NeoadjuvantChemotherapy in the CombinedModality Approachof ... · junction with local therapy to treat locally advanced breast cancer. In a retrospective review from the Harvard-Joint

[CANCER RESEARCH 47, 3889-3894, July 15. 1987)

NeoadjuvantChemotherapy in the Combined Modality Approach of LocallyAdvancedNonmetastatic Breast Cancer

Sandra M. Swain, Richard A. Sorace, Caroline S. Bagley, David N. Danforth, Jr., Judith Bader, Margaret N. Wesley,Seth M. Steinberg, and Marc E. Lippman1

Medicine Branch (S. M. S., C. S. B., M. E. L.J, Radiation Oncology Branch Â¡R.A. S., J. B.J, Surgery Branch fD. N. D.J, Bioslatislics and Data Management Section[M. N. W., S. M. S.J, and Clinical Oncology Program. Division of Cancer Treatment, National Cancer Institute, Belhesda, Maryland 20892

ABSTRACT

We have treated 76 patients with locally advanced breast cancer, 31with stage IIIA, 41 with stage 1MB. and 4 with stage IV disease, withprimary induction chemotherapy including an attempted hormonal synchronization in 70 patients. All were treated to maximum objective clinicalresponse before proceeding to any local therapy. Patients achieving acomplete response with a negative repeat biopsy generally receivedradiation therapy while patients with residual disease, partial response(PR) or no change (NC) status received debulking surgery prior toradiation therapy. Regardless of response to induction chemotherapy,patients received at least 6 additional months of chemotherapy followinglocal therapy. Initial doses of combination chemotherapy were escalatedto targeted myelosuppression. The objective response rate to inductionchemotherapy was 93% with 49% complete response (CR), 44% PR, and7% NC. The median numbers of cycles of chemotherapy to achieve aCR, PR, or NC were 5, 3, and 5, respectively. Three patients whocurrently have PRs are still on chemotherapy with continued tumorregression. Of 37 patients achieving a CR to chemotherapy, 35 wereassessed by biopsies to determine pathological evidence of response.Twenty-three of the 37 patients (62%) were proven to be completeresponders with negative biopsies. Twenty-four patients have relapsed, 6

with stage IIIA, 16 with stage HIB, and 2 with stage IV. Five patientshave had locoregional relapses alone, 4 locoregional and distant, and 15distant alone. Median time to progression is 35.9 months for stage IIIAand 34.2 months for stage III IÂ».Median survival is 35.3 months for stageIIIB and is indeterminate for stage IIIA. This aggressive primary chemotherapy regimen with hormonal synchronization followed by localtherapy appears to provide excellent local control and encouraging earlyresults on systemic disease control.

INTRODUCTION

Locally advanced, nonmetastatic breast cancer is associatedwith poor local control and dismal survival. The results ofHaagensen and Stout (1,2) emphasize the poor outlook forthis group of breast cancer patients; they reported that radicalmastectomy failed to cure even a single patient in their seriesof 120 patients followed for up to 8 years. They regarded theselesions as "categorically inoperable" and suggested that they be

treated by other measures. Improved radiotherapy techniqueshave allowed the delivery of higher doses of radiation to thesepatients' tumors with acceptable rates of local complications,

thus increasing chances of achieving and maintaining localcontrol (3). Results with radiotherapy alone have yielded 5-yearlocal control rates of 28 to 74% and 5-year survival in the 12to 38% range (4-13). Combined programs of surgery andradiation have resulted in 5-year survivals of 35 to 55% in someselected series (5-8, 10, 13, 14). Direct comparisons of resultsare difficult because of the wide latitude in prognostic variablesin these series. In spite of some improvement in local control,the majority of patients succumb to metastatic disease, sug-

Received 8/8/86; revised 1/20/87, 4/6/87; accepted 4/7/87.The costs of publication of this article were defrayed in part by the payment

of page charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

' To whom requests for reprints should be addressed, at Medicine Branch.National Cancer Institute, Bldg. 10, Room I2N226, Bethesda, MD 20892.

gesting that disseminated micrometastases exist at initial presentation in virtually all patients.

Therefore investigators began using chemotherapy in conjunction with local therapy to treat locally advanced breastcancer. In a retrospective review from the Harvard-Joint Centerfor Radiation Therapy (7), 41 patients with stage III breastcancer, who received postirradiation adjuvant therapy witheither chemotherapy alone or combined with an endocrine-ablative procedure, had a significantly better relapse free survival and improved local control at 4 years (51 and 85%,respectively) as compared to a matched group of patients notreceiving adjuvant therapy (29 and 63%, respectively). In arandomized prospective study of patients with operable stageIII breast cancer (T3, N0-2), GrÃ¶hnet al. (15) showed improved5-year survival (57 to 73%) and freedom from distant relapse(50% to 87.5%) when patients received 6 cycles of cyclophos-phamide, doxorubicin, and vincristine chemotherapy after amodified radical mastectomy compared to those patients whoreceived only postoperative radiation therapy. Combinationchemotherapy prior to local therapy was introduced by theMilan group (16). In 110 patients, they achieved a 70% objectiveresponse rate (15% CR2) with 4 cycles of AV chemotherapyprior to local therapy and reported an overall 3-year survival of53% compared to 41% for a historical control group receivingradiation alone. In subsequent studies comparing surgery versusradiation after primary chemotherapy, they achieved equivalentfindings with respect to local control, patterns of recurrence,and overall survival (17). A similar treatment strategy was usedin 52 noninflammatory patients at the M. D. Anderson Hospital (18) in which 3 cycles of FAC plus Bacillus Calmette-GuÃ©rinwere administered prior to local therapy with eitherradiation alone or surgery followed by radiation. An objectiveresponse rate to chemotherapy of 82% (15% CR) was achievedwith 40% actuarial 5-year survival. Thus, induction chemotherapy for a fixed number of cycles prior to local therapy resultedin considerable tumor regression that enhanced the ability toeither perform a surgical resection or deliver a tumoricidal doseof radiation in both of these institutional experiences.

