Ductal carcinoma in situ (intraductal carcinoma) of the breast treated with breast-conserving surgery and definitive irradiation correlation of pathologic parameters with outcome of

2532

Ductal Carcinoma In Situ (Intraductal Carcinoma) of the Breast Treated with Breast-Conserving Surgery and Definitive Irradiation Correlation of Pathologic Parameters with Outcome of Treatment

Lawrence J. Solin, M.D.,* I-Tien Yeh, M.D.,t John Kurtz, M.D.,S Alain Fourquet, M.D.,§ Abram Recht, M.D., 11 Robert Kuske, M.D.,lT Beryl McCormick, M.D.,# Michael A. Cross, M.D.,** Delray 1. Schultz, Ph.D.,tt Robert Amalric, M.D.,SS Virginia A. LiVolsi, M.D.,t Michael J . Kowalyshyn, M.D.,& Ioachim Torhorst, M.D., 11 11 Jocelyne Jacquemier, M.D.,lTlT Cindy D. Westermann, M.D.,** Gwen Mazoujian, M.D.,## Brigitte Zafrani, M.D.,§ Paul P. Rosen, M.D.,*** Robert L. Goodman, M.D.,* and Barbara L. Fowble, M.D.*

Background. To evaluate the pathologic characteristics of the primary tumor relative to local control, survival, and freedom from distant metastases, an analysis was performed of 172 patients with ductal carcinoma in situ (intraductal carcinoma) of the breast treated with breast-conserving surgery and definitive breast irradiation.

From the *Department of Radiation Oncology, University of Pennsylvania School of Medicine and the Fox Chase Cancer Center, Philadelphia, Pennsylvania; the Departments of tPathology and Lab- oratory Medicine and ttBiostatistics, Cancer Center, University of Pennsylvania School of Medicine, Philadelphia; the Departments of $Radiation Oncology and 11 11 Pathology, University Hospital, Basel, Switzerland; the SInstitut Curie, Paris, France; the )I Joint Center for Radiation Therapy, Department of Radiation Therapy, Harvard Medi- cal School, Boston, Massachusetts; the llDepartment of Radiation On- cology, Mallinckrodt Institute of Radiology and the ##Department of Pathology, Washington University School of Medicine, St. Louis, Missouri; the Departments of #Radiation Oncology and ***Pathol- ogy, Memorial Sloan-Kettering Cancer Center, New York, New York; the **University of Cincinnati, Cincinnati, Ohio; the Departments of $$Radiotherapy and llllPathology, Institut Paoli-Calmettes, Marseille, France; and the §§Department of Pathology, Fox Chase Cancer Center, Philadelphia.

Address for reprints: Lawrence J. Solin, M.D., Department of Radiation Oncology, Hospital of the University of Pennsylvania, 3400 Spruce Street, Philadelphia, PA 19104.

Accepted for publication November 12, 1992.

Methods. The clinical records and pathology slides were reviewed from 172 women with ductal carcinoma in situ treated with breast-conserving surgery and definitive breast irradiation at multiple institutions in Europe and the United States. Central pathology review was performed by one pathologist without knowledge of the clinical outcome. The clinical outcome was measured in terms of local control, overall survival, cause-specific survival, and freedom from distant metastases. The median follow-up time was 84 months (range, 17-177 months).

The pathologic parameters evaluated were histologic subtype, nuclear grade, amount of necrosis, and final pathology margin. The only pathologic parameter that correlated with the rate of local recurrence was the presence versus the absence of the combination of the histologic subtype of comedo carcinoma plus nuclear grade 3 (8-year actuarial rate of local recurrence of 20% versus 5%, respectively; P = 0.009 on univariate analysis; P = 0.017 on multivariate analysis). None of the pathologic parameters evaluated correlated with overall survival (all P 2 O.lS), cause-specific survival (all P 2 0.13), or freedom from distant metastases (all P 2 0.13).

These results have demonstrated that there are important differences in the rate of local recurrence based on the pathologic characteristics of the primary tumor for women with ductal carcinoma in situ treated with breast-conserving surgery and definitive irradiation. However, the differences in local recurrence

Results.

Conclusions.

Intraductal Carcinoma of the BreastlSolin et al . 2533

have not been associated with differences in survival or freedom from distant metastases. Careful follow-up for patients at increased risk for local recurrence is warranted because of the potential ability to salvage patients with local recurrence. Cancer 1993; 712532-42.

Key words: ductal carcinoma in situ, intraductal carcinoma, breast carcinoma, breast-conservation treatment, definitive breast irradiation, local recurrence.

Recent clinical studies have reported the outcome of women treated with breast-conserving surgery and definitive irradiation for ductal carcinoma in situ (intraductal carcinoma) of the breast.'-17 The outcome of treatment in clinical studies has focused mainly on local control and survival. Recent reports have suggested that there may be important differences among the various histologic subtypes of intraductal carci-

ecurrence in the breast has been correlated with the pathologic characteristics of the intraductal carcinoma for women treated with excision of the primary tumor without radiation treatment.'"''

