Surgery as Part of a Combined Modality Approach for Inflammatory Breast Carcinoma

PETER SCHAFER, MD," PIERRE ALBERTO, MD,t MICHEL FORNI, MD.' DRAGAN OBRADOVIC. MD,* GUDRUN PIPARD, MD,§ AND FELIX KRAUER, MD'

From November 1977 to February 1984 21 patients with inflammatory breast carcinoma (IBC) were treated in a prospective, nonrandomized study with initial chemotherapy, consisting of a combination of chlorambucil, methotrexate, 5-fluorouracil, and doxorubicin. Modified radical mastectomy was performed after three courses of chemotherapy followed by another six courses of the same chemotherapy. Locore- gional radiotherapy concluded the treatment program. All patients responded clinically to the initial chemotherapy. Modified radical mastectomy provided a direct histologic control of the effectiveness of the treatment and, in case of tumor persistence, allowed cytoreduction. No scarring problems were observed and postoperative chemotherapy could be administered without any delay. Radiotherapy was decisive for locoregional tumor control which was obtained in all patients who were treated strictly according to the protocol. Dermal lymphatic involvement represented in this series of clinical inflammatory breast carcinoma no pejorative feature and positive hormone receptors failed to be of predictive value for response to treatment. Considerably variable tumor persistence after initial chemotherapy and the results relative to local tumor control advocate surgery as part of a combined modality approach for inflammatory breast carcinoma.

Cancer 59:1063-1067, 1987.

NFLAMMATORY BREAST CARCINOMA (IBC) is defined I by characteristic clinical features such as erythema, "peau &orange", and augmentation of breast volume.' For its ominous prognosis, almost all researchers consid- ered this type of carcinoma of the breast to be an absolute contraindication to the performance of a radical mastec- tomy, either initially or after radiotherapy.'-8 In a few institutions radical mastectomy was done in selected pa- tients?." Initial chemotherapy, particularly doxorubicin, permitted to delay the appearance of distant metastasis, but not to improve the locoregional control."-' Radical mastectomy, therefore, was reconsidered in the context of locoregional treatment .8J 4 3 1

We present here a series of 2 1 patients with clinical IBC who were included in a prospective nonrandomized study using a combination of chemotherapy, surgery, and ra- diotherapy.

From the *Department of Gynecology, YDivision Onco-Hematology, $Division of Gynecologic Pathology, and §Division of Radiotherapy, University Hospital of Geneva, Switzerland.

Address for reprints: Peter Schafer, MD, Department of Obstetrics and Gynecology, University Hospital, 20, rue Alcide Jentzer, 12 I I Ge- neva 4, Switzerland.

The authors thank Dr. G. Rosset for the steroid hormone receptors assays, Bernadette Mermillod for statistical assistance, and Francoise Misset for manuscript preparation.

Accepted for publication October 20, 1986.

Patients and Methods

Between November 1977 and March 1984, 2 1 previ- ously untreated patients with clinical IBC and without evidence of distant metastasis were subjected to the fol- lowing procedure.

Diagnosis

Clinical diagnosis was completed by infrared thermog- raphy, bilateral mammography, fine-needle aspiration for cytologic study and surgical biopsy for histologic typing and hormone receptor analysis. Distant metastasis was excluded by a chest x-ray, a bone scan, and a liver ultra- sound.

Chemotherapy

Chemotherapy consisted in a four-drug combination with chlorambucil, methotrexate, 5-fluorouracil, and doxorubicin. Chemotherapy dose schedule included chlorambucil, 5 mg/m2 orally for 14 days; methotrexate, 10 mg/m2 intravenously (IV) on days 1 and 8; 5-fluoro- uracil, 500 mg/m2 IV on days 1 and 8; and doxorubicin, 10 mg/m2 IV on days I and 8. This chemotherapy was repeated three times at 28-day intervals (Table 1).

1063

1064 CANCER March 15 1987 Vol. 59

TABLE 1. Inflammatory Breast Carcinoma Treatment Plan

Initial Modified Postoperative chemotherapy > radical > chemotherapy > Locoregional

3 courses mastectomy 6 courses radiotherapy

Initial and postoperative chemotherapy Chlorambucil Methotrexate 10 m&m2 IV d 1 + 8 5-Fluorouracil 500 mg/m2 IV d 1 + 8 Doxorubicin 10 mg/m2 IV d 1 + 8

5 mg/m2 orally d 1-14

Duration of cycle: 28 days

Radiotherapy To chest wall and draining

To residual disease lymphatics Cobalt 60 50 cGy in 5 wk

Electrons 10-15 cGy in 1-2 wk

IV: intravenously.