In an effort to maximally reduce the tumor bulk prior todefinitive local therapy and to destroy occult micrometastaticdisease, since 1977 the National Cancer Institute has givenwomen with locally advanced breast cancer combination chemotherapy to the point of maximum objective clinical response.Unlike other protocols with a fixed number of initial chemotherapy cycles, each patient received the number of cycles ofchemotherapy needed for a best response. Concomitant endocrine therapy was used with chemotherapy by tamoxifen andpremarin sequentially before the administration of cell cyclespecific agents in the hopes of synchronizing DNA synthesis ina proportion of these tumor cells and thus making them moresusceptible to cytotoxic chemotherapy.

2The abbreviations used are: CR, complete response; PR, partial response;NC, no change; AV, doxorubicin-vincristine; FAC, 5-fluorouracil-doxorubicin-cyclophosphamide; TNM, tumor-nodes-metastasis classification.

3889

on June 27, 2020. © 1987 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

LOCALLY ADVANCED BREAST CANCER

The rationale for this approach has its foundation in priorstudies from our laboratory (19-21) which demonstrated thatanti-estrogen-induced inhibition of breast cancer cell growthcould be reversed by estrogen rescue. Weichselbaum et al. (22)showed that this technique could increase the sensitivity ofbreast cancer cells to cytotoxic drug treatment. Green et al.(23), using a flow cytometry technique, demonstrated that tam-oxifen induced a G, arrest in MCF-7 human breast cancer cellsin culture. Our laboratory has shown that subsequent estrogentreatment could induce a synchronous wave of DNA synthesisin human breast cancer cells (24). Hence, this strategy was usedin the present study prior to the administration of the cell cycle-specific agents methotrexate and 5-fluorouracil, in hope ofsynchronizing a portion of tumor cells and making them moresusceptible to these cytotoxic drugs. Allegra et al. (25) hastreated a small group of breast cancer patients with antiestrogentherapy followed by estrogen rescue plus methotrexate and 5-fluorouracil and has reported a complete remission rate of 56%.In a randomized study from our institution (26) using cyclo-phosphamide-doxorubicin-methotrexate-5-fluorouracil chemotherapy with or without hormonal synchronization in patientswith metastatic and inflammatory breast cancer, improvementin overall survival and time to progression was observed inobjective responders in the treatment arm including tamoxifenand premarin. In this study 13 of 14 patients with inflammatorybreast cancer had an objective response rate of 93%. Of theinflammatory nonmetastatic patients, 4 of 7 patients treated inthe hormonal synchronization arm had a clinical completeresponse versus 1 of 6 who had no hormones included. It wasfor this reason that hormonal synchronization was chosen forthe present study.

MATERIALS AND METHODS

Patient Selection. From January 1977 through October 1986, 76women with previously untreated locally advanced breast cancer weretreated with combined modality therapy. These patients' stages were

defined as stage III A and 1MBaccording to the 1983 American JointCommittee staging system. To be eligible for treatment, all patientswere required to have a histologically documented diagnosis of mammary cancer with evidence of measurable disease. Inflammatory carcinoma was defined by the presence of dermal lymphatic tumor invasionor clinically as signs of inflammatory cancer in greater than one-thirdof the breast. Karnofsky performance index was 90 in 96% of patientswith the remaining ranging from 60 to 80. Patients were excluded whogave a past history of malignant neoplasms aside from curatively treatedbasal cell or squamous cell carcinoma of the skin or surgically curedcarcinoma of the cervix in situ. Patients were required to have aleukocyte count greater than 4,000/mm' and a platelet count greaterthan 100,000/mm3. Similarly, renal and hepatic functions were required

to be within normal limits. All patients gave written informed consentbefore starting treatment. Four patients who met the above criteria butwho were retrospectively found to have had metastatic disease at thetime of entry in the study are included in the analysis.

Initial and Follow-up Evaluations. Prior to treatment all patients hada complete history and physical examination (including calculation ofbody surface area), blood cell count, urinalysis, chemistry profile,electrocardiogram, radionuclide scans of bone and liver, mammograms,and radiographie examinations of the chest and skeleton. Blood countswere followed on a weekly basis. With the institution of each newchemotherapy cycle, chemistry profiles were repeated along with formalquantitation of any palpable tumor burden. Mammograms were performed at best response and annually thereafter. Liver-spleen scans andbone scans were performed every 12 weeks.

