Annals of Medicine
ISSN: 0785-3890 (Print) 1365-2060 (Online) Journal homepage: http://www.tandfonline.com/loi/iann20
Hormone Receptors as Prognostic Factors in Female: Breast Cancer Sirpa Aaltomaa, Pertti Lipponen, Matti Eskelinen, Veli-Matti Kosma, Sinikka Marin, Esko Alhava & Kari Syrjänen To cite this article: Sirpa Aaltomaa, Pertti Lipponen, Matti Eskelinen, Veli-Matti Kosma, Sinikka Marin, Esko Alhava & Kari Syrjänen (1991) Hormone Receptors as Prognostic Factors in Female: Breast Cancer, Annals of Medicine, 23:6, 643-648, DOI: 10.3109/07853899109148097 To link to this article: http://dx.doi.org/10.3109/07853899109148097
Published online: 08 Jul 2009.
Submit your article to this journal
Article views: 14
View related articles
Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=iann20 Download by: [Universite Laval]
Date: 02 April 2016, At: 04:21
Original Article
Hormone Receptors as Prognostic Factors in Female Breast Cancer Sirpa Aaltomaa' , Pertti Lipponen2,Matti Eskelinen' Veli-Matti Kosma2,Sinikka Marin2, Esko Alhava' and Kari Syrjanen2
Downloaded by [Universite Laval] at 04:21 02 April 2016
I
The oestrogen (ER) and progesteronereceptor (PR) status in a series of 281 women with breast cancer (followed up for a mean of 8.5 years) was correlated with eight histological variables, seven nuclear morphometric factors and two mitotic indices. Sex steroid receptor status was not significantly related to tumour size, axillary lymph node status or tumour recurrence. Sex steroid receptors were related to histological grade (PcO.OOOl), nuclear grade (PeO.OOOl), tumour necrosis (P=0.0003), tumour circumscription (P=0.0027), inflammatory cell reaction (P=0.007), intraductal growth pattern (P=0.0378), and tubule formation (P=0.0432). Receptor status was also significantly related to nuclear morphometric variables (PcO.001) and mitotic indexes (PcO.001). In univariate analysis, ER (P=0.2) and PR (P=O.l) negatively predicted the recurrence free survival. Cancer related survival was predicted by ER (P=0.0068) and PR (P=0.0027). In small (diameter 12 mm) axillary lymph node negative tumours, ER (P=O.O9) and PR (P=O.l) had some value in predicting the survival. The survival advantage of steroid receptor positive tumours was not due to adjuvant hormone therapy. Sex steroid receptors had independent predictive value in multivariate survival analysis and also in small (diameter 220 mm) tumours. The results indicate that sex steroid receptor negativity is related to several malignant histological features in breast cancer and hormone receptors have prognostic value. Their prognostic influence seems to be mediatedthrough the different proliferation rates in receptor positive and negative breast carcinomas. Key words: sex steroid receptor; histological grade; nuclear morphometry; mitotic index; survival; breast cancer. (Annals of Medicine 23: 643-648, 1991)
Introduction Oestrogen (ER) and progesterone (PR) receptors are established prognosticfactors in female breast cancer (1-5), although some sceptical reports on their predictive value have been presented (6).At present, thesexsteroid receptors can be measured biochemically (7) or by using immunohistochemical methods (8).The sex steroid receptor status has been related to tumour histology (8-1 I ) , ERand PR-negativetumours exhibiting more pronounced nuclearatypia (&10,12,13). Similarly,these receptorsseem to be related to the intensity of the inflammatory cell reaction in the tumour tissue (12, 13). From the 'Department of Surgery and2Departmentof Pathology, Kuopio University Hospital, Kuopio, Finland. This study was supported by research grants from the North Savo Cancer Fund and Paavo Koistinen Foundation. Address and reprint requests: Sirpa Aaltomaa, MD., Department of Surgery, Kuopio University Hospital, SF-70210 Kuopio, Finland. Received: May 24, 1991; revision accepted: October 3, 1991.