Surgery

The operation performed was a modified radical mas- tectomy. The pectoral muscles were conserved and axillary dissection was limited to the inferior two thirds of the axilla.

Postoperative Chemotherapy

The same drug combination at the same dosages as for initial chemotherapy was reinstituted within 2 weeks of mastectomy for another 6 courses.

Radiotherapy

Radiotherapy delivered a 50-cGy tumor dose within 5 weeks to the chestwall, to the axilla, and to the internal mammary and supraclavicular lymph nodes. Areas of known residual disease after surgery were boosted with additional 10 to 15 cGy (Table 1).

Steroid Hormone Receptors

Cytosol was prepared from fresh tissue of surgical spec- imen after pathologic examination and assayed by the dextran charcoal separation. Positive levels were consid- ered above 10 fmol/mg protein for estrogen (E) and pro- gesterone (P).

Statistics

The probability of survival was calculated according to Kaplan and MeierI6 and the 95% confidence intervals were established according to Pet0 et

Results

Four hundred thirty-nine patients with histologically confirmed breast carcinoma had been registered at the Department of Gynecology of the University Hospital of

Geneva between November 1977 and February 1984. Thirty patients were introduced in the trial. Six patients had to be excluded from analysis because they did not fulfill the criteria of clinically IBC. In two patients his- tologic study of the surgical biopsy specimen showed ne- crosis in a circumscribed type of carcinoma according to Haagensen;’ both patients had clinically and histologically uninvolved axillary lymph nodes. They were excluded from the series, but submitted to the same treatment course and are alive without recurrence, 33 and 32 months after diagnosis, respectively. One patient with primary IBC was withdrawn from the study for tumor progression after the first course of initial chemotherapy.

Analysis includes all patients (N = 2 1) registered up to February 1984 and information is updated to October 1985. Median follow-up of the whole group is currently 37 months (range, 20-95 months).

Diagnosis

Characteristic clinical features of “primary IBC”5979’8 consisting of a primary enlargement and induration of the breast, redness, heat, and at least one third of the overlying skin presenting edema, were noted in 15 pa- tients. Eleven of them complained of pain in the affected breast. Diagnosis was confirmed in 12 cases by surgical biopsy. In the first two cases included in the study and in a 7 1 -year-old patient in very poor general condition at presentation, treatment was initiated on the bases of pos- itive aspiration cytology. Histologic typing was exclusively evaluated on the mastectomy specimen. Neoplastic in- filtration of the intramammary lymphatics was observed in all 15 cases of primary IBC and neoplastic infiltration of the dermal lymphatics was observed in 11 of the 15 patients (Table 2). Six patients developed inflammatory signs secondary to long-standing breast masses; four of them had also anamnestic global or partial retraction of the breast before the inflammatory evolution. None of the six patients with “secondary IBC”5,7,18 complained about breast pain. Histologic study of the surgical biopsy specimen showed neoplastic invasion of the intramam- mary lymphactics in three cases; neoplastic invasion of the dermal lymphatics was observed in one of them (Ta- ble 3).

Clinically involved lymph nodes were present in all pa- tients: in 5 cases classified as Nlb, in 14 cases as N2, and in 2 cases as N3. Median age was 47 years, and 14 patients were premenopausal (Tables 2 and 3).

Initial Chemotherapy

No wound-healing problems were encountered after diagnostic surgery, and the introduction of initial che- motherapy was possible within optimal time. All patients received the programmed three cycles of initial chemo- therapy at the calculated doses. Tumor response was

No. 6 SURGERY FOR INFLAMMATORY BREAST CA - Schuyer et al. 1065

characterized primarily by the disappearance of pain and by the regression of the inflammatory signs, followed by the shrinking of the tumor masses in the breast and in the axilla.

It was judged impossible to distinguish correctly, by clinical examination, between complete and partial re- sponse. Chemotherapy was interrupted deliberately fol- lowing the treatment plan after three courses to perform modified radical mastectomy. Tolerance was excellent: only one patient manifested severe hematologic toxicity (leukocyte count 1 700/p1); severe digestive toxicity or al- opecia were not observed.

Surgery

Surgery was never delayed for reason of toxicity. No major technical problem was encountered in any of the 21 cases and wound healing per primam resulted in all patients.