Chemotherapy Regimens. The initial 13 patients with locally advanced breast cancer were included as part of a randomized protocoldesigned to prospectively compare two chemotherapy regimens as

primary induction therapy in advanced breast cancer (26). These regimens are shown in Table 1 and termed 160B. Thirteen patients receivedcycIophosphamide-doxorubicin-methotrexate-5-fluorouracil chemotherapy with cyclophosphamide at 750 mg/m2 and methotrexate at 40mg/m2; 6 of these did not receive hormones. This protocol was closed

as a randomized trial in June 1983. Subsequent patients with locallyadvanced breast cancer were accrued to the hormonal arm of the priorrandomized study, with the dose adjustments shown in Table 1 andtermed 160C. In 63 patients, initial chemotherapy doses consisted ofcyclophosphamide (500 mg/m2) and doxorubicin (30 mg/m2 i.V., day1); tamoxifen (40 mg p.o., days 2-6); premarin (0.625 mg p.o. every 12h in 3 doses beginning on day 7); methotrexate (300 mg/m2 i.v. followedin l h by 5-fluorouracil 500 mg/m2 i.v. on day 8) and leucovorin (10mg/m2 p.o. every 6 h for 6 doses beginning 24 h after MTX). A 25%dose escalation of either methotrexate or 5-fluorouracil in alternatecycles was used to achieve maximum doses of these drugs.

All patients were treated with induction therapy to the point ofmaximum objective clinical response as described previously (27). Detailed response criteria used in these studies have been publishedelsewhere (28) (i.e., complete response, partial response, and nochange). Maximum objective clinical response or best response wasformally scored when response parameters remained stable during twoconsecutive determinations separated by 6 weeks. For the purposes ofthe time to best response analysis, the time was taken to be the pointat which the maximum objective clinical response was first observed.Dose modification criteria for individual treatment cycles have beendescribed previously (27).

Local Therapy. After achieving maximum objective clinical responseto chemotherapy, patients were evaluated for local therapy. This wasdone routinely by the same two physicians (S. S. and M. L.) using pre-and postchemotherapy mammograms and physical examinations. If apatient was considered a clinical complete responder, an incisionalbiopsy was taken at the site of the original lesion with a biopsy of anysurrounding suspicious areas seen at the time of the procedure. If skininvolvement was present at diagnosis a skin biopsy was obtained. If anegative biopsy was obtained, the patients received a course of radicalradiotherapy to the intact breast and adjacent draining lymph nodeareas. No axillary surgery was performed. For patients with residualdisease in their biopsy, or patients achieving only partial response, orno change to induction chemotherapy, a total mastectomy was performed, with removal of any gross axillary disease. This was followedby a course of radical radiotherapy. All patients continued to receivechemotherapy with the exclusion of doxorubicin while radiotherapywas administered. For patients on 160B, 6 received chemotherapy for1 year and 7 received chemotherapy for 6 months after local therapywas completed. Complete responders to induction therapy receiveddoxorubicin every other cycle on 160B. All 160C patients receivedchemotherapy for 6 months after local therapy was completed with

Table 1 Treatment regimens160B (first 13 patients were randomly assigned)"

C Cyclophosphamide 750 mg/m2 i.v. day 1 vs. CA Doxorubicin 30 mg/m2 i.v. day l A T Tamoxifen 10

mg p.o. twicea day, days 2-6

M Methotrexate 40 mg/m2 i.v. day 8

F 5-Fluorouracil 500 mg/m2 i.v. day 8

160C (subsequent 63 patients received)C Cyclophosphamide 500 mg/m2 i.v. day 1

M P Premarin 0.652mg p.o. every12 h in 3doses, day 7

A DoxorubicinM MethotrexateF 5-FluorouracilT TamoxifenP Premarin

L Leucovorin

30 mg/m2 i.v. day 1300 mg/m2 i.v. day 8500 mg/m2 i.v. day 8

40 mg p.o. days 2-60.625 mg p.o. every12 h in 3 doses, day 7

10 mg/m2 p.o. every 6

h in 6 doses, day 9Â°One patient received CAMF without being randomized.

3890




doxorubicin included in every cycle. In all cases, total dose of doxorub-icin did not exceed 525 mg/m2.

Patients achieving a complete response received radiation to theintact breast and peripheral lympathic areas via breast tangents (including the internal mammary nodes), a supraclavicular field, and a posterior axillary boost. Patients who had a PR, NC, or residual tumor onrestaging biopsy underwent mastectomy and removal of any grosslyinvolved axillary lymph nodes. That was followed by chest wall tangentsand similar peripheral lymphatic irradiation. Only that portion of theaxilla that was not dissected received radiation. Bolus was used everyother day for all chest wall tangents and for any intact breast whichhad inflammatory involvement at presentation. Dose (and dose perfraction) ranged from 4400 to 5400 cGy to the prescribed 150 dose fortangents and to the supraclavicular field at 3 cm. A boost (to 6000 cGytotal dose) was given to the original tumor volume for intact breasts.Also boosted were areas of gross residual nodal disease and positivechest wall margin following mastectomy.