There is limited information available that corre- lates the pathologic characteristics of the intraductal carcinoma with the outcome of treatment for women treated with breast-conserving surgery and definitive irradiation, and only a small number of studies with small numbers of patients have been rep~rted. ' ,~,"- '~ The current study was performed to evaluate in detail the importance of the pathologic characteristics of the intraductal carcinoma relative to outcome for 172 women treated with breast-conserving surgery and definitive irradiation.

noma~3,8,11-13,18-20 R

Materials and Methods

A database of women treated with breast-conserving surgery and definitive irradiation for ductal carcinoma in situ (intraductal carcinoma) of the breast had been established by combining the data from the treatment of these patients from multiple institutions in Europe and the United States.I5 A comprehensive analysis of local control and survival has been reported previ-

To analyze pathologic parameters potentially correlating with outcome, multiple institutions contrib- uted pathology slides for a central pathology review, which was performed by one pathologist (I.Y.) without the knowledge of the clinical outcome. One institution from the previousliy reported study15 elected not to par- ticipate in the central pathology review. For the current study, 172 patienfs were confirmed as having pure intraductal carcinoma of the breast on central pathology review (Table 1). When there was uncertainty regarding

ously.15

Table 1. Distribution of Patients According to Location Where Treated

Europe University Hospital, Basel, Switzerland Institut Curie, Paris, France Cancer Institute, Marseille, France

Joint Center for Radiation Therapy,

Memorial Sloan-Kettering Cancer

University of Cincinnati, Cincinnati, OH University of Pennsylvania School of

Medicine and the Fox Chase Cancer Center, Philadelphia, PA

St. Louis, MO

United States

Boston, MA

Center, New York, NY

Mallinckrodt Institute of Radiology,

Total

NO. (Yo)

10 (6) 37 (22) 17 (10)

26 (15)

6 (3) 17 (10)

36 (21)

23 (13) 172 (100)

the confirmation of the diagnosis of intraductal carcinoma (i.e., the central pathology review diagnosed atypical ductal hyperplasia or intraductal carcinoma with microinvasion), an independent pathology review was performed by a second pathologist (V.A.L.). Pathology slides from many of the patients previously reportedl5 could not be obtained for central pathology review because a number of the cases were old and slides from many of the cases from outside institutions could not be obtained.

All women included in the current study had Amer- ican Joint Committee on Cancer (AJCC) clinical stage Tis NO MO ductal carcinoma in situ (intraductal carcinoma) of the breast." No histologic subtype of intraductal carcinoma was excluded. Terms that have been used synonymously include ductal carcinoma in situ (DCIS), intraductal carcinoma, and noninvasive ductal carcinoma.

All women included in the current study had unilat- eral intraductal carcinoma of the breast. Patients with bilateral intraductal carcinoma of the breast at presentation were not included. Breast-conserving surgery was followed by definitive breast irradiation for all patients. Criteria for patient selection and treatment have been reported in detail.15 With the exception of one patient, the 172 patients reported in the current study are a sub- set of the patients previously r ep~r t ed . '~ In addition, many of the patients included in the current study have been included in previous single or limited institution

Details of treatment have been reported previ- 0us1y.'~ A brief summary of the method of treatment for these patients is as follows. All patients underwent breast-conserving surgery, which included complete gross excision of the primary tumor. Re-excisional

reports. 1,2,5,7-10,14,16.17

2534 CANCER April 25, 2993, Volume 71, No. 8

biopsy was performed in 14% (24 of 172) of the patients. The axillary lymph nodes were pathologically staged in 66 (38%) of the patients, and all were pathologically node negative. All patients had definitive irradiation of the whole breast, and a boost to the primary tumor site was used in 67% (115 of 172) of the patients. The median total rahation dose was 6000 cGy (mean, 6144 cGy; range, 3120-8200 cGy). A total radiation dose of 5000 cGy or more was administered in 95% (163 of 172) of the patients, and a total radiation dose of 6000 cGy or more was administered in 62% (107 of 172) of the patients. Regional nodal irradiation to one or more regional nodal areas was used in 30% (52 of 172) of the patients. The regional nodal areas treated were the axilla (n = 21), the supraclavicular region (n = 43), and/or the internal mammary nodal regions (n = 27). Systemic therapy was used in nine patients, all from a single institution. The systemic therapy used was peri- operative or intraoperative thiotepa (n = 6), oophorec- tomy (n = 2), or tamoxifen (n = 1).

The original pathology reports were reviewed for determination of gross pathology features. The number of slides reviewed per patient was recorded for 169 of the 172 patients, and the median number of slides reviewed per patient was 3 (mean, 6; range, 1-72). When a limited number of slides were reviewed for an individual patient, the slides reviewed were representative of the case as determined by the institutional pathologist at the institution where the patient was treated. In general, representative slides, rather than all available ones, for each patient were referred for central pathology review for the current study.

The architectural pattern, nuclear grade, and amount of necrosis were scored separately for the current study. The primary histologic architectural pattern of ductal carcinoma in situ was categorized as comedo, cribriform, micropapillary, papillary, or solid.22 A pattern was scored as primary if more ducts showed the pattern than any other pattern, but the primary pattern did not necessarily involve the majority of ducts because many of the individual tumors contained sev- eral patterns. Comedo intraductal carcinoma was defined as intraductal carcinoma having marked central necrosis. Cribriform intraductal carcinoma was defined by rounded spaces forming acinar-like structures. Mi- cropapillary intraductal carcinoma was defined by papillary projections of epithelial cells, often without fibrovascular cores, involving multiple ducts that were mod- erately distended. Papillary intraductal carcinoma was defined by a single large duct containing papillae with thin fibrovascular cores and with cytologic features of malignancy. Solid intraductal carcinoma was defined by a solid intraductal proliferation, at times with foci of individual cell necrosis, but without marked central ne-

crosis. Because the pattern of clinging intraductal carcinoma was seen only rarely and only in conjunction with other patterns of intraductal carcinoma, clinging intraductal carcinoma was not included in the analysis of the histologic subtypes of intraductal carcinoma.