Studies using histoquantitativemorphometry have shown that the nuclear size seems to be larger in sex steroid receptor negative breast carcinomas than in their receptor positive counterparts (9). Flow cytometric (FCM) studies suggest that the proliferation rate of hormone receptor negative tumours might be higher than that of receptor positive tumours (14). As yet, however, the relations between tumour ploidy and steroid receptors are incompletely elucidated (14, 15). So far as the authors are aware the associationsbetween the mitotic frequency and sex steroid receptors have not been analysed in detail in any previousstudies. The present study was conducted to eludicate the relations between biochemically determined sex steroid receptor status, and histological type, histological grade, nuclear grade, tumour necrosis, tumour circumscription, tubule formation, intraductalgrowth pattern, inflammatorycell reaction,mitotic activity, and seven morphometrically measured nuclear factors in 281 women with breast cancer. In addition, the predictive value of hormone receptors was evaluated in a survival analysis. Ann Med23
644
Aaltomaa
Lipponen Eskelinen Kosma Marin Alhava Syrjanen
Downloaded by [Universite Laval] at 04:21 02 April 2016
Patients and Methods The series comprised281 consecutivewomen with primary breast cancer treated and followed up at Kuopio University Hospital during 1978-90. Complete follow up histories of the women and peroperative biopsy specimens from their primary tumours were available for the present analysis. Axillary lymph node status was assessed histologically in 269/281 (96 %) of patients and a clinical judgement was available in 12 (4 O h ) of them. Tumour size was recordedas the maximum tumour diameter as measured in a fresh mastectomy specimen. The measurement was based on the estimate of the operating surgeon. Metastases were detected by routine chest and bone x-rays, and laboratory tests reflecting bone and liver metabolism. Metastases were verified histologicallywhen appropriate. The follow up was conducted at three month intervalsduring the first year, at six month intervalsduringthe next two years and annually thereafter. Other pertinent clinical data of patients are sumrnarised in Table 1. Cytosol oestrogen and progesteronereceptor measurements were performed using charchoal dextran assay described in detail elsewhere (7). The cut off level (5 fmol/mg cytosol protein) for oestrogen (ER) and progesterone receptor (PR) positivity was adopted from the literature (2). Peroperative biopsy specimens were fixed in buffered formalin (pH 7.0) and embedded in paraffin. Five pm thick sections were cut from these archivalspecimens and stained with hematoxylin and eosin. The histological grading (16) and typing (17) of the tumours was completed by two board certified pathologists (V-MK, SM) using a consultation microscope in a blinded manner, i.e., the examiners were unaware of the clinical data. No grading was done in cases where intraductalgrowthonly was present (18). The special histological features recorded and their semiquantitative grading are shown in Table 2. The mitotic fiqures were counted by two examiners (PL, SA) using a consultation microscope and an objective magnification of x40 (field diameter 490 pm). Mitoticfigures were identified, as detailed earlier (17), from the most cellular areas of the tumour margins, avoiding the frequently necrotic central areas. The mitotic activity index was the number of mitotic figures in 10 consecutive fields. The volume corrected mitotic index (MN index) was estimated using the method described by Haapasalo et al. (19). In the M/V index method the number of mitotic figures is identified in each microscopic field in a similar manner to the measurement of MAI. Simultaneously with the mitosis count the fractionof neoplastic tissue is identifiedin each filled and the final M N index thus describes the mitotic activity/mm2 neoplastic tissue. Finallyboth mitotic indexes were adjusted to correspond to the mitotic activity per 1 mm2of the section. In morphometric measurements (SA), the IBAS 1 & 2 image analyser was used. The images of most atypical microscopic fields were focused on a video screen through a video camera attached to the microscope (magnification x40). A mean of 75 nuclei were traced using a digitisertablet and a mouse connected to the computer. The computer automatically calculated mean nuclear area (NA), standard deviation of nuclear area (SDNA), nuclear perimeter (PE), standard deviation of nuclear perimeter (SDPE), largest nuclear diameter (Dmax), shortest nuclear diameter (Dmin) and the mean area of the 10 largest nuclei (NAlO). The basic statistical analyses were completed using the SPSS/PC+ program package (20) in a Toshiba T3200
Ann Med 23
Table 1. Clinical characteristicsof the patients.
