSION Combination chemotherapy was effective in producing objective tumor regression in this group of patients with locally and regionally advanced breast cancer. Despite the large tumor volume to be treated, 84% of our patients achieved complete or partial response after three courses of chemotherapy and 94% were rendered free of disease with subsequent local therapy. Because of this favorable response to systemic therapy, sufficient tumor reduction was achieved in 43 patients for them to be technically operable. In fact, no residual tumor was found in two mastectomy specimens. This degree of tumor reduction in most cases permitted less radical forms of surgery, lower doses of radiotherapy, and in 35% of patients no mastectomy at all was necessary and only moderate doses of radiotherapy given. This multimodal approach significantly prolonged disease-free survival time, especially in the group of patients with nodal stage N3. These results
Treatment
of locorcgionally
advanced breast cancer 0 G. N. HORTOBAGYIet al.
are especially interesting in view of the poor complicance with systemic therapy. Analysis of our results suggested several means to improve therapy and, therefore, the prognosis for this group of patients. Patients in whom systemic therapy was resumed soon after conclusion of local treatment remained disease-free longer than those who started after a longer delay. Furthermore, the disease-free survival period was better for both groups of patients than for those who refused systemic therapy after local treatment, suggesting that early adjuvant therapy may prolong disease-free survival. Age, race, and menopausal status had no significant influence on outcome. The predominance of hormonedependent tumors would explain why primary tumors grow to an advanced stage without clinical evidence of distant metastasis, but the unusual distribution needs to be confirmed in a larger patient population. However, this initial lead suggests that the addition of hormonal therapy at some stage of multimodal treatment might be an appropriate strategy. The best combination and sequence of therapies to use at this stage of the disease still is not known and further studies are needed. Only 14 of the 52 patients entered in this study completed the two-year treatment program as planned. Unnecessary delays were mainly because of poor patient compliance. in retrospect, this should not be surprising since these patients had ignored the initial diagnosis and many months passed before they sought medical care for their primary breast cancer. Awareness of this personality trait should lead to closer collaboration between the various specialists and the establishment of strong reinforcement techniques to better motivate these patients. Careful and timely scheduling of surgery and radiotherapy might prevent the delays disclosed in our study. Since this series is being compared to historical controls, it should be pointed out that staging for some of the control patients did not include routine bone and liver scanning; therefore, some patients with occult metastatic disease may have been included in this group. This could have conferred a poorer prognosis on the control group. Systemic treatment after relapse was different in the control and FAC-treated gropus, making the comparison of overall survivals difficult. Serrou ef al.” reported on the prolongation of disease-
649
free and overall survival of patients with T,T, breast cancer treated with radiotherapy followed by adjuvant chemotherapy. Morris ef aL9 recently reported three patients with locally advanced breast cancer who were treated with combination chemotherapy followed by cytoreductive or “debulking” surgery. in all three patients, systemic treatment clearly made local therapy easier and less radical. Sponzo et ~1.‘~achieved complete remission in 11 of 12 such patients treated with radiotherapy and combination chemotherapy consisting of cyclophosphamide, 5-fluorouracil. and prednisone. Our own experience with inflammatory breast cancer treated with a similar regimen shows a modest prolongation of disease-free and overall survival. More important, an increase in long-term disease-free survival was observed in postmenopausal patients.*
in a multimodal program almost identical to the one described in this paper, De Lena et af.* found that following the initial cycles of chemotherapy and local treatment, 83% of patients had achieved complete remission. in those who received no maintenance therapy, the median duration of such remission was 11 months, while in those who did receive therapy this interval was 19 months. This difference was highly significant. Whether maintenance chemotherapy reduces the incidence of metastases or only delays their appearance has not been established. The reportsby Serrou et al.” and by De Lena ef al.’ as well as our findings clearly point to the usefuiness of systemic maintenance therapy. A longer follow-up of these patients will be necessary to determine whether the length of survival has increased. Clearly, local therapy alone is inadequate, and advanced primary breast cancers remain a challenge to the oncologists. Improvement in local control may be achieved by combination surgery and radiotherapy,‘-’ combination radiotherapy with hyperthermia or hypoxic sensitizers, or by improving the efficacy of systemic therapy. The incorporation of hormonal therapy into multimodality treatment may be a valuable, relatively nontoxic addition. While optimal local control continues to be a major goal of treatment, methods of controlling metastatic disease should have the highest priority. Only carefully designed clinical trials will define the optimal sequence of the various modalities of treatment in the management of patients with advanced breast cancer.
