Clinicopathological factors associated with survival in patients with breast cancer brain metastasis

    Clinicopathologic Factors Associated with Survival in Patients with Breast Cancer Brain Metastasis Rong Li MD, PhD, Kui Zhang PhD, Gene P. Siegal MD, PhD, Shi Wei MD, PhD PII: DOI: Reference:

S0046-8177(17)30109-0 doi: 10.1016/j.humpath.2017.03.022 YHUPA 4178

To appear in:

Human Pathology

Received date: Revised date: Accepted date:

31 January 2017 17 March 2017 30 March 2017

Please cite this article as: Li Rong, Zhang Kui, Siegal Gene P., Wei Shi, Clinicopathologic Factors Associated with Survival in Patients with Breast Cancer Brain Metastasis, Human Pathology (2017), doi: 10.1016/j.humpath.2017.03.022

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Clinicopathologic Factors Associated with Survival in Patients with Breast Cancer Brain

T

Metastasis

Department of Pathology, the University of Alabama at Birmingham, Birmingham,

SC

1

RI P

Rong Li, MD, PhD,1 Kui Zhang, PhD,2 Gene P. Siegal, MD, PhD,1 Shi Wei, MD, PhD1

NU

Alabama 35249, and 2Department of Mathematical Sciences, Michigan Technological

CE

PT

ED

MA

University, Houghton, Michigan 49931

Disclosure: The authors declare no conflicts of interest.

AC

Running title: Prognostic factors in breast cancer brain metastasis Corresponding author: Shi Wei, M.D., Ph.D. Department of Pathology University of Alabama at Birmingham NP 3545A, 619 19th St. South Birmingham, AL 35249-7331 Phone: (205) 975-8880 Fax: (205) 975-5242 E-mail: [email protected]

1

ACCEPTED MANUSCRIPT ABSTRACT Brain metastasis from breast cancer generally represents a catastrophic event yet demonstrates

T

substantial biological heterogeneity. There have been limited studies solely focusing on the

RI P

prognosis of patients with such metastasis. In this study, we carried out a comprehensive analysis in 108 consecutive patients with breast cancer brain metastases between 1997 and 2012 to further

SC

define clinicopathologic factors associated with early onset of brain metastasis and survival

NU

outcomes after development of them. We found that lobular carcinoma, higher clinical stages at diagnosis, and lack of coexisting bone metastasis were significantly associated with a worse

MA

brain relapse-free survival when compared to brain only metastasis. High histologic grade, triplenegative breast cancer, and absence of visceral involvement were unfavorable prognostic factors

ED

after brain metastasis. Further, high histologic grade, advanced tumor stages and lack of

PT

coexisting bone involvement indicated a worse overall survival. Thus, the previously established prognostic factors in early-stage or advanced breast cancers may not entirely apply to patients

CE

with brain metastases. Further, the prognostic significance of the clinicopathologic factors

AC

differed before and after a patient develops brain metastasis. This knowledge might help in establishing an algorithm to further stratify breast cancer patients into prognostically significant categories for optimal prevention, screening and treatment of their brain metastasis.

Keywords: breast cancer; brain metastasis; prognosis; molecular subtypes

2

ACCEPTED MANUSCRIPT INTRODUCTION Breast cancer remains the most common malignancy and the second leading cause of

T

cancer mortality among American women, with an estimated 246,660 new cases of invasive

RI P

disease and 40,450 cancer deaths in 2016 [1]. Despite the recent advances in early detection and wide application of systemic therapies, approximately 5 to 10% of patients are still diagnosed

SC

with metastatic diseases at initial presentation, and there is a 20 to 30% likelihood of eventually

NU

developing distant metastasis when diagnosed with early-stage breast cancer [2, 3]. On the other hand, it has long been known that cancer metastasis is a non-random

MA

process, thus the distant sites to which breast cancer preferentially relapse, of which bone, liver, lung and brain are among the most common organs, are of clinical significance, and are closely

ED

related to the patients’ prognostic outcome [3-5]. Further, the incidence of brain metastasis has

PT

been continuously rising due to the technological advances in earlier detection, and also as a result of longer survival in the setting of controlled extracranial disease secondary to novel

CE

chemotherapies or human epidermal growth factor receptor 2 (HER2)-targeted therapies, as most