Statistical Analysis. Time to progression and survival time wereestimated actuarially using the Kaplan-Meier procedure (29).

RESULTS

Distribution of Prognostic Variables. At the time of thisevaluation, 76 patients have been started on treatment and areÃ©valuablefor toxicity, time to progression, and survival. One ofthe 76 patients is a recent entrant and is excluded from responseanalysis. As shown in Table 2, stage III \ and IIIB patientswere comparable with respect to the distribution of estrogenreceptor status and menopausa! status, while stage III A patientswere somewhat younger. The TNM status of patients at presentation is shown in Table 3.

Chemotherapy Response. Overall objective response rate (CR+ PR) was 93%. Thirty-seven patients (49%) achieved a CR,33 patients (44%) achieved a PR, and 5 patients (7%) showedno change. Three of the patients scored as PRs are still oninduction chemotherapy with continued tumor regression andhave the potential for achieving a CR response. Median numbers of chemotherapy cycles to achieve a CR, PR, or NC were5, 3, and 5, respectively, with a range of cycles of 2 to 11, 2 to9, and 4 to 6, respectively. The time course for these patientsachieving their best response to chemotherapy is shown in Fig.1.

Table 4 shows the distribution of various prognostic factors

Table 2 Distribution of prognostic factors

VariableNo.

ofpatientsHistology

InflammatoryNoninflammatoryStage

IIIA31031Estrogen

receptor statusPositive 6Negative 14Unknown 11Menopausal

statusPremenopausal

PostmenopausalMedian

age (yr)Table

3T2

T,T4TotalIS1642.0Stage

IIIB4132

97

1321142753.0Stage

IV Allpatients4

761

333431

142 291331

3034660.7

51.7Clinical

stage of patients atpresentationN.

N,0

011 1091620

26N2

N,2

182

9419

7M

Total0

3233

2404

76 patients

ERCENTATBESTRESPONSE

Ã«Ã«ÃªSSSSSgo-

10-H2/4

6_a-f810TIMEINMONTHSFig.

1. Time from start of therapy to bestresponse.Table

4 Distribution of prognostic factors according to bestchemotherapyresponseBest

clinicalresponseVariableOverall

treatmentTamoxifen+premarinNo

hormonesEstrogen

receptorstatusERCpositiveER

negativeERunknownStageIDAIIIBInflammatory'InflammatoryIVNodal

metastasisN.N,N,N,Menopausal

statusPremenopausalPostmenopausalMedian

age (yr)CR3736

(4)'KD7171316183151372142352.1PR33Â°29(3)4(4)510181113S44IIII7132051.3NC54KD22131112112346.4Tooearly11111035.0

" Three patients are still on induction chemotherapy and may improve further.'' Numbers in parentheses, patients who were randomized on regimen I60B.' ER, estrogen receptor.

according to each chemotherapy response group. There was noindication that estrogen receptor status, histology, nodal status,menopausal status, or age were predictive of a patient's best

clinical response to induction chemotherapy. A higher proportion of CRs were N0 rather than N|-N3. However, eight ofthese patients (T3 or T4, N0) had inflammatory cancer.

Toxicities. The WBC nadir reached 3000-3999 in 13% ofpatients, 2000-2999 in 39%, 1000-1999 in 45%, and below1000 in 3%. There were four infection-related admissions occurring during periods of neutropenia while on chemotherapybut no deaths. Two patients developed cardiomyopathies secondary to doxorubicin. One patient received a total dose of 490mg/m2 of doxorubicin and is doing well with compensated

congestive heart failure. The second patient received 513 mg/m2 of doxorubicin and died of progressive congestive heart

failure with no evidence of breast cancer. There has been oneserious wound infection after local treatment with surgery andradiation, one radiation neuropathy resulting in a nonfunction-ing arm which is improving, and one methotrexate pneumoni-tis. One patient developed a severe rash with blistering skinfrom 5-fluorouracil and this agent was discontinued in thispatient only.

3891




In conclusion, this chemotherapy regimen compares favorably with other chemotherapy regimens in that it was welltolerated by patients with low toxicity rates. It may be thatfurther dose escalations will be warranted.

HistolÃ³gica! Evaluation of Response. Patients achieving aclinical CR were evaluated histologically to determine if a CRwith negative biopsy had been attained (Table 5). Thirty-fivepatients underwent biopsy evaluation in the area of the breastpreviously known to be involved with tumor. Twenty-threepatients had negative biopsies. Two biopsied patients showedresidual microscopic disease, one of which showed a few scattered tumor cells that appeared to be nonviable treated tumorcells. Both of these patients proceeded to local therapy withradiation. The remaining ten biopsied patients had residualgross disease and underwent a simple mastectomy before proceeding to radiation therapy. Two patients went on to radiationtherapy without histological evaluation. This course was pursued in one of these patients because her initial breast tumorwas a T2Nj lesion that was excisionally biopsied before induction chemotherapy; hence she was thought not to have grossdisease in her breast at the start of treatment. The other patientdid not undergo biopsy because of the clinical judgment of themanaging physicians. Patients achieving a PR had total mastectomy in 26 cases. One of the patients achieving a PR (stageIV) had no tumor in her mastectomy specimen. Two patientsachieving a PR did not have mastectomies; one of these patients(stage IV) had a clinical CR in her breast and had no biopsyand the other had a biopsy positive for tumor. Three patientsachieving a PR are still on induction therapy, one patientrefused biopsy, and one is lost to follow-up. Therefore pathology is available on 27 of 33 PR patients at the present time.All 5 patients who were NC had tumor at mastectomy.