The amount of necrosis was recorded as none, focal, or marked. Necrosis was scored as focal if a minor- ity (i.e,, <50%) of the ducts contained central necrosis. Necrosis was scored as marked if the majority (i.e., >5o0/o) of the ducts showed central necrosis. Tumors in which the primary histologic architectural pattern was comedo, but in which the comedo pattern did not involve the majority of the ducts, were scored as comedo for histologic subtype, but were not scored as marked for necrosis. Nuclear grade was scored as grade 1, 2, or 3.23 Nuclear grade was scored as grade 1 for monomor- phic nuclei with fine chromatin. Nuclear grade was scored as grade 3 for pleomorphic nuclei with coarse chromatin and nucleoli. Nuclear grade was scored as grade 2 for intermediate features. For the correlation of pathologic parameters with outcome of treatment, the combination of the histologic subtype of comedo carcinoma plus nuclear grade 3 was evaluated as a separate parameter.24

The final pathology margin was defined as the pathology margin of the re-excisional biopsy, if performed, or as the pathology margin of the primary sur- gical excision, if a re-excisional biopsy was not performed. The final pathology margin was determined from review of the pathology reports in the patient records and was not re-evaluated at the time of central pathology review. The final pathology margin was scored as negative when all tumor identified was more than 2 mm from the inked margin of the biopsy specimen or when no tumor was identified in the re-excisional biopsy specimen. The final pathology margin was scored as positive when tumor was identified at the inked margin of the biopsy specimen. The final pathology margin was scored as close when tumor was identified 2 mm or less from, but not at, the inked margin of the biopsy specimen.

The Kaplan-Meier method was used to calculate actuarial curves.25 The time period was calculated from the start of definitive radiation treatment, not from the time of diagnosis of intraductal carcinoma. The median follow-up time was 84 months (mean, 88 months; range, 17-177 months). The Mantel-Cox test was used for statistical comparisons between actuarially calculated curves.26 The Cox regression model was used for multivariate analy~is.~’

For analysis of local control, a local failure was scored for a failure that occurred within the treated breast. For analysis of regional control, a regional failure was scored for a failure that occurred within the

Intraductal Carcinoma of the Breast/Solin et al. 2535

ipsilateral axillary, supraclavicular, infraclavicular, or internal mammary nodal regions. For analysis of overall survival, a failure was scored for a death of any cause. For analysis of cause-specific survival, a failure was scored only for a death secondary to carcinoma of the breast. Therefore, a patient who died of causes other than carcinoma of the breast was scored as a failure for overall survival, but not as a failure for cause- specific survival. For analysis of freedom from distant metastases, a failure was scored at the time of first evidence of metastatic disease.

Table 3. Site(s) of First Failure

No. (%)

None Local only Regional only Distant only Local and regional Local and distant Regional and distant Local, regional, and distant Other* Total

Results * A contralateral, pathologically node-positive invasive breast carcinoma followed by distant metastatic disease developed in one patient (see Solin et aI.l5 for details).

Patient and tumor characteristics are detailed in Table 2. For the tumors for which the clinical size was known, 80% (104 of 130) were 2.0 cm or smaller. The final

Table 2. Patient and Tumor Characteristics

~ - _ _ Clinical tumor size (cm)

5 2 . 0 2.1-5.0 5.1-6.0 Unknown

Negative Positive Close ( 2 ; 2 mm) Unknown

Histologic subtype Cornedo Cribriform Papillary Micropapillary Solid

1 2 3

None Focal Marked

Physical examination only* Mammography onlyt Physical examination and

Final pathology margin

Nuclear grade

Amount of necrosis

Clinical presentation

mammography Menopausal status

Premenopausal Postmenopausal Perimenouausal

NO. ( o h )

104 (60) 25 (15)

42 (24) 1 (< 1)

55 (32) 27 (16)

6 (3) 84 (49)

67 (39) 51 (30)

28 (16) 12 (7)

14 (8)

35 (20) 68 (40) 69 (40)

29 (17) 91 (53) 52 (30)

35 (20) 69 (40)

68 (40)

82 (48)

13 (81 77 (45)

Includes four patients with. positive thermography. t Includes one Datient with oositive thennomaohv.

pathology margin of the tumor excision was unknown in 49% (84 of 172) of the patients because routine ink- ing of surgcal biopsy specimens was not commonly performed during the earlier years of the study. Tumors in which the primary histologic architectural pattern was comedo, but in which the comedo pattern did not involve the majority of the ducts, were scored as comedo for histologic subtype, but were not scored as marked for necrosis (see Materials and Methods). There- fore, there were more tumors scored as having comedo histologic subtype (n = 67) than marked necrosis (n = 52). The median patient age was 51 years (mean, 51 years; range, 27-77 years).

Table 3 shows the site(s) of first failure. Fifteen of the 16 local failures were isolated first failures without evidence of simultaneous regional or distant failure. The one patient whose first site of failure was simultaneous local and regional failure had 1 of 12 positive axillary lymph nodes at the time of salvage mastectomy. No other patient had regional nodal failure at any time during the course of follow-up.