Pathologic Findings

Pathologic examination of the mastectomy specimen demonstrated that resection of the tumor was incomplete in eight cases, predominantly because of the infiltration of the muscles. Complete absence of tumor in the breast and in the axilla was noted in one case, absence of tumor in the breast with persistence of tumor in the axilla in two cases, and persistence of tumor in the breast and in the axilla with the usual histologic modifications due to che- motherapy in 18 cases.

Receptors for E and for P were detected with the fol- lowing distribution: E+/P+, 5 cases; E+/P-, 5 cases; E-/ P+, 0 cases; and E-/P- 11 cases.

Postoperative Chemotherapy

Postoperative chemotherapy could always be reintro- duced within 2 weeks after mastectomy. Sixteen patients accomplished the six cycles of postoperative chemother- apy without evidence of disease. Three patients with pri- mary IBC developed distant recurrence during the post- operative chemotherapy; they were subsequently submit- ted to another chemotherapy regimen. In one patient with secondary IBC metastases were detected before radio- therapy.

Radiotherapy

Radiotherapy was given without unusual side effects to eight patients with primary IBC and to all six patients with secondary IBC. Two patients refused radiotherapy. For another two patients this treatment modality was judged to be of no benefit since the tumor extended at presentation far beyond the planned fields of irradiation.

TABLE 2. Initial Characteristics for Patients With Primary Inflammatory Breast Carcinoma

Neoplastic infiltration of lymphatics N

Clin stage N Age Intramammary Dermal (UICC) Postop Path

1 26 + + I1 0/6* 2 31 + I1 016 t 3 31 + + I1 12/20 4 40 + - I1 11/25 5 41 + + I1 15/15 6 44 + + Ib 21/37 7 46 + + I1 6/13 8 46 + + I1 417 9 46 + I1 919

10 47 + - Ib 016' 11 49 + + I1 8/15 12 58 + + I1 +$ 13 59 + + Ib 15/17 14 64 + + I11 10/22 15 71 + + I11 +t:

-

-

* Fine-needle aspiration of axillary nodes positive before treatment. t Diagnostic axillary biopsy specimen: 313 nodes positive. t: Important axillary tumor mass, nodes not recognized. Clin: clinical; Postop: postoperative; Pathol: pathology: UICC: Inter-

national Union Against Cancer.

Locoregional Recurrence

Among the eight patients with primary inflammatory carcinoma who underwent locoregional radiotherapy, only one failure was observed. In deviation to the study protocol this patient had received only 33.5 cGy to the chest wall and local recurrence occurred 34 months after mastectomy.

None of the six patients with secondary inflammatory carcinoma, all of whom were irradiated, has shown lo- coregional failure until now.

AH seven patients with primary IBC who were not sub- mitted to locoregional radiotherapy for the different rea- sons discussed before developed local recurrence 8 to 17 months after mastectomy (Table 4).

TABLE 3. Initial Characteristics for Patients With Secondary Inflammatory Breast Carcinoma

Neoplastic infiltration of lymphatics N

Clin stage N Age Intramammary Dermal (UICC) Postop Path

I1 719 - - 1 45 2 52 + + I1 13/13

- IB 4/30 3 53 4 59 + - I1 2/10

- I1 9/18 5 61 6 68 + 1B 212

-

-

-

Clin: clinical; Postop: postoperative; Path pathology.

1066 CANCER March 15 1987 VOl. 59

TABLE 4. Time of Appearance of Locoregional Recurrent Disease (>20 Months’ Follow-Up for All Survivors)

Months Chemotherapy, Chemotherapy after mastectomy no irradiation and irradiation

0-6 017 0114 7-12 417 0114

13-24 717 0/14 25-36 717 ]*/I4

* Only 32.5 cGy.

Development of Distant Metastases and Survival

As of October 1985, four patients who had accom- plished the whole therapeutic program were alive without recurrence, 29, 30, 35, and 66 months, respectively, after diagnosis (two patients with primary inflammatory car- cinoma and two patients with secondary inflammatory carcinoma). Three patients were alive with recurrence. Fourteen patients have died of tumor progression. Median survival is 43 months.