In summary, 35 of 37 CR patients, 27 of 33 PR patients, and5 of 5 NC patients for a total of 67 of 76 patients have pathologyavailable after best response. Of these 67, 24 (36%) were pathologically free of disease.

Local Therapy. After achieving the best clinical response toinduction chemotherapy, local therapy was given as detailed inTable 6. All patients achieving a clinical CR received radiationtherapy with the exception of a single patient who refused anylocal therapy after achieving a complete response with negativerepeat biopsy. One patient is still in the process of completingradiation therapy. Of the 38 PR and NC patients, 31 patientshave completed a course of local therapy composed of debulkingsurgery followed by radiation and two patients are still receivingradiation after debulking surgery. Debulking surgery consistedof a total mastectomy in 10 patients, a modified radical mas-

Table5 Histologicalevaluationofpatientsachievinga clinicalCRwithchemotherapy

ProcedureBiopsy

onlyNo biopsyTotalNo.

ofpatients35

237No.

of patientswithout tumor

inspecimen23

023 (62%)No.

of patients withtumor inspecimen12

012(33%)No.

of patients with

biopsy0

22(5%)

Table 6 Local therapy after best chemotherapy response

Chemotherapy No. ofresponsepatientsCR

PRNC37Â°33* 5Mastectomy

only02

0Mastectomy

+radiation10

245Radiation

only26

20

" One patient refused local treatment.* Three on induction chemotherapy: one refused local treatment: one lost to

follow-up.

tectomy in 18 patients, and Halstead radical mastectomy inone patient. Early in the study, two patients including one withinflammatory disease received modified radical mastectomiesalone as local therapy after achieving PR to chemotherapy.Two patients achieving PR to induction therapy received radiation alone. In one case the patient had stage IV disease withbony disease still apparent, but a CR in the breast. The secondpatient had chest wall fixation of her remaining tumor anddebulking surgery was impossible. However, after several morecycles of chemotherapy and radiation, a repeat biopsy of theresidual mass was negative for tumor. One patient achieving aPR to induction therapy was lost to follow-up. One patientachieving a PR refused local therapy. Three patients are currently on induction chemotherapy or are awaiting rebiopsy andhave not proceeded to local therapy.

Outcome of Patients Completing Combined Modality Treatment. Sixty-nine patients have proceeded to the local therapyportion of the program (4 patients are still on chemotherapyand 3 patients refused). Sixty-six of these patients have completed all local therapy (3 patients are still receiving radiationtherapy). All patients except one with chest wall fixation havebeen rendered disease free after completing local therapy bythis combined modality treatment approach. However, thispatient was rendered disease free after several additional cyclesof therapy.

There have been 24 patients (32%) who have relapsed todate. Thus far no predictive value for eventual relapse has beenfound for estrogen receptor status, menopausal status, age,addition of hormonal synchronization, or number of cycles tobest chemotherapy response. The most important predictor ofrelapse was the clinical TNM stage at the start of treatment.The distribution of relapsed patients according to initial TNMstage is illustrated in Table 7. Patients with advanced nodalmetastasis who relapsed tended to have inflammatory breastcancer. Of the 24 relapsed patients, 6 were stage IIIA (19% ofpatients with IIIA disease), 16 were stage IIIB (39% with stageIIIB), and 2 were stage IV (50% with stage IV). Twelve relapsedpatients (38% of IIIB inflammatory) had inflammatory breastcancer.

Table 8 reveals the site of disease at relapse. Five patientsrelapsed locoregionally alone, 4 locoregionally and distant, and15 distant alone. This is an overall 14% incidence of anylocoregional recurrence. Two locoregional relapses were in thebreast parenchyma, one was in the breast skin only, one in anipsilateral axillary node, one in an ipsilateral supraclavicularnode, and 4 chest wall recurrences. At 2 years the probabilityof local control is 83% for all patients (Fig. 2).

There have been 9 patients who were clinical complete re-sponders (24%) that have relapsed. Seven of these were corn-

Table7 Distributionofrelapsesaccordingto initialTNMstageNo. of patients relapsed/total no. of patients with that

stage

T2T,T4

TotalNo0/02/112/94/20N,0/03/10

7/1610/26N20/2

1/84/9

5/19N,1/1

1/21/43/7M01/21/22/4Total1/3

8/3315/4024/76

Table 8 Site of disease at first relapse

Analysis of first failures: N = 24 (24/75 = 32%)

Local/regional L/R + distant" Distant

5/24(21%) 4/24(17%) 15/24(63%)" One failure in contralateral axilla.