REFERENCES Brown, G.R., Horiot, J-C, Fletcher, G.H., White, E.C., Ange, D.W.: Simple mastectomy and radiationtherapy for locally advanced breast cancers technically suitable for radical mastectomy. Am. J. Roentgenol. 120: 67-73, 1974. De Lena, M., Zucali, R., Viganotti, G., Valagussa, P., and Bonadonna, G.: Combined chemotherapy-radiotherapy approach in locally advanced ( TJh-T,) breast cancer. Cuncer Chemofher. Pharmacol. 1: 53-59, 1978. Fletcher, G.H., Montague, E.D.: Radical irradiation of advanced breast cancer. Am. J. Roentgenol. 93: 573-584. 1965.
Haagensen, C.D.: The clinical classification of carcinoma and the choice of treatment. In Discuses ofthe Breosf, 2nd edition, C.D. Haagenson (Ed.). Philadelphia, ders Co. I97 1, pp. 617-668.
W.B. Saun-
Henderson, l.c., Canellos, G.P.: Cancer of the breast: the past decade. N. Engl. J. Med. 302: 17-30, 1980. Hortobagyi, G.N., Gutterman, J.U., Blumenschein, G.R., Tashima, C.K., Burgess, M.A., Einhorn, L., Buzdar, A.U., Richman, S.P., Hersh, E.M.: Combination chemoimmunotherapy and radiation therapy for locally advanced breast
650
7.
8.
9.
10.
11.
Radiation Oncology 0 Biology??Physics
cancers technically suitable for radical mastectomy. Cancer 43: 1225-1233, 1979. Kaplan, E.L., Meier. P.: Nonparametric estimation from incomplete observations. J. Amer. Stat. Assoc. 53: 457481, 1958. Krutchik, A.N., Buzdar. A.U., Blumenschein, G.R., Hortobagyi, G.N., Tashima, C.K., Gutterman, J.U., Yap, H-Y, Hersh, E.M.: Combined chemoimmunotherapy and radiation therapy of inflammatory breast carcinoma. J. Surg. Oncoi. I I: 325-332.1979. Morris, D., Aisner, J., Elias, E.G., Wiernik, P.H.: Mastectomy as an adjunct to combination chemotherapy. Arch. Surg. 113: 282-284, 1978. Rubens, R.D., Armitage, P., Winter, P.J., Tong, D.. Hayward, J.L.: Prognosis in inoperable stage III carcinoma of the breast. Europ. J. Cancer 13: 805-811, 1977. Serrou, B., Sancho-Gamier, H., Cappelaere, P., Plagne, R.,
May 1983, Volume 9, Number 5
Metz, R., Schneider, M... Chollet, P., Namer, H., Pujol. H., Gary-Bobo, J., Meyer, G., Mathe, G.: Results of a randomized trial of prophylactic chemotherapy in T,-T, breast cancer patients previously treated by radiotherapy. In Adjuvant Therapies and Markers of Post-surgical Minimal Residual Disease Il. Adjuvant Therapies of the Various Primary Tumors, G. Bonadonna, G. Math&, and S.E. Salmon (Eds.). New York, Springer-Verlag. 1979, pp. 105-108. 12. Sponzo. R.W., Cunningham, T.J., Caradonna, R.R.: Management of non-resectable (Stage III) breast cancer. ht. J. Radiat. Oncol. Biol. Phys. 5: 1415-1478, 1979. 13. Zucali. R., Uslenghi, C., Kenda. R., Bonadonna, G.: Natural history and survival of inoperable breast cancer treated with radiotherapy and radiotherapy followed by radical mastectomy. Cancer 31: 1422-l 43 1, 1976.