All of the 16 patients with a recurrence in the breast were rendered free of disease with initial salvage treatment. Median follow-up after salvage treatment was 42 months (mean, 41 months; range, 0-96 months). The histology at the time of local recurrence was intraductal carcinoma in nine (56%) of the patients, including one patient with associated Paget disease, and invasive carcinoma in seven (44%) of the patients. For the 16 local recurrences, the median time to recurrence was 62 months (mean, 61 months; range, 19-119 months). Two of the 15 patients with an isolated local recurrence subsequently developed distant metastatic disease, one of which occurred simultaneously with a chest wall recurrence. The histology of the local recurrence was invasive ductal carcinoma for both of these two patients.

For all 172 patients, the actuarial rates of local recurrence at 5 years and 8 years were 4% and 9%, respec-


tively. Table 4 shows the univariate analysis of pathologic parameters for local control. A comparison of the histologic subtype of comedo carcinoma versus other histologic subtypes showed no difference in the rate of local recurrence (P = 0.10). However, the combination of the histologic subtype of comedo carcinoma plus nuclear grade 3 was highly correlated with local recurrence (P = 0.009). None of the remaining pathologic parameters evaluated were significantly correlated with local control (all P 2 0.10). Figure 1 shows curves for local recurrence as a function of comedo versus noncomedo intraductal carcinoma (Fig. 1A) and the presence versus the absence of the combination of the histologic subtype of comedo carcinoma plus nuclear grade 3 (Fig. 1B).

The time to local recurrence was shorter for patients with the combination of the histologic subtype of comedo carcinoma plus nuclear grade 3 than for patients without this combination (median, 38 months versus 78 months, respectively; mean, 42 months versus 80 months, respectively; range, 19-67 months versus 23- 119 months, respectively). However, the location of the local recurrence relative to the primary tumor (five of eight versus four of eight, respectively, for either true recurrence or marginal miss'') was similar between the two groups. The histology of the local recurrence was invasive carcinoma in two of eight patients versus five

of eight patients, respectively. For the 44 patients with the histologic subtype of comedo carcinoma plus nuclear grade 3, a local recurrence subsequently developed in 0% (0 of 5) of the patients with a negative final pathology margin, 18% (2 of 11) of the patients with a positive final pathology margin, and 21% (6 of 28) of the patients with an unknown final pathology margin.

A multivariate analysisz7 that included all 172 patients was performed for local control relative to pathologic characteristics of the primary intraductal carcinoma. The multivariate analysis included the following parameters:

1. 2. 3.

4. 5.

histologic subtype of the primary tumor; nuclear grade; the combination of the histologic subtype of comedo carcinoma plus nuclear grade 3; amount of necrosis; and final pathology margin.

With the level of significance for the multivariate analysis set at P = 0.05, only the combination of the histologic subtype of comedo carcinoma plus nuclear grade 3 was significantly correlated with local control (P = 0.017; relative risk, 3.4 with 95% confidence intervals of 1.3-9.1). None of the remaining pathologic parameters was significantly correlated with local control (all P 2 0.18).

Table 4. Univariate Analysis of Pathologic Parameters for Local Control

5-yr 8-yr Total no.

Total no. of patients Yo actuarial No. of Yo actuarial No. of of patients with local local patients local patients at risk recurrence recurrence a t risk recurrence a t risk P value'

Histologic subtype 0.10

Nuclear grade 0.20

Comedo 67 9 8 48 14 19 Other 105 7 2 92 6 36

1 35 3 3 29 12 10 2 68 4 1 57 4 25 3 69 9 7 54 13 20

Histologic subtype of comedo carcinoma plus nuclear grade 3 0.009

Yes 44 8 11 32 20 11 No 128 8 2 108 5 44

None 29 3 0 22 t t Amount of necrosis 0.70

Focal 91 7 3 80 5 32 Marked 52 6 8 38 13 17

Negative 55 2 2 43 5 15

Unknown 84 10 5 72 11 29

Final pathology margin 0.29

Positive/close 33 4 6 25 10 11

* Mantel-Cox testz6 t Too few patients at risk (ie., < 10) to report data.

Intraductal Carcinoma of the Breast/Solin et al. 2537

NUMBER AT RISK

67 67 66 62 58 48 38 26 19- 105 105 104 102 97 92 75 53 36--

50 w U 3 40

3 J 4 30

0 -I g 20

E 10

u 0

a

~

A

P = .I0

Comedo 7 14%

I /TF I r'---- I ,6Yo Noncomedo

--,,,5----- 0

0 1 2 3 4 5 6 7 8

YEARS

NUMBER AT RISK 44 44 43 40 36 32 23 15 11- 128 128 127 124 119 108 90 64 44--

50- 6 w U 2 40- 2 J 0 a 30{

9

2oi nuclear grade 3

720yo " -

0 1 2 3 4 5 6 7 8

YEARS Figure 1. Local recurrence. (A) Comparison of the histologic subtype of comedo (solid line) versus noncomedo (dashed line) intraductal carcinoma. (B) Comparison of the presence (solid line) versus the absence (dashed line) of the combination of the histologic subtype of comedo carcinoma plus nuclear grade 3.

For all 172 patients, the actuarial rates of overall survival at 5 years and 8 years were 98% and 96%, respectively. Table 5 shows the univariate analysis of

pathologic parameters for overall survival. None of the pathologic parameters was significantly correlated with overall survival (all P L 0.16).