3892




Â¡Iz,Ã¶

LU ~Zn <

8

100

90

80

70

60

50

40

30

20

10

0

3 All Patients

6 12 18 24 30 36 42 48 54 60

TIME TO LOCAL/REGIONAL RECURRENCE

Fig. 2. Time to locoregional recurrence.

Table 9 Summary of survival and lime to progression

66

GroupAll

patientsStageMIAStage1MBStage

1MBnoninflammatoryStage

lili! inflammatoryStage

IIIA andIIIBnoninflammatoryNo.

ofpatients76314193240Median

survival(mo)Â°391*35.3NRC35.3kMedian

time toprogression(mo)Â°34.235.934.234.223.234.2

" Median follow-up is 26.4 months.* Not reliable due to limited number of events.' NR, not reached.

o

St-zUJ

Fig. 3.

12 18 24 30 36 42 48 54 60 66

TIME TO PROGRESSION IN MONTHS

Time to progression.

pÃeteresponders with negative repeat biopsies. Two relapsed inthe breast and received mastectomies, and one relapsed in theskin of the breast and had excision of this lesion only. Sixteenof 23 (69%) complete responders with negative biopsies continue to be free of disease. Eleven PRs (33%) have relapsed and4 patients with no change (80%) have relapsed.

The median survival and time to progression for all patientsin this series are summarized in Table 9. Median time toprogression is 35.9 months for stage IIIA patients and 34.2months for stage IIIB (Fig. 3). The median survival is 35.3months for stage IIIB and indeterminate for stage IIIA (Fig. 4).

DISCUSSION

We treated locally advanced breast cancer by giving initialchemotherapy for as many cycles as necessary to achieve maximal response. This approach has yielded significant responserates in both stage IIIA and stage IIIB breast cancer. Stage IIIApatients had an objective response rate of 90% with 53% CR.Stage IIIB had objective response rate of 95%, with 51% CR.This approach achieved the maximum cytoreductive benefit ofchemotherapy before proceeding to local therapy. The admin-

CL

O)KZLUOÅ’

12 18 24 30 36 42 48

SURVIVAL IN MONTHS

60 66

Fig. 4. Overall survival.

istration of subsequent local therapy was guided by the patient's

response to induction chemotherapy. We believed that patientsachieving a CR with negative repeat biopsy had their tumorburdens reduced to the magnitude of minimal microscopiclevels and were therefore candidates for consolidative radiationalone. Patients with residual gross disease received debulkingsurgery.

The concept of giving a variable initial amount of chemotherapy is in contrast to other studies using fixed number ofchemotherapy cycles before local therapy. DeLena et al. (16)used 4 cycles of AV and reported an objective response in 70%with 15% CR and 55% PR. Thirty-six % of the patients in thatstudy had inflammatory histology; no difference in responserates were noted between inflammatory and noninflammatorypatients. Improved survival was noted in the noninflammatorygroup when compared with their subsequent study (17), 3 cyclesof AV were used, excluding patients with inflammatory histology, with objective responses in 53% (CR and PR rates of 4and 49%, respectively). This latter study used a 20% reductionin the doxorubicin dose compared to their earlier work, a factperhaps explaining their lower response rates. Rubens et al.(30) gave 4 cycles of AV to 12 patients and achieved a 50%objective response rate with 17% CR and 33% PR. The M. D.Anderson group (18) treated 52 patients with noninflammatoryhistology with 3 cycles of FAC plus B. Calmette-Guerin beforeinitiating local therapy. They observed an 82% objective response rate with 15% CR and 67% PR. Using the same induction regimen for patients with inflammatory histology, thisgroup obtained similar response rates with 12% of patientsachieving a CR (31). By giving a variable number of chemotherapy cycles until the maximum objective clinical responsewas attained, our induction chemotherapy regimen producedobjective responses, without significant morbidity, in 93% ofpatients with 49% CR and 44% PR. As shown in Fig. 1, if only3 or 4 cycles of chemotherapy had been given to our patients,as in these previously mentioned trials, only 32 or 50% of ourresponding patients, respectively, would have had a response toinduction chemotherapy. The median numbers of cycles ofchemotherapy in our series were 3 for partial responders and 5for complete responders. The substantial gains in objectiveresponse rate in our series compared to others in the literatureresulted in part from the increased proportion of patientsachieving complete response status. Therefore, higher numbersof cycles of chemotherapy needed to achieve this responseappear justified.

In the present study, patients achieving a clinical CR tochemotherapy underwent biopsy evaluation in the area of thebreast previously known to be involved with tumor. Using thisapproach the overall complete response rate with negative

3893




biopsy for this induction chemotherapy regimen was 36%. In arecent report from the M. D. Anderson Hospital (32) where 21patients with inflammatory breast cancer and 69 patients withlocally advanced noninflammatory breast cancer were treatedwith 3 cycles of FAC chemotherapy prior to mastectomy, 15patients (17%) had no evidence of gross disease but only 6patients (7%) had no evidence of microscopic disease. Whilegreat efforts were made to utilize clinical and radiologicalinformation in guiding each biopsy in our series, this methodis still subject to sampling errors. Nevertheless, the multiplebiopsy procedure has appeared useful in identifying clinical CRpatients with no or only microscopic residual tumor burden.These patients appear likely to do well with radiation alone,thereby avoiding the need for disfiguring surgery. At the presenttime, 24 of our patients with locally advanced breast cancerafter achieving a complete response have finished combinedmodality therapy and been rendered disease free with a costnet-ically intact breast. While cosmetic issues have not been ourprimary concern in the treatment of locally advanced breastcancer, the possibility of being able to maintain an intact breastafter aggressive treatment may provide an important psychological incentive in a population of patients, some of whominitially delayed seeking medical attention for fear of losingtheir breast.