AC

of these therapeutic agents do not cross the blood-brain barrier [5, 6]. Up to 17% of patients with advanced breast cancer develop brain metastases synchronously or metachronously [7]. Moreover, breast cancer is the second most common source of brain metastases derived from solid malignancies [8]. A recent comprehensive review of 106 studies over the past 35 years revealed an incidence of 24% for brain metastasis from breast cancer [9]. Comparative analyses to identify risk factors for brain metastasis and to compare prognostic outcomes in patients with and without cranial involvement have been extensively studied. Brain metastases in breast cancer generally represent a catastrophic event that presages dismal survival [10-14]. However, this malignancy represents a group of disease with substantial

3

ACCEPTED MANUSCRIPT biological and clinical heterogeneity. This is more evident with additional novel therapeutic agents available over the last decade. Reflecting this actuality, the limited recent studies on

T

clinicopathologic prognostic factors in patients with breast cancer brain metastasis have resulted

RI P

in variable observations [12, 15-17]. Thus, it is essential to conduct additional systemic studies to contemplate the diversity of the disease in the pursuit of precision medicine. The aim of this

SC

study was to carry out a comprehensive analysis to further define clinicopathologic factors

NU

associated with early onset of brain metastasis and survival outcomes after development of brain metastasis. This might help stratify patients into prognostically significant categories for

MA

personalized surveillance, thus tailor individualized clinical management.

ED

MATERIALS AND METHODS

PT

After approval by the institutional review board of the authors’ institution, the tumor registry was searched to identify breast cancer cases with associated brain metastasis between

CE

1997 and 2012. A total of 108 patients with clinicopathologic, treatment and follow-up data

AC

available were identified. The patients’ demographic information and the pathologic features of the tumor were recorded, including the age at diagnosis, race, histologic type and grade, and the statuses of estrogen/progesterone receptors (ER/ PR) and HER2. The accuracy of the data was further validated for each patient using the electronic medical record and/or surgical pathology database. The diagnosis of metastatic breast cancer in distant organs was based on imaging studies (conventional radiography, computed tomography, or magnetic resonance imaging) with or without confirmation by tissue biopsy. With regard to brain metastases, 53 patients underwent brain biopsy or cerebral fluid fine needle aspiration at the authors’ institution. The clinical stage

4

ACCEPTED MANUSCRIPT was based on the American Joint Committee on Cancer (AJCC) Cancer Staging Manual 7th edition [18].

T

The ER, PR and HER2 statuses were assessed by immunohistochemistry or in situ

RI P

hybridization analysis as previously described [19]. A positive ER or PR was defined as ≥1% of tumor cell nuclei with immunoreactivity. HER2 overexpression/amplification was determined as

SC

either a 3+ immunohistochemistry score (uniform and intense membrane staining of >30% of

NU

tumor cells) or a positive in situ hybridization result. Of note, receptor studies were performed in 25 brain biopsy specimens. The results of ER, PR and HER2 statuses were used as surrogate

MA

markers for classification of the major breast cancer subtypes as previously reported, including luminal (ER+ and/or PR+), HER2 (ER−/PR−/HER2+) and triple negative breast cancer

ED

(ER−/PR−/HER2−; TNBC) [5, 19]. The luminal carcinomas and TNBC were not further

PT

subclassified given the lack of reliable markers and clear clinical indication. All patients received local and/or systemic therapies. All but 7 patients received systemic

CE

therapy at the authors’ institution. Thirty-eight out of 51 patients with ER/PR-positive tumors

AC

received endocrine therapy, based on physician and patient discussions. Nighty-three patients received cytotoxic chemotherapy. Eighty-seven patients had gamma knife/radiotherapy. Brain relapse-free survival (calculated from the date of diagnosis to the date of brain metastasis), survival after brain metastasis (calculated from the date of brain metastasis to the date of death or the follow-up cutoff) and overall survival (calculated from the date of diagnosis to the date of death or the follow-up cutoff) were plotted on Kaplan-Meier curves. Of the 83 patients who died in the study period, 75 died of disease while the causes of death for remaining 8 patients were unknown. Patients who survived at the study cutoff (N=19, 17.6%) or were lost to follow-up (N=6, 5.6%) were considered censored data in the analysis. The log-rank test was

5

ACCEPTED MANUSCRIPT used to compare groups, followed by performance of a Cox proportional hazards regression analysis. A P-value of <.05 was considered statistically significant. The statistical analysis was

RI P

T

conducted using R.3.0.1 (www.r-project.org) software.