Number of patients Age at diagnosis, mean (SD) Follow-uptime, mean (SD) Diameter of tumour, mean (SD) (SE) 0-2.0 cm, number 2.1-5.0 cm, number >5.0cm, number Type of primaly treatment modified mastectomy postoperativeradiotherapy yesho hormone therapy yeslno chemotherapy yeslno Hormone receptor status ER negativelpositive PR negativelpositive Number a cases with recurrences Causes of death during the follow-up breast cancer other
281 61.4 (13.6) years 8.6 (2.0) years 3.4 (2.1) (1.2) cm 105 138 38 277 1351146 581223 441237 1171164 1281153 108 109 32
computer. The assessment of the deaths due to breast cancer was based on both death certificates and clinical data. Breast cancer survival was calculated by counting as breast cancer deaths only those patients who died with known metastasesand breast cancer disease. Recurrence free survival was defined as the time elapsed between the first surgical treatment and the first confirmed metastasisor recurrentgrowth. The univariatesurvival analysiswas based on the life table method with the Lee-Desu statistics (21). Multivariate survival analysis (22) was done (PL) with the BMDP (2L) program package.
ResuIts A total of 164/281(58 %) tumours were ER positive and 153/ 281 (54 Yo)were positivefor PR. The mean (SE) concentration of PR was 73.1 (9.5) fmolhg cytosol protein. The receptor positivity was not related to tumour size (P=0.7), axillary lymph node stage (P=0.4), metastasis at diagnosis (P=O.9), recurrence (P=0.3) or age (pre-/ postmenopausal) (P=O.l) during the follow up. The prediction of recurrence in postmenopausal patients was better than in premenopausal patients; the difference was not, however, significant. Hormone receptor positivity was related significantly to several histological features (Table 2), whereas receptor content was independent of the histological types (P=0.3). Both ER and PR positive tumours had significantly lower mitotic frequency (MA1 & M N index) than in their receptor negative counterparts (Table 3). The morphometric variables were higher in hormone receptor negative tumours than in receptor positive tumours (t-test, Pc0.001) exept Dmin which did not differ significantly between the groups ( P>O.05). The relationship between hormone receptors, SD of nuclear area and MA1 (X2-statistics)are shown in Table 4. Significantly lower values of both MA1 and SDNA were associated with ER and PR positive tumours. ER predicted the recurrencefree survival (X2=1.5,P=0.2) and disease related survival (X2=7.3, P=0.0068, Fig. 1). Progesterone receptors were related to recurrence free survival (X2=2.4,P=O.l )anddisease related survival (X2=9.0,
Hormone Receptors as Prognostic Factors in Female Breast Cancer
645
Table 2. Histologicalfeatures related to hormone receptors. Feature neg.
pos.
P*
neg.
pos.