Significance of Dermal Lymphatic Involvement

Eleven patients with primary IBC and 1 patient with secondary IBC had dermal lymphatic involvement on histologic examination (Table 5). In four patients with

TABLE 5. Dermal Lymphatic Involvement, Hormone Receptors, and Distant Disease-Free Survival

Receptors Distant fmolfmg prot disease-free Dermal

survival lymphatic Age E P (mo) involvement

I 26 0 0 2 31 0 0 3 31 0 0 4 44 0 0 5 46 0 0 6 47 16 0 7 49 0 0 8* 52 0 0 9 58 38 0

10 59 0 0 1 1 64 50 20 12 71 0 0

13 31 0 0 14 40 14 47 15* 45 0 0 16 46 79 137 17 47 245 0 18* 53 22 446 19* 59 90 276 20* 67 65 0 21* 68 97 0

66t 23 22 41 15 30 t 14 1 1 13 10 29 28

10 21

8 4

30

35t 12 7

29t

Present Present Present Present Present Present Present Present Present Present Present Present

Absent Absent Absent Absent Absent Absent Absent Absent Absent

* Secondary inflammatory breast carcinoma. t Alive. Prot: protein.

primary IBC and in five patients with secondary IBC no dermal lymphatic involvement could be demonstrated (Table 5). Median age of the two groups was not different (Table 5). Estrogen receptors were identified in the tumors in 3 of 12 patients with dermal lymphatic involvement, and in 7 of 9 patients without dermal lymphatic involve- ment (Table 5). Distant disease-free survival for all 21 patients is listed in Table 5 . Probability of distant disease- free survival and probability of survival for the two his- tologic subgroups were not calculated since their number was too small.

Discussion

Published data on local tumor control and survival may vary because of different interpretations of the term “in- flammatory carcinoma.” A review of the three largest se- ries of IBC in which patients were included according to their clinical features shows that the incidence of dermal lymphatic invasion in adequately examined surgical specimens varies between 50% and 80%.53’8~19

Ellis and Teitelbaum” suggested that the term “clinical inflammatory carcinoma” should be abandoned and be substituted by the name of “dermal lymphatic carcino- matosis of the breast” as determined by histologic sub- dermal tumor emboli in the lymphatics. In our series, all 15 patients with primary IBC showed infiltration of the intramammary lymphatics, but only in 11 cases (73%) could the presence of dermal lymphatic involvement be demonstrated (Table 2).

Diagnosis of “secondary IBC” is based on the evolution of its clinical features and defines a very heterogeneous group of patients. The absence of local pain is character- istic for these tumors and may be related to the rare ob- servation of dermal lymphatic involvement. A few of these cancers may be slow-growing lesions; they may account for the patients reported in older series as cured by mas- tectomy. In the majority of these locally advanced breast cancers, the appearance of peau d’orange and a sudden increase of the breast volume indicate a relatively ad- vanced stage of the evolution. Distant metastasis are al- ready present, although not yet detectable. No difference between the two groups of IBC can be demonstrated his- tologically at this stage. Computerized cytometric analysis is currently under evaluation in order to discriminate dif- ferent cell populations. Rapid evolution and short-term prognosis, however, seem to justify the definition of IBC on a clinical basis as well as by pathologic criteria.

Camp3 considered IBC to be an absolute contraindi- cation for radical mastectomy, either initially or after ra- diation therapy.

Ellis and Teitelbaum’’ had examined the pathologic material and/or reports of all but one patient reported free of disease at least 5 years after radical mastectomy:

No. 6 SURGERY FOR INFLAMMATORY BREAST C A - Schuyeer et a/. 1067

none of these patients had dermal lymphatic metastasis. In addition, no patient with erythema of the breast or clinical inflammatory carcinoma who survived 5 years without recurrence after radical mastectomy at the Barnes Hospital of St. Louis (Missouri) over a 10-year period had dermal lymphatic metastasis. l 9 They came to the conclu- sion, since the presence of dermal lymphatic carcino- matosis indicates that the malignancy has extended al- ready beyond the confines of the mammary gland, that radical mastectomy was not justified. For the same reasons locoregional radiotherapy would not be the appropriate treatment. A definite improvement was achieved only when chemotherapy was introduced, thus delaying the appearance of distant metastasis.

In our series a difference in the outcome of the patients with and without dermal lymphatic involvement could not be observed (median distant disease-free survival, 22.5 and 12 months, respectively, not significant). Tumors with dermal lymphatic involvement responded to treatment as well as tumors without dermal lymphatic involvement (Table 5) . Also, the presence of hormone receptors was of no predictive value (Table 5) . In our study-design, which in 1977 represented a totally new approach, a low- dose initial chemotherapy had been chosen in order not to compromise surgery and radiotherapy. Retrospectively we think that a more intensive chemotherapy could have been tolerated since toxicity was negligible. Clinical re- sponse of locoregional tumor to the low-dose initial che- motherapy was excellent. The correlation between clinical response and pathologic response was, however, disap- pointing: in the majority of the cases clinical response was not correlated with total disappearance of the tumor.

