Breast Cancer Research and Treatment 17: 9-13, 1990. © 1990 Kluwer Academic Publishers. Printed in the Netherlands.

Report

N u c l e a r p o l y m o r p h i s m - a p r o g n o s t i c p a r a m e t e r to e v a l u a t e local r e c u r r e n c e o f f e m a l e breast c a n c e r

H.R. Rosen 1, M. Stierer I and R. Weber 2

1Department of Surgery and elnstitute of Pathology, Hanusch Hospital, Vienna, Austria

Key words: local recurrence, breast cancer, prognostic factors, nuclear polymorphism Summary Between 1980 and 1986 676 patients underwent surgery for primary breast cancer. Of these, 35 patients developed locoregional recurrence. Retrospective analysis of the spontaneous postoperative development revealed 2 groups: group A had subsequent distant metastases, group B was tumor-free after surgical treatment of local recurrence. Analysis of the commonly employed characterization criteria of primary tumors (tumor size, lymph node involvement, estrogen receptors, histologic grading of primary tumors, and excised locoregional recurrence) showed no statistically significant difference between the two groups. However, a more detailed differentiation of the subcriteria for the histologic grading according to Bloom and Richardson revealed a prevalence of anaplastic nuclei in the primary tumors of group A (9/11). Even in this small patient population the parameter of nuclear polymorphism revealed a highly significant statistical difference between the two groups.

Introduction While breast cancer has been established as a primary systemic disease, local tumor control continues to be a major therapeutic goal [1]. The patient is emotionally strained by the re-appearance of tumor tissue in the area of the surgical intervention and/or irradiation, and the resulting deterioration of prognosis. Only in about 10% of patients [2-4] does recurrence not lead to synchronous or metachronous dissemination and consequent deterioration of prognosis. 5-year survival rates are reported to be a mere 30% both for local recurrence after radical surgery (skin, subcutis, chest wall) and for regional recurrence (metastases in the regional lymph drainage system) [5]. The etiology of locoregional recurrence is probably linked to two major factors: (1) nests of tumor

cells left in the remaining skin-subcutis layer [6]; and (2) primary manifestation of generalization [51. It is well-known that the incidence of locoregional recurrence increases in relation to the size of the primary tumor and to the degree of lymph node involvement. However, we cannot give any reliable statement on the potential development of metachronous metastases in the presence of locoregional recurrence. We have analysed our patient population to evaluate parameters related to the primary tumor that are able to predict generalization following local recurrence after radical mastectomy.

Address for offprints: H.R. Rosen, Department of Surgery, Hanusch Hospital, Heinrich Collinstr 30, A - 1140 Vienna, Austria

10

H R Rosen et al.

PER CENT 100 -

I-L~ L'J I

. . . . I

"1

GROUP B

lh J J . . . . I III

~

. . . . .

~-II.

. . . . . . .

J.

II

50-

GROUPA o

2~

4'8

~2

9~

110 MONTHS

Fig. 1. Survival after recurrence. Group A: 13,5 months 75%. Group B: 30.5 months 75%. p < 0.01. Group A: Local recurrence with

subsequent metastases, Group B: Local recurrence without subsequent metastases.

Patients and methods

From 1980 to 1986, 676 patients with primary breast cancer were operated on with curative intention at the Department of Surgery of the Hanusch Hospital in Vienna. During follow-up, 35 patients with an average age of 60 years (36-87 years) had locoregional recurrences after modified radical mastectomy (or simple mastectomy in three cases, respectively). All patients were treated surgically for local recurrence without any further adjuvant therapy. Retro-

spective analysis of the spontaneous development after extirpation of local recurrence revealed two groups of patients, which differed significantly in terms of survival after recurrence and overall survival (Table 1): (1) group A developed subsequent distant metastases after detection of local recurrence and surgical removal; (2) group B also underwent mere extirpation of local recurrence, but demonstrated no further manifestation of the primary disease by the control date (Dec. 31, 1989).

Table 1. Distribution of locoregional recurrence in the two prognostic groups 35 local recurrences (1980-1986)

Median age: 60 (36-87)

number median age local recurrence lymph node recurrence median observation disease-free interval (median) Group A = LR with subsequent metastases. Group B = 'simple' LR without subsequent metastases.