The combined modality approach used in this series hasachieved a local and systemic disease-free state for all thosepatients who completed both induction chemotherapy and localtherapy, irrespective of their clinical stages at presentation orbest local response to induction therapy. Although the medianfollow-up is still short, this aggressive regimen appears toprovide excellent local control and encouraging early information on systemic disease control for stage IIIA patients. Whilestage IIIB patients have derived benefit from this combinedmodality approach, they still have a high rate of relapse and anoverall shorter survival. This is most apparent in those patientswith inflammatory breast cancer, where 8 of 12 such relapsedpatients presented with distant metastasis as a component oftheir initial treatment failure. This is comparable to inflammatory breast cancer patients treated with radiation alone (33)and attests to the aggressive nature of this form of breast cancer,even when treated with early intensive chemotherapy. An effortis being made to improve systemic control in inflammatorypatients in this study. After they are rendered disease free bythis combined modality regimen, they are randomized to receiveeither high dose melphalan plus autologous bone marrow transplant followed by maintenance chemotherapy or maintenancechemotherapy alone. It is hoped that this aggressive chemotherapy can improve survival in these poor prognosis patients.

REFERENCES

1. Haagensen, C. D. Diseases of the Breast, Ed. 2, p. 629. Philadelphia, W. B.SaundersCo., 1971.

2. Haagensen, C. D., and Stout, A. P. Carcinoma of the breast. II. Criteria ofoperability. Ann. Surg., US: 859-870, 1032-1051, 1943.

3. Badesse, F. Roentgen therapy alone as the method of treatment of cancer ofthe breast. Am. J. Roentgenol., 62: 311-319, 1949.

4. Amalric, R., SantamarÃa, F., Robert, F., Seigle, J., Altschuler, C., Kurtz, J.M., Spitalier, J. M., Brandone, H., Ayme, Y., Pollet, J. F., Burnmeister, R.,and Abed, R. Radiation therapy with or without primary limited surgery foroperable breast cancer. Cancer (Phila.), 49: 30-34, 1982.

5. Balawajder, I., Antich, P. P., and Boland, J. An analysis of the role ofradiotherapy alone in combination with chemotherapy and surgery in themanagement of advanced breast carcinoma. Cancer (Phila.), 51: 574-580,1983.

6. Bouchard, J. Advanced cancer of the breast treated primarily by irradiation.Radiology 84:823-842, 1965.

7. Bruckman, J. E., Harris, J. R., Levene, M. B., Chaffey, J. T., and Hellman,S. Results of treating stage III carcinoma of the breast by primary radiationtherapy. Cancer (Phila.), 43: 985-993, 1979.

8. Fletcher, G. H. Local results of irradiation in the primary management oflocalized breast cancer. Cancer (Phila.), 29: 545-551, 1972.

9. Langlands, A. O., Kerr, G. R., and Shaw, S. The management of locallyadvanced breast cancer by X-ray therapy. Clin. Oncol., 2: 365-371, 1976.

10. Pearlman, N. W., Guerra, O., and Fracchia, A. A. Primary inoperable cancerof the breast. Surg. Gynecol. Obstet., Â¡43:909-913, 1976.

11. Rubens, R. D., Armitage, P., Winter, P. J., Tong, D., and Hayward, J. L.Prognosis in stage III carcinoma of the breast. Eur. J. Cancer, 13: 805-811,1977.

12. Treurniet-Donker, A. D., Hop, W. C. J., and Hoed-Sijtsema, S. Radiationtreatment of stage III mammary carcinoma: a review of 129 patients. Int. J.RadiÃ¢t.Oncol. Biol., Phys., 6: 1477-1482, 1980.

13. Zucali, R., Uslenghi, C., Kenda, R., and Bonadonna, G. Natural history ofsurvival of inoperable breast cancer treated with radiotherapy and radiotherapy followed by radical mastectomy. Cancer (Phila.), 37: 1422-1431, 1976.

14. Bedwinek, J., Rao, D. V., Perez, C., Lee, J., and Finberg, B. Stage III andlocalized stage IV breast cancer: irradiation alone versus irradiation plussurgery. Int. J. Radial. Oncol. Biol. Phys., 8: 31-36, 1982.

15. GrÃ¶hn,P., Heinonen, E., Klefstrom, P.. and Tarkkanen, J. Adjuvant postoperative radiotherapy, chemotherapy and immunotherapy in stage III breastcancer. Cancer (Phila.), 54: 670-674, 1984.

16. DeLena, M., Zucali, R., Viganotti, G., Valagussa, P., and Bonadonna, G.Combined chemotherapy-radiotherapy approach in locally advanced (T3b-T4)breast cancer. Cancer Chemother. Pharmacol., /: 53-59, 1978.