For all 172 patients, the actuarial rates of cause- specific survival at 5 years and 8 years were 99% and 96%, respectively. Table 6 shows the univariate analysis of pathologic parameters for cause-specific survival. None of the pathologic parameters was significantly correlated with cause-specific survival (all P 2 0.13).

For all 172 patients, the actuarial rates of freedom from distant metastases at 5 years and 8 years were 99% and 96%, respectively. Table 7 shows the univariate analysis of pathologic parameters for freedom from distant metastases. None of the pathologic parameters was significantly correlated with freedom from distant metastases (all P 2 0.13).

Of the 172 patients, 7 (4%) had a subsequent contralateral carcinoma of the breast. The median interval to development of a contralateral breast carcinoma was 41 months (range, 15-95 months; mean, 51 months). The histologic subtype of the original ipsilateral intraductal carcinoma was comedo in three patients, papillary in two patients, cribriform in one patient, and solid in one patient. The contralateral breast carcinoma was invasive ductal carcinoma (with or without associated intraductal carcinoma) in four patients, intraductal carcinoma in one patient, invasive carcinoma (not other- wise specified) in one patient, and unknown in one patient.

Recent studies have prospectively identified highly selected patients with intraductal carcinoma of the breast for treatment with excision only without breast irradiation.ll,12,18,19 Only 12 (7%) of the patients in the current study potentially would have met the selection criteria for treatment with excision only (i.e., clinical tumor size of 2.5 cm or smaller, detection by mammo- graphic microcalcifications only with a negative physical examination, and pathologically confirmed negative margins of resection). There were no failures in this group of 12 treated patients in the current study, with a median follow-up time of 79 months (mean, 88 months; range, 28-149 months). The histologic subtype of the intraductal carcinoma was comedo in four patients, cribriform in four patients, micropapillary in three patients, and solid in one patient. Only one of these 12 patients presented with the combination of the histologic subtype of comedo carcinoma plus nuclear grade 3.

Discussion

Recent analyses have suggested that there may be important biologic differences associated with various histologic characteristics of ductal carcinoma in situ (intraductal carcinoma). Comedo intraductal carcinoma and


Table 5. Univariate Analysis of Pathologic Parameters for Overall Survival

5-yr 8-yr

Total no. Total Yo actuarial No. of Yo actuarial No. of of patients no. of overall patients overall pa tien t s at risk deaths survival a t risk survival a t risk P value*

Histologic subtype Comedo Other

Nuclear grade 1 2 3

0.16 67 4 97 52 91 23

105 2 99 94 99 39

35 2 100 30 92 12 68 2 97 58 97 26 69 2 98 58 96 24

0.72


Yes 44 2 97 36 94 15 No 128 4 98 110 96 47

Amount of necrosis 0.45 None 29 2 100 22 t t Focal 91 2 98 83 98 34 Marked 52 2 98 41 95 19

Negative 55 2 98 44 98 16

Unknown 84 3 99 75 94 34



* Mantel-Cox test.z6 t Too few patients at risk (i.e., < 10) to report data.

Table 6. Univariate Analysis of Pathologic Parameters for Cause-Specific Survival

5-yr 8-yr

Total no. Total no. of % actuarial No. of YO actuarial No. of of patients deaths from cause-specific patients cause-specific patients at risk breast cancer survival a t risk survival a t risk P value*

Histologic subtype 0.13 Comedo 67 3 98 52 92 23 Other 105 1 99 94 99 39

1 35 1 100 30 92 12 2 68 1 99 58 99 26 3 69 2 98 58 96 24

Nuclear grade 0.81


Yes 44 2 97 36 94 15 No 128 2 99 110 97 47

None 29 1 100 22 t t Focal 91 1 99 83 99 34 Marked 52 2 98 41 95 19

Negative 55 0 100 44 100 16

Unknown 84 3 99 75 94 34

Amount of necrosis 0.53



* Mantel-Cox test.26 t Too few patients at risk (i,e,, < 10) to report data.

Intraductal Carcinoma of the Breast/Solin e t al . 2539

Table 7. Univariate Analysis of Pathologic Parameters for Freedom From Distant Metastases

5-yr 8-yr

Total no. % actuarial Yo actuarial Total no. of patients freedom No. of freedom No. of of patients with distant from distant patients from distant patients at risk metastases metastases at risk metastases a t risk P value*

Histologic subtype Comedo Other

Nuclear grade 1 2 3

0.13 67 3 98 52 92 23

105 1 99 94 99 39

35 1 100 30 92 12 68 1 99 58 99 26 69 2 98 58 96 24

0.81


Yes 44 2 98 36 94 15 No 128 2 99 110 97 47

Amount of necrosis 0.53 None 29 1 100 22 t t Focal 91 1 99 83 99 34 Marked 52 2 98 41 95 19

Final pathology margin 0.44 Negative 55 0 100 44 100 16

Unknown 84 3 99 75 95 34 Positive/close 33 1 97 27 97 1 2

* Mantel-Cox test." t Too few patients at risk ( ~ e , < 10) to report data

high nuclear grade have been associated with more ag- gressive lesion^.^^^^'^^'^^'^^^^ F isher et aL3 reported that poor nuclear grade was associated with comedo necrosis. Lagios et al.'8,19 have reported 79 patients treated with local excision without breast irradiation. Patho- logic features predicting for local recurrence were found to be comedo carcinoma, necrosis, and high nuclear grade. The incidence of local recurrence after excision without radiation was 25% (9 of 36) for high-grade lesions, 10% (1 of 10) for intermediate-grade lesions, and 0% (0 of 33) for low-grade lesions." Lagios" also found that high nuclear grade lesions were more commonly aneuploid than were intermediate or low nuclear grade lesions. Carpenter et al.29 reported a significantly increased incidence of aneuploid lesions for ductal carcinoma in situ with microinvasion compared with ductal carcinoma in situ or proliferative atypia ( P < 0.001).