Group A

Group B

11 54 (36-80) 9 2 38 months (20-50) 19 months

24 59 (41-87) 12 12 68 months (25--115) 20 months

Nuclear polymorphism and prognosis

11

PER CENT

100

I, L"~''------'I GROUP B 50

[ o

~ GROUPA 72

4'8

2',

~

11o MONTHS

Fig. 2. Overall survival. Group A: 40.3 months 75%, Group B: 53.3 months 75%. p < 0.01. Group A: Local recurrence with subsequent metastases, Group B: Local recurrence without subsequent metastases.

In the patients treated by simple mastectomy, even though the axilla was only 'clinically' tumor-free, recurrences occurred in the scar rather than the axilla. All lymph node recurrences were encountered at more than 24 months after primary surgery. While in the group with simple local recurrence (Group B) 75% of the patients survived 30.5 months after detection of recurrence, survival was only 13.5 months in the group with subsequent metastases (Group A) (Fig. 1). While in Group A Table 2. Lymph node involvement and size of the primary tumors Primary tumor status (Pathologic staging) Group A T1N0 T1 N 1 T2N0 T2N1 T3N0 T3N1 T1Nx N+:N0 8:2

Group B 1 5 1 1 2 1 n.s.

T1N0 T1 N 1 T2N0 T2N1 T3N0 T3N1 T1Nx N+:N0 13:8

5 8 2 4 1 2 2

N x = lymph node involvement clinically evaluated as N 0; no axillary histology available; simple mastectomy performed.

all patients died secondary to the formation of distant metastases, the cause of death was unrelated to breast cancer in Group B patients (pulmonary embolism, stroke, myocardial infarction). The overall 75% survival from primary surgery was 53.3 months with simple local recurrence compared to 40.3 months with local recurrence and subsequent metastases (Fig. 2). We have attempted to identify a prognostic factor for local recurrence in these prognostically different groups with the commonly employed characterization criteria of their primary tumors. Therefore, all patient records were retrospectively analyzed regarding tumor size, axillary lymph node status, and estrogen receptor levels, if available. All histologic sections of primary tumors as well as of excised recurrences were retrospectively investigated by a single pathologist who had no knowl-

Table 3. Receptor status of the primary tumor as determined in 19 of 35 cases (54.2%); cut-off > 10 fmol/mg cytoso! protein Estrogen receptor status

Group A

Group B

positive negative ?

1 4 6

4 10 10

12

HR Rosen et al.

Table 4. Histological grading according to Bloom and Richard-

son Factors

Differentiation

Nuclear polymorphism

Mitotic status

glandular

low-grade

mixed

medium-grade

1 mitosis/HPF 1 2 mitoses/HPF 2 3 and m o r e / H P F 3

solid/disseminated high-grade

Points

Grading according to total points (differentiation + nuclear polymorphism + mitotic status) Grade I

Grade II

3-5 points

6 - 7 points

Grade I I I 8-9 points

edge of the patient's outcome to formulate the histologic grading and grading subfactors. At our Department of Pathology, the histologic grading of primary tumors is performed according to the method of Bloom and Richardson [7]. This method has proved to be an excellent tool to evaluate the degree of malignancy. It is based on the graded histologic and cytologic analysis of individual tumor structures in several places. Three parameters are used for the investigation: differentiation (glandular - tubular, mixed, or solid), nuclear polymorphism (low-, medium- or high-grade), and the number of mitoses at 40-fold magnification. Each parameter is assigned 1-3 points, and the overall count of all three evaluated criteria constitutes the grading (I to III) according to Bloom and Richardson (Table 4).

Statistics

For comparison of percentage results the Chi square test was used. To compare survival data the Mantel Cox test was applied, p < 0.01 was chosen as the level of significance.

Results Tables 2 and 3 list primary tumor size and lymph node involvement, as well as the estrogen receptor levels which were determined in 54% of the patients (19/35). Comparison of positive and negative lymph node involvement yielded a ratio of 8 : 1 3 (Group B) versus 2 : 8 (Group A). Analysis of these data with the Chi-square test produced no statistical significance. The results of the histologic grading of primary tumors as well as of excised recurrences are illustrated in Table 5. Use of the histologic grading as a prognostic factor to predict the development of metastases following local recurrence yielded no statistical significance. When, however, the gradTable 6. Analysis of subfactors of histologic grading (according to Bloom and Richardson)

Differentiation glandular mixed solid/disseminated

Group A

Group B

PT

LR

PT

0 10 1

1 9 1

1

2

19 4

18 4

LR

n.s.