P'
6 45 59
12 105 46
0.0001
7 50 66
11 100 39
co.ooo1
0 5 108
4 18 137
0.0272
0 7 119
4 16 126
0.0432
Tumour margin circumscription 1 definite 2 questionable 3 none
23 86 4
16 138 5
0.0543
27 93 6
12 131 3
0.0027
lntraductal growth 0 none 1 some 2 principal
69 30 14
106 48 5
0.0128
84 29 13
91 49 6
0.0378
Tumour necrosis 0 none 1 spotty 2 moderate 3 severe
81 20 7 5
138 18 3 0
88 26 7 5
131 12 3 0
Nuclear pleomorphism 1 slight 2 moderate 3 severe
7 41 65
6 106 47
ISSN: 0785-3890 (Print) 1365-2060 (Online) Journal homepage: http://www.tandfonline.com/loi/iann20
Hormone Receptors as Prognostic Factors in Female: Breast Cancer Sirpa Aaltomaa, Pertti Lipponen, Matti Eskelinen, Veli-Matti Kosma, Sinikka Marin, Esko Alhava & Kari Syrjänen To cite this article: Sirpa Aaltomaa, Pertti Lipponen, Matti Eskelinen, Veli-Matti Kosma, Sinikka Marin, Esko Alhava & Kari Syrjänen (1991) Hormone Receptors as Prognostic Factors in Female: Breast Cancer, Annals of Medicine, 23:6, 643-648, DOI: 10.3109/07853899109148097 To link to this article: http://dx.doi.org/10.3109/07853899109148097
Published online: 08 Jul 2009.
Submit your article to this journal
Article views: 14
View related articles
Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=iann20 Download by: [Universite Laval]
Date: 02 April 2016, At: 04:21
Original Article
Hormone Receptors as Prognostic Factors in Female Breast Cancer Sirpa Aaltomaa' , Pertti Lipponen2,Matti Eskelinen' Veli-Matti Kosma2,Sinikka Marin2, Esko Alhava' and Kari Syrjanen2
Downloaded by [Universite Laval] at 04:21 02 April 2016
I
The oestrogen (ER) and progesteronereceptor (PR) status in a series of 281 women with breast cancer (followed up for a mean of 8.5 years) was correlated with eight histological variables, seven nuclear morphometric factors and two mitotic indices. Sex steroid receptor status was not significantly related to tumour size, axillary lymph node status or tumour recurrence. Sex steroid receptors were related to histological grade (PcO.OOOl), nuclear grade (PeO.OOOl), tumour necrosis (P=0.0003), tumour circumscription (P=0.0027), inflammatory cell reaction (P=0.007), intraductal growth pattern (P=0.0378), and tubule formation (P=0.0432). Receptor status was also significantly related to nuclear morphometric variables (PcO.001) and mitotic indexes (PcO.001). In univariate analysis, ER (P=0.2) and PR (P=O.l) negatively predicted the recurrence free survival. Cancer related survival was predicted by ER (P=0.0068) and PR (P=0.0027). In small (diameter 12 mm) axillary lymph node negative tumours, ER (P=O.O9) and PR (P=O.l) had some value in predicting the survival. The survival advantage of steroid receptor positive tumours was not due to adjuvant hormone therapy. Sex steroid receptors had independent predictive value in multivariate survival analysis and also in small (diameter 220 mm) tumours. The results indicate that sex steroid receptor negativity is related to several malignant histological features in breast cancer and hormone receptors have prognostic value. Their prognostic influence seems to be mediatedthrough the different proliferation rates in receptor positive and negative breast carcinomas. Key words: sex steroid receptor; histological grade; nuclear morphometry; mitotic index; survival; breast cancer. (Annals of Medicine 23: 643-648, 1991)
Introduction Oestrogen (ER) and progesterone (PR) receptors are established prognosticfactors in female breast cancer (1-5), although some sceptical reports on their predictive value have been presented (6).At present, thesexsteroid receptors can be measured biochemically (7) or by using immunohistochemical methods (8).The sex steroid receptor status has been related to tumour histology (8-1 I ) , ERand PR-negativetumours exhibiting more pronounced nuclearatypia (&10,12,13). Similarly,these receptorsseem to be related to the intensity of the inflammatory cell reaction in the tumour tissue (12, 13). From the 'Department of Surgery and2Departmentof Pathology, Kuopio University Hospital, Kuopio, Finland. This study was supported by research grants from the North Savo Cancer Fund and Paavo Koistinen Foundation. Address and reprint requests: Sirpa Aaltomaa, MD., Department of Surgery, Kuopio University Hospital, SF-70210 Kuopio, Finland. Received: May 24, 1991; revision accepted: October 3, 1991.