The term comedo carcinoma in situ or comedo intraductal carcinoma has been used to denote intraductal carcinomas exhibiting prominent central necrosis. Ne- crosis also is commonly observed in ducts filled with solid, cribriform, or micropapillary patterns of intraductal carcinoma. For this reason, Moriya and Silver- bergz4 have proposed redefining the term comedo carcinoma to include lesions with central necrosis and high nuclear grade in conjunction with any other pattern.

Other studies also have noted that the comedo subtype appears to be distinctive, which supports reporting these lesions ~ e p a r a t e l y . ' ~ , ' ~ , ~ ~ Comedo carcinomas are more frequently aneuploid and exhibit a higher thymi- dine labeling index than do other subtypes of intraductal c a r c i n ~ m a . ' ~ ~ ~ ~ Gene analyses have demonstrated higher proportions of c-erb-B2 overexpression for comedo intraductal carcinoma than for other subtypes of intraductal c a r c i n ~ m a . ~ ~ , ~ ~

Patchefsky et a1." reported the pathologic findings in 55 mastectomy specimens for patients with intraductal carcinoma or microinvasive ductal carcinoma. Co- medo carcinoma showed a significantly higher incidence of high nuclear grade (89%; 17 of 19) and microinvasion (63%; 12 of 19) than did other subtypes of intraductal carcinoma ( P < 0.01). For small lesions, defined as involving fewer than 50 ducts, a higher incidence of microinvasion was seen for the comedo carcinomas (42%; 5 of 12) than for other histologic subtypes ( P < 0.01), suggesting a propensity for invasion by comedo carcinoma, even for smaller tumors. Of the microinvasive tumors, comedo carcinoma was found in 57% (12 of 21).

Silverstein et al." reported the outcome of 227 patients with intraductal carcinoma treated with mastectomy, excision plus radiation, or excision without radia-


tion. Of the 103 patients treated with excision plus radiation, 52 had comedo intraductal carcinoma, and 51 had noncomedo intraductal carcinoma. Although there were more recurrences for comedo lesions than for noncomedo lesions (1 3% versus 6%, respectively), the number of deaths was similar between the two groups (2% versus O%, respectively). Only 26 patients were treated with excision only without radiation. Of the nine patients treated with excision only for comedo intraductal carcinoma, the mean tumor size was 0.9 cm, and there were no recurrences and no deaths in this subgroup of patients. Of the 17 noncomedo carcinomas treated with excision only, the mean tumor size was 1.1 cm, and there were two recurrences and no deaths.

The current study has analyzed pathologic parameters relative to 8-year outcomes for local control, survival, and freedom from distant metastases in women treated with breast-conserving surgery and definitive irradiation for ductal carcinoma in situ (intraductal carcinoma) of the breast. Potential limitations of the current study include the limited number of slides evaluated per patient (median, 3; mean, 6), and the fact that pathology slides could not be obtained for central pathology review for a substantial number of patients from the previously reported study.I5 In general, the slides referred for central pathology review consisted of representative slides for each individual patient. Therefore, the results of the current study must be interpreted in light of the fact that representative slides, not all slides, underwent central pathology review and that a substantial number of patients did not have slides available for central pathology review. Nonetheless, analysis of pathologic parameters relative to local control showed that only the histologic subtype of comedo carcinoma plus nuclear grade 3 was significantly correlated with local control (Table 4 and Fig. 1B). None of the pathologic parameters evaluated was significantly correlated with outcome in terms of overall survival (Table 5), cause-specific survival (Table 6), or freedom from distant metastases (Table 7).

For the 16 patients with local recurrence, follow-up after salvage treatment was short (median, 42 months; range, 0-96 months). Because one of the major issues for breast-conservation treatment of intraductal carcinoma is the long-term ability to salvage local recurrences, longer follow-up is needed to confirm that the influence of the pathologic features of the primary tumor on local recurrence does not translate into a difference in survival or freedom from distant metastases.

The time course of local recurrence for patients with comedo intraductal carcinoma is shorter than for patients with other histologic subtypes of intraductal carcinoma. Local recurrences after the treatment of comedo intraductal carcinoma tend to occur earlier than

do local recurrences after the treatment of noncomedo intraductal carcinoma. Of note is that invasive carcinoma was found in only two of the eight local recurrences for patients with the combination of comedo carcinoma plus nuclear grade 3 in the current study. For patients in the current study with favorable histologic characteristics of the intraductal carcinoma, longer follow-up may be required to evaluate the potential for late failures.