Table 5. Grade of the primary tumor (PT) and the local recurrence (LR)

Grading

Group A PT

GI G II G III

1 8 2

Group B LR 4 6 1

PT 3 17 4

LR 5 15 4

G I = low-grade malignancy, G II = medium-grade malignancy, G I I I = high-grade malignancy.

Nuclear polymorphism low-grade medium-grade high-grade Number of mitoses/HPF 1 mitosis/high power field 2 mitoses/-'3 and more/-'-

0 2 9*

1 7 3

2 20 2*

4 15 5

1 8 2

1 7 3

3 18 3

4 14 6

n.s.

PT = primary tumor, LR = local recurrence. * 9/11 primary tumors in group A showed high-grade nuclear polymorphism versus 2/24 in group B (p < 0.001).

Nuclear polymorphism and prognosis ing data are classified into their individual components (tubular differentiation, nuclear polymorphism, and number of mitoses), the component 'nuclear polymorphism' of primary tumors in our patient population reveals a pronounced prevalence of grade 3 polymorphism in patients with subsequent metastases (Group A) (Table 6). Despite the relative paucity of data, comparison with the Chi-square test revealed a statistically highly significant difference (p < 0.001). The two other subfactors were not significant (Table 6).

Discussion

In recent years, tumor kinetic parameters have increasingly become the focus of attention, and recent results reported by McGuire on the importance of ploidy and the number of cells in the S-phase as potential prognostic factors have created a new line of thought [8]. However, despite the relatively small patient population we have investigated here, we feel that our histopathologic observations deserve further notice. Without the technical and personal input required for cytophotometric analysis, we have employed an optical microscopic parameter, i.e. the semiquantitative evaluation of nuclear polymorphism, as a prognostic factor to detect potential generalization after locoregional recurrence. Thus a high-risk group could be identified already at the time of the first intervention. The method could also serve to establish the indication for systemic adjuvant therapy after local surgical management, in the presence of massive nuclear polymorphism. Buzdar [3] treated 68 patients after surgical management of local recurrence by administration of FAC-BCG as adjuvant therapy. 66% of the patients were recurrence-free for 2 years, compared with a historical control population of 38 untreated

13

patients, where only 25% remained recurrencefree for 2 years. Beck [1] recommended the application of an adjuvant drug therapy at the time of primary surgery for early recurrence (less than 24 months after primary surgery) and with advanced tumors. Compared with the findings reported by Buzdar, the survival in his population was longer. Although the present analysis is based on a relatively small number of cases, the statistical significance of nuclear polymorphism in primary breast cancer as an indicator of metachronous generalization after locoregional recurrence is striking. In clinical routine, determination of this factor requires minimal input in terms of time and labor. Confirmation of our results in a larger series of patients is mandatory.

References 1. Beck TM, Hart NE, Woodard DA: Local or regionally recurrent carcinoma of the breast. Results of therapy in 121 patients. J Clin Oncol 1: 41Y46, 1983 2. Bruce J, Carter DC, Fraser J: Patterns of recurrent disease in breast cancer. Lancet i: 433-437, 1970 3. Buzdar AU, Blumenschein GR, Smith TL: Adjuvant chemoimmunotherapy following regional therapy for isolated recurrences of breast cancer (Stage IV NED). J Surg Oncol 12: 27-32, 1979 4. Clark RM: Alternatives to mastectomy-the Princess Margaret Hospital experience. In: Harris JR, Hellman S, Silen W (eds) Conservative Management of Breast Cancer. JB Lippincott Company, Philadelphia, 1983, p 35 5. Dao TL, Nemoto T: The clinical significance of skin recurrence after radical mastectomy in women with cancer of the breast. Surg Gynecol Obstet 117: 447-453, 1963 6. Peters T, Donegan WL, Burg EA: Minimal breast cancer: A clinical appraisal. Ann Surg 186: 704-710, 1977 7. Bloom HJG, Richardson WW: Histological grading and prognosis in breast carcinoma. Br J Cancer 11: 359-364, 1957 8. McGuire WL, Dressier LG: The emerging impact of flow cytometry in predicting recurrence and survival in breast cancer patients. J Natl Cancer Inst 75: 405Mll, 1985