Studies using histoquantitativemorphometry have shown that the nuclear size seems to be larger in sex steroid receptor negative breast carcinomas than in their receptor positive counterparts (9). Flow cytometric (FCM) studies suggest that the proliferation rate of hormone receptor negative tumours might be higher than that of receptor positive tumours (14). As yet, however, the relations between tumour ploidy and steroid receptors are incompletely elucidated (14, 15). So far as the authors are aware the associationsbetween the mitotic frequency and sex steroid receptors have not been analysed in detail in any previousstudies. The present study was conducted to eludicate the relations between biochemically determined sex steroid receptor status, and histological type, histological grade, nuclear grade, tumour necrosis, tumour circumscription, tubule formation, intraductalgrowth pattern, inflammatorycell reaction,mitotic activity, and seven morphometrically measured nuclear factors in 281 women with breast cancer. In addition, the predictive value of hormone receptors was evaluated in a survival analysis. Ann Med23
644
Aaltomaa
Lipponen Eskelinen Kosma Marin Alhava Syrjanen
Downloaded by [Universite Laval] at 04:21 02 April 2016
Patients and Methods The series comprised281 consecutivewomen with primary breast cancer treated and followed up at Kuopio University Hospital during 1978-90. Complete follow up histories of the women and peroperative biopsy specimens from their primary tumours were available for the present analysis. Axillary lymph node status was assessed histologically in 269/281 (96 %) of patients and a clinical judgement was available in 12 (4 O h ) of them. Tumour size was recordedas the maximum tumour diameter as measured in a fresh mastectomy specimen. The measurement was based on the estimate of the operating surgeon. Metastases were detected by routine chest and bone x-rays, and laboratory tests reflecting bone and liver metabolism. Metastases were verified histologicallywhen appropriate. The follow up was conducted at three month intervalsduring the first year, at six month intervalsduringthe next two years and annually thereafter. Other pertinent clinical data of patients are sumrnarised in Table 1. Cytosol oestrogen and progesteronereceptor measurements were performed using charchoal dextran assay described in detail elsewhere (7). The cut off level (5 fmol/mg cytosol protein) for oestrogen (ER) and progesterone receptor (PR) positivity was adopted from the literature (2). Peroperative biopsy specimens were fixed in buffered formalin (pH 7.0) and embedded in paraffin. Five pm thick sections were cut from these archivalspecimens and stained with hematoxylin and eosin. The histological grading (16) and typing (17) of the tumours was completed by two board certified pathologists (V-MK, SM) using a consultation microscope in a blinded manner, i.e., the examiners were unaware of the clinical data. No grading was done in cases where intraductalgrowthonly was present (18). The special histological features recorded and their semiquantitative grading are shown in Table 2. The mitotic fiqures were counted by two examiners (PL, SA) using a consultation microscope and an objective magnification of x40 (field diameter 490 pm). Mitoticfigures were identified, as detailed earlier (17), from the most cellular areas of the tumour margins, avoiding the frequently necrotic central areas. The mitotic activity index was the number of mitotic figures in 10 consecutive fields. The volume corrected mitotic index (MN index) was estimated using the method described by Haapasalo et al. (19). In the M/V index method the number of mitotic figures is identified in each microscopic field in a similar manner to the measurement of MAI. Simultaneously with the mitosis count the fractionof neoplastic tissue is identifiedin each filled and the final M N index thus describes the mitotic activity/mm2 neoplastic tissue. Finallyboth mitotic indexes were adjusted to correspond to the mitotic activity per 1 mm2of the section. In morphometric measurements (SA), the IBAS 1 & 2 image analyser was used. The images of most atypical microscopic fields were focused on a video screen through a video camera attached to the microscope (magnification x40). A mean of 75 nuclei were traced using a digitisertablet and a mouse connected to the computer. The computer automatically calculated mean nuclear area (NA), standard deviation of nuclear area (SDNA), nuclear perimeter (PE), standard deviation of nuclear perimeter (SDPE), largest nuclear diameter (Dmax), shortest nuclear diameter (Dmin) and the mean area of the 10 largest nuclei (NAlO). The basic statistical analyses were completed using the SPSS/PC+ program package (20) in a Toshiba T3200
Ann Med 23
Table 1. Clinical characteristicsof the patients.