A locoregional treatment (in order to achieve local control, even in presence of distant metastasis) is man- datory, even more so in that it contributes to the im- provement of the quality of life. The literature confirms that the association of chemotherapy and locoregional radiotherapy did not attain this objective in about one third of the case^.^,^,'^ In our series, all patients with pri- mary IBC who were not submitted to radiotherapy de- veloped local recurrence within a short time after mas- tectomy, whereas local control was obtained in all patients who were treated strictly with a combination of surgery, chemotherapy, and radiotherapy (Table 4).

Mastectomy as treatment modality for IBC has a double function: first, it provides, when realized consequently to primary chemotherapy, a pathologic control of its effec- tiveness and, second, it has a cytoreductive function on persisting tumor masses. Locoregional radiotherapy is probably best placed at the end of the treatment schedule and it is certainly decisive for locoregional tumor control.

Haagensen stated in 1972 that it seemed to him entirely unreasonable to treat these patients by radical mastectomy when no cures can be expected and no definitive evidence of prolongation of life can be shown.’ The introduction of chemotherapy has profoundly modified the ominous prognosis of IBC. Inflammatory breast carcinoma must actually be considered as a curable disease in certain pa- tients.

Even if distant metastases lead to the lethal outcome of the disease, local tumor control can be achieved and a treatment plan for IBC has to take into consideration the curability of the disease as well as the aspect of quality of life.

REFERENCES

I . Haagensen CD. Diseases of the breast. ed. 2. Philadelphia: W.B. Saunders, 1971; 576-584.

2. Brambda C, Rossi A. I1 carcinoma infiammatorio della mammella: Ruolo della chemioterapia. In: Veronesi V, Staudacher V, Salvadori B, Emanuelli H, Zucali R, Santoro A, eds. I1 Carcinoma Della Mammella: Diagnosi e Terapia. Milano: Ambrosiana, 1983; 303-3 10.

3. Camp E. Inflammatory carcinoma of the breast. Am J Surg 1976;

4. Dao TL, McCarthy JD. Treatment of inflammatory carcinoma of the breast. Surg Gynecol Obstet 1957; 105:289-294.

5. Lee BJ, Tannenbaum NE. Inflammatory carcinoma of the breast: A report of 28 cases from the breast clinic of the Memorial Hospital. Surg Gynecol Obstei 1924; 39580-595.

6. Robbins GF, Shah J, Rosen P, Chu F, Taylor J. Inflammatory carcinoma of the breast. Surg Clin North Am 1974; 542301-810.

7. Taylor GW, Meltzer A. “Inflammatory carcinoma” of the breast. Am J Cancer 1938; 33:33-49.

8. Treves N. The inoperability of inflammatory carcinoma of the breast. Surg Gynecol Obstet 1959; 109:240-242.

9. Chu AM, Wood WC, Doucette JA. Inflammatory breast carcinoma treated by radical radiotherapy. Cancer 1980; 45:2730-2737.

10. Rogers CS, Fitts WT. Inflammatory carcinoma of the breast: A critique of therapy. Surgery 1956; 39:367-370.

I I . Barker JL, Montague ED, Peters LJ. Clinical experience with ir- radiation of inflammatory carcinoma of the breast with and without elective chemotherapy. Cancer 1980 49625-629.

12. Buzdar AU, Montague ED, Barker JL, Hortobagyi GN, Blumen- schein GR. Management of inflammatory carcinoma of breast with combined modality approach: An update. Cancer I98 1; 47:2537-2542.

13. De Lena M, Varini M, Zucali R et al. Multimodal treatment for locally advanced breast cancer. Cancer Clin Trials 1981; 4:229-236.

14. Hagelberg RS, Jolly PC, Anderson RP. Role of surgery in the treatment ofinflammatory breast carcinoma. Am JSurg 1984; 148: 125- 131.

15. Zylberberg B, Salat-Baroux J, Ravina JH et al. Initial chemo- therapy in inflammatory carcinoma of the breast. Cancer 1982; 49:1537- 1543.

16. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Statist Assoc 1959; 53:457-48 I .

17. Pet0 R, Pike MC, Armitage P et al. Design and analysis of ran- domized clinical trials requiring prolonged observation of each patient. Br J Cancer 1977; 35:l-39.

18. Lucas FV, Perez-Mesa C. Inflammatory carcinoma of the breast. Cancer 1978; 41:1595-1605.

19. Ellis DL, Teitelbaum St.L. Inflammatory carcinoma of the breast: A pathological definition. Cancer 1974; 33:1045-1047.

131:583-586.