17. DeLena, M., Varini, M., Zucali, R., Rovini, D., Viganotti, G., Valagussa, P.,Veronesi, U., and Bonadonna, G. Multimodality treatment for locally advanced breast cancer. Cancer Clin. Trials, 4: 229-236, 1981.

18. Hortobagyi, G. N., Blumenschein, G. R., Spanos, W., Montague, E. D.,Buzdar, A. U., Yap, H. Y., and Schell, F. Multimodal treatment of locore-gionally advanced breast cancer. Cancer (Phila.), 51: 763-768, 1983.

19. Donehower, R., Allegra, J. C., Lippman, M. E., and Chabner, B. Combinedeffects of methotrexate and 5-fluoropyrimidine on human breast cancer cellsin serum-free culture. Eur. J. Cancer, 16: 655-661, 1980.

20. Lippman, M. E., BolÃ¡n,G., and Huff, K. The effects of estrogens andantiestrogens on hormone-responsive human breast cancer in long-termtissue culture. Cancer Res., 36: 4595-4601, 1976.

21. Nawata, H., Chong, M., Bronzert, D., and Lippman, M. E. Estradiol independent growth of a subline of MCF-7 human breast cancer cell line inculture. J. Biol. Chem., 256: 6895-6902, 1981.

22. Weichselbaum, R. R., Hellman, S., Piro, A. J., Nove, J. J., and Little, J. B.Proliferation kinetics of a human breast cancer cell line in vitro followingtreatment with 17/J-estradiol and 1-0-D-arabinofuranosylcytosine. CancerRes., 38: 2339-2345, 1978.

23. Green, M.D., Whybourne, A. M., Taylor, I. W., and Sutherland, R. L. Effectsof antiestrogens on the growth and cell cycle kinetics of cultured humanmammary carcinoma cells. In: R. L. Sutherland and V. C. Jordon (eds.),Non-Steroidal Antiestrogens. pp. 397-413. New York: Academic Press,1981.

24. Aitken, S. C., and Lippman, M. E. Hormonal regulation of net DNAsynthesis in MCF-7 human breast cancer cells in tissue culture. Cancer Res.,42: 1727-1735, 1982.

25. Allegra, J. C., Woodcock, T. M., Richman, S. P., Bland, K. L, and Wittliff,J. F. A phase II trial of tamoxifen, breast cancer premarin, methotrexate and5-fluorouracil in metastatic breast cancer. Breast Cancer Res. Treat., 2: 93-100, 1982.

26. Lippman, M. E., Cassidy, J., Wesley, M., and Young, R. C. A randomizedattempt to increase the efficacy of cytotoxic chemotherapy in metastaticbreast cancer by hormonal synchronization. J. Clin. Oncol., 2: 28-36, 1984.

27. Sorace, R. A., Bagley, C. S., Lichter, A. S., Danforth, D. N., Wesley, M. W.,Young, R. C., and Lippman, M. E. The management of nonmetastatic locallyadvanced breast cancer using primary induction chemotherapy with hormonalsynchronization followed by radiation therapy with or without debulkingsurgery. World J. Surg., 9: 775-785, 1985.

28. Breast Cancer: Suggested Protocol Guidelines for Combination Chemotherapy Trials and for Combined Modality Trials. The Breast Cancer Task ForceTreatment Committee NCI National Cancer Institute. Publication 78-1192.Washington, DC: United States Department of Health, Education, andWelfare, 1978.

29. Kaplan, E. L., and Meier, P. Nonparametric estimation from incompleteobservations. J. Am. Statist. Assoc., 53:457-481, 1958.

30. Rubens, R. D., Sexton, R., Tong, D., Winter, P. J., Knight, R. K., andHayward, J. L. Combined chemotherapy and radiotherapy for locally advanced breast cancer. Eur. J. Cancer, 16: 351-356, 1980.

31. Barker, J. L., Montague, E. D., and Peters, L. J. Clinical experience withirradiation of inflammatory carcinoma of the breast with and without electivechemotherapy. Cancer (Phila.), 45: 625-629, 1980.

32. Feldman, L., Hortobagyi, G., Buzdar, A., and Blumenschein, G. Pathologiccomplete remission in patients with inflammatory breast cancer and locallyadvanced breast cancer. Proc. Am. Soc. Clin. Oncol., 25: 197, 1984.

33. Chu, A. M., Wood, W. C., and Doucette, J. A. Inflammatory breast carcinoma treated by radical radiotherapy. Cancer (Phila.), 45: 2730-2737, 1980.

3894



1987;47:3889-3894. Cancer Res Sandra M. Swain, Richard A. Sorace, Caroline S. Bagley, et al. Approach of Locally Advanced Nonmetastatic Breast CancerNeoadjuvant Chemotherapy in the Combined Modality

Updated version

http://cancerres.aacrjournals.org/content/47/14/3889

Access the most recent version of this article at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/47/14/3889To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]



Documents

NeoadjuvantChemotherapy in the CombinedModality Approachof ... · junction with local therapy to treat locally advanced breast cancer. In a retrospective review from the Harvard-Joint