Number of patients Age at diagnosis, mean (SD) Follow-uptime, mean (SD) Diameter of tumour, mean (SD) (SE) 0-2.0 cm, number 2.1-5.0 cm, number >5.0cm, number Type of primaly treatment modified mastectomy postoperativeradiotherapy yesho hormone therapy yeslno chemotherapy yeslno Hormone receptor status ER negativelpositive PR negativelpositive Number a cases with recurrences Causes of death during the follow-up breast cancer other
281 61.4 (13.6) years 8.6 (2.0) years 3.4 (2.1) (1.2) cm 105 138 38 277 1351146 581223 441237 1171164 1281153 108 109 32
computer. The assessment of the deaths due to breast cancer was based on both death certificates and clinical data. Breast cancer survival was calculated by counting as breast cancer deaths only those patients who died with known metastasesand breast cancer disease. Recurrence free survival was defined as the time elapsed between the first surgical treatment and the first confirmed metastasisor recurrentgrowth. The univariatesurvival analysiswas based on the life table method with the Lee-Desu statistics (21). Multivariate survival analysis (22) was done (PL) with the BMDP (2L) program package.
ResuIts A total of 164/281(58 %) tumours were ER positive and 153/ 281 (54 Yo)were positivefor PR. The mean (SE) concentration of PR was 73.1 (9.5) fmolhg cytosol protein. The receptor positivity was not related to tumour size (P=0.7), axillary lymph node stage (P=0.4), metastasis at diagnosis (P=O.9), recurrence (P=0.3) or age (pre-/ postmenopausal) (P=O.l) during the follow up. The prediction of recurrence in postmenopausal patients was better than in premenopausal patients; the difference was not, however, significant. Hormone receptor positivity was related significantly to several histological features (Table 2), whereas receptor content was independent of the histological types (P=0.3). Both ER and PR positive tumours had significantly lower mitotic frequency (MA1 & M N index) than in their receptor negative counterparts (Table 3). The morphometric variables were higher in hormone receptor negative tumours than in receptor positive tumours (t-test, Pc0.001) exept Dmin which did not differ significantly between the groups ( P>O.05). The relationship between hormone receptors, SD of nuclear area and MA1 (X2-statistics)are shown in Table 4. Significantly lower values of both MA1 and SDNA were associated with ER and PR positive tumours. ER predicted the recurrencefree survival (X2=1.5,P=0.2) and disease related survival (X2=7.3, P=0.0068, Fig. 1). Progesterone receptors were related to recurrence free survival (X2=2.4,P=O.l )anddisease related survival (X2=9.0,
Hormone Receptors as Prognostic Factors in Female Breast Cancer
645
Table 2. Histologicalfeatures related to hormone receptors. Feature neg.
pos.
P*
neg.
pos.
P'
6 45 59
12 105 46
0.0001
7 50 66
11 100 39
co.ooo1
0 5 108
4 18 137
0.0272
0 7 119
4 16 126
0.0432
Tumour margin circumscription 1 definite 2 questionable 3 none
23 86 4
16 138 5
0.0543
27 93 6
12 131 3
0.0027
lntraductal growth 0 none 1 some 2 principal
69 30 14
106 48 5
0.0128
84 29 13
91 49 6
0.0378
Tumour necrosis 0 none 1 spotty 2 moderate 3 severe
81 20 7 5
138 18 3 0
88 26 7 5
131 12 3 0
Nuclear pleomorphism 1 slight 2 moderate 3 severe
7 41 65
6 106 47