Two early reports by Silverstein et al.'2,13 found that there was an increased incidence of local recurrence for patients treated with mastectomy or radiation therapy for comedo intraductal carcinoma (with or without microinvasion) compared with patients with noncomedo intraductal carcinoma (with or without microinvasion) with median follow-up times of 45-51 months. In one of the s t ~ d i e s , ' ~ the difference was sta- tistically significant (P = 0.04). However, with longer median follow-up of 56 months, a recent update of the experience by Silverstein et al." found that there was no difference (P = 0.47) between the rates of local recurrence for comedo versus noncomedo intraductal carcinoma (without microinvasion). The three studies by Silverstein et al.11-13 suggest that the early pattern of failure of comedo intraductal carcinoma may lead to early differences in the rate of local recurrence that may disappear with longer follow-up time. The findings in the current study of an earlier pattern of failure for comedo intraductal carcinoma compared with noncomedo intraductal carcinoma are similar to the findings of Silverstein et al.11-13 For the patients in the current study, additional follow-up is needed to determine if the differences in rate of local recurrence for patients with comedo intraductal carcinoma plus nuclear grade 3 (Table 4 and Fig. 1B) continue to be present or disappear with longer follow-up, as was found in the studies by Silverstein et al.11-13

Comparison of the current study with other reports in the literature of the treatment of intraductal carcinoma of the breast is difficult. An important influence on the outcome of treatment is patient selection. For the patient with intraductal carcinoma on initial breast biopsy, there is a known risk of occult microinvasive carcinoma in the remainder of the breast found in the mastectomy specimen. Although the patient with occult microinvasive carcinoma is excluded from reports of treatment with mastectomy, the patient with occult microinvasive carcinoma is included in reports of treatment with breast conservation surgery with or without definitive irradiation. Patients treated with excision only without definitive irradiation are highly selected for favorable tumor characteristics (e.g., small tumor size and pathologically confirmed negative margins of excision). Patients with unfavorable tumor characteris-

Intraductal Carcinoma of the Breast/SoZin e t al. 2541

tics (e.g., large tumor size, positive margins of excision, diffuse microcalcifications, or gross multicentric disease) are generally treated with mastectomy. In a comparison of 227 patients with intraductal carcinoma of the breast reported by Silverstein et al.," differences were found among the three treatment groups (mastectomy versus excision plus radiation versus excision without radiation) for mean tumor size (3.7 cm versus 1.4 cm versus 1.0 cm, respectively), positive margins on initial breast biopsy (68% versus 31% versus 31%, respectively), and percentage of nonpalpable lesions (62% versus 86% versus 85%, respectively).

Given the difficulty of comparing studies in the literature, the role of radiation treatment in the care of the patient with an excisional biopsy that shows intraductal carcinoma remains to be fully established. Multiple prospective, randomized trials that compare excision alone with excision plus definitive breast irradiation are being conducted33 and will help to define the role of breast irradiation in the patient with intraductal carcinoma on breast biopsy. In the study of Lagios," the incidence of local recurrence after treatment with excision alone without breast irradiation was reported to be 25% (9 of 36) for high-grade lesions, 10% (1 of 10) for intermediate-grade lesions, and 0% (0 of 33) for low-grade lesions, respectively. In the current study, only 12 (7%) of the 172 patients would have met the patient selection criteria of Lagios et al.,"*19 and the local recurrence rate was 0% for these 12 patients. Because there is a high incidence of local recurrence in high-grade lesions treated with excision only without breast irradiation, definitive breast irradiation appears to be indicated in high-risk groups to reduce the risk of local recurrence. In lower-risk groups, the low incidence of local recurrence after excision only without breast irradiation sug- gests that the role of definitive breast irradiation treatment is uncertain, although there were no local recurrences for the small number of comparable patients in the current study treated with definitive breast irradiation. Pathologic analyses of prospective, randomized trials will help to define the role of radiation after excision only for various histologic subsets of intraductal carcinoma. 33

In summary, for patients with ductal carcinoma in situ treated with breast-conserving surgery and definitive irradiation, the current study has demonstrated that there are important differences in the rate of local recurrence based on the pathologic characteristics of the primary tumor. However, the pathologic features of the primary tumor did not have an impact on outcome, as measured by survival or freedom from distant metastases. Careful follow-up of patients at increased risk for local recurrence is warranted because of the potential ability to salvage patients with local recurrence.

References

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

Bornstein BA, Recht A, Connolly JL, Schnitt SJ, Cady B, Kouf- man C, et al. Results of treating ductal carcinoma in situ of the breast with conservative surgery and radiation therapy. Cancer

Findlay P, Goodman R. Radiation therapy for treatment of intraductal carcinoma of the breast. Am ] Clin Oncol 1983; 6:281-5. Fisher ER, Leeming R, Anderson 5, Redmond C, Fisher 8, collab- orating NSABP investigators. Conservative management of intraductal carcinoma (DCIS) of the breast. ] Surg Oncol 1991;

Fowble B. Intraductal noninvasive breast cancer: a comparison of three local treatments. Oncology 1989; 3:51-8. Fowble BL, Solin LJ, Goodman RL. Results of conservative surgery and radiation for intraductal noninvasive breast cancer. [abstract]. Am ] Clin Oncol 1987; 10:110-1. Haffty BG, Peschel RE, Papadopoulos D, Pathare P. Radiation therapy for ductal carcinoma in situ of the breast. Conn Med

Kurtz JM, Jacquemier J, Torhorst J, Spitalier J-M, Amalric R, Hunig R, et al. Conservation therapy for breast cancers other than infiltrating ductal carcinoma. Cancer 1989; 63:1630-5. Kuske RR, Bean JM, Garcia DM, Perez CA, Andriole D, Philpott G, et al. Breast conservation therapy for intraductal carcinoma of the breast. Int ] Radiat Oncol Biol Phys. In press. McCormick B, Rosen PP, Kinne D, Cox L, Yahalom J. Duct carcinoma in situ of the breast: an analysis of local control after conservation surgery and radiotherapy. Jnt ] Radiat Oncol Bid Phys

Recht A, Danoff B, Solin LJ, Schnitt 5, Connolly J, Botnick L, et al. lntraductal carcinoma of the breast: results of treatment with excisional biopsy and irradiation. ] Clin Oncol 1985; 3:1339-43. Silverstein MJ, Cohlan BF, Gierson ED, Furmanski M, Gama- gami P, Colburn WJ, et al. Ductal carcinoma in situ: 227 cases without microinvasion. Eur ] Cancer 1992; 28:630-4. Silverstein MJ, Waisman JR, Gamagami P, Gierson ED, Colburn WJ, Rosser RJ, et al. lntraductal carcinoma of the breast (208 cases): clinical factors influencing treatment choice. Cancer

Silverstein MJ, Waisman JR, Gierson ED, Colburn W, Gama- gami P, Lewinsky BS. Radiation therapy for intraductal carcinoma: is it an equal alternative? Arch Surg 1991; 126:424-8. Solin LJ, Fowble BL, Schultz DJ, Yeh I-T, Kowalyshyn MJ, Goodman RL. Definitive irradiation for intraductal carcinoma of the breast. In t J Radiat Oncol Biol Phys 1990; 19:843-50. Solin LJ, Recht A, Fourquet A, Kurtz J, Kuske R, McNeese M, et al. Ten-year results of breast-conserving surgery and definitive irradiation for intraductal carcinoma (ductal carcinoma in situ) of the breast. Cancer 1991; 68:2337-44. Stotter AT, McNeese M, Oswald MJ, Ames FC, Romsdahl MM. The role of limited surgery with irradiation in primary treatment of ductal in situ breast cancer. [nt ] Radiat Oncol Biol Phys 1990;

Zafrani B, Fourquet A, Vilcoq JR, Legal M, Calle R. Conservative management of intraductal breast carcinoma with tumorectomy and radiation therapy. Cancer 1986; 57:1299-301. Lagios MD. Duct carcinoma in situ: pathology and treatment. Surg Clin North Am 1990; 70:853-71. Lagios MD, Margolin FR, Westdahl PR, Rose MR. Mammogra- phically detected duct carcinoma in situ: frequency of local recurrence following tylectomy and prognostic effect of nuclear grade on local recurrence. Cancer 1989; 63:618-24. Patchefsky AS, Schwartz GF, Finkelstein SD, Prestipino A,

1991; 677-13.

471139-47.

1990; 54~482-4.

1992; 21~289-92.

1990; 66:102-8.

18:283-7.


21.

22.

23.

24.

25.

26.

27.

Sohn SE, Singer JS, et al. Heterogeneity of intraductal carcinoma of the breast. Cancer 1989; 63:731-41. American Joint Committee on Cancer. Manual for staging of cancer. 3rd ed. Philadelphia: JB Lippincott, 1988:145-50. Page DL, Anderson TJ, with Rogers LW. Carcinoma in situ (CIS). In: Page DL, Anderson TJ. Diagnostic histopathology of the breast. New York: Churchill Livingstone, 1987:157-92. Bloom HJG, Richardson WW. Histological grading and progno- sis in breast cancer. Br Cancer 1957; 11:359-77. Moriya T, Silverberg SG. Intraductal carcinoma (DCIS) of the breast: characterization and significance of different histologic patterns [abstract]. Presented at the Annual Meeting of the United States and Canadian Academy of Pathology, March 1991. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. ] A m Stat Assoc 1958; 53:457-81. Mantel N. Evaluation of survival data and two new rank order statistics arising in its consideration. Cancer Chemother Rep 1966;

Kalbfleisch JD, Prentice RL. The statistical analysis of failure time data. New York: John Wiley and Sons, 1980.

5 0 3 63-70.

28. Recht A, Silver B, Schnitt S, Connolly J, Hellman S, Hams JR. Breast relapse following primary radiation therapy for early breast cancer: I. classification, frequency and salvage. Int Ra- diat Oncol Biol Phys 1985; 11:1271-6.

29. Carpenter R, Gibbs N, Matthews J, Cooke T. Importance of cel- lular DNA content in pre-malignant breast disease and pre-invasive carcinoma of the female breast. Br ] Surg 1987; 74:905-6.

30. Meyer JS. Cell kinetics of histologic variants of in situ breast carcinoma. Breast Cancer Res Treat 1986; 7:171-80.

31. Lodato RF, Maguire HC, Greene MI, Weiner DB, LiVolsi VA. Immunohistochemical evaluation of c-erbB-2 oncogene expres- sion in ductal carcinoma in situ and atypical ductal hyperplasia of the breast. Mod Pathol 1990; 3:449-54.

32. van de Vijver MJ, Peterse JL, Mooi WJ, Wisman P, Lomans J, Dalesio 0, et al. Neu-protein overexpression in breast cancer: association with comedo-type ductal carcinoma in situ and limited prognostic value in stage I1 breast cancer. N Engl ] M e d 1988;

33. van Dongen ]A, Holland R, Peterse JL, Fentiman IS, Lagios MD, Millis RR, et al. Ductal carcinoma in-situ of the breast: Second EORTC Consensus Meeting. Eur ] Cancer 1992; 28:626-9.

319:1239-45.

Documents

Ductal carcinoma in situ (intraductal carcinoma) of the breast treated with breast-conserving surgery and definitive irradiation correlation of pathologic parameters with outcome of