RESULTS

SC

Clinicopathological characteristics

NU

The demographic data from the 108 patients included in the study and the pathologic characteristics of the tumors are summarized in Table 1. In brief, most patients (66%) were

MA

between 40-59 years of age (median 48 years) and were Caucasians (71%). These cases displayed a trend toward a high histologic grade. Near half of the patients harbored luminal

ED

carcinomas while the HER2 subtype and TNBC each consisted of approximately one quarter of

PT

all cases, respectively. These latter two subtypes constituted a much higher proportion of breast cancers than encountered in general clinical practice, thus reflecting the unfavorable clinical

CE

outcomes for these patients. The majority of these patients developed brain only metastases

AC

(57%). This was followed by both brain and visceral (i.e., liver, lung) metastases (20%), whereas the combination of brain and bone metastases with or without visceral involvement was far less common (each 11%). The median brain relapse-free interval, survival after brain metastasis, and overall survival were 761, 340, and 1165 days, respectively. The metastatic tumors in the brain biopsies showed near exclusively poorly differentiated carcinomas, despite the fact that some of the primary tumors were of Grade II (N=11), with only one exception in which primary and metastatic carcinomas were both Grade II. Of the 25 brain biopsies in which receptor studies were performed, there were four cases with discrepancies for ER status between primary and metastatic tumors (2 positive to negative, 2 negative to positive),

6

ACCEPTED MANUSCRIPT 6 with discrepancies for PR status (1 positive to negative, 5 negative to positive), and 1 with a discrepancy for HER2 status (equivocal to positive). Of note, the metastatic tumors with negative

T

to positive hormonal receptor alterations typically showed weak and/or focal ER/PR expression

RI P

(i.e., 1+; <10%) with one exception in which ER was strongly and diffusely positive (Figure 1). Factors associated with brain relapse-free survival

SC

Lobular carcinoma was independently associated with a worse brain relapse-free

NU

survival, thus reflecting the highly infiltrative nature of this histologic type (Table 2, Figure 2). Not surprisingly, the clinical stage at diagnosis significantly correlated with a worse outcome.

MA

The combination of brain and visceral metastases (lack of bone involvement) predicated a shorter brain relapse-free interval when compared to brain only metastasis, but failed to reach a

ED

statistical significance by multivariate analysis. Of note, ten patients presented with brain

PT

metastases at the time of diagnosis of their breast cancer, and thus were excluded from this analysis.

CE

Factors associated with survival after brain metastasis

AC

In contrast to brain relapse-free survival for which the lack of coexisting skeletal involvement was a poorer prognostic indicator, the presence of bone metastasis (but lack of visceral involvement) was an independent factor associated with a worse outcome after development of brain metastasis when compared to brain only metastasis. Similarly, high histologic grade and the TNBC subtype were both associated with a shorter survival after brain metastasis although they did not reach a statistical significance by multivariate analysis (Table 3, Figure 3). Moreover, the HER2 subtype was associated with a prolonged survival after brain metastasis when compared to TNBC by univariate analysis (P=0.009) but only had marginal significance by multivariate analysis (P=0.059).

7

ACCEPTED MANUSCRIPT Factors associated with overall survival In analysis of overall survival, high histologic grade and advanced tumor stage were both

T

significant prognostic factors by univariate and multivariate analyses. The absence of coexisting

RI P

bone involvement indicated a worse outcome, yet it failed to be significant when applying a

SC

multivariate analysis (Table 4, Figure 4).

NU

DISCUSSION

The propensity for metastasis to varying distant sites is a major challenge in the

MA

management of breast cancer. Recent discoveries of novel therapeutic agents have significantly improved the prognostic outcomes in patients with advanced breast cancer. Yet, secondary

ED

cancer to the brain has been increasing over time. The brain represents the fourth most common

PT

organ for ‘recurrent’ breast cancer as it may act as a sanctuary site for metastatic disease [20]. Thus, development of adequate preventive strategies is desperately required in addition to the

CE

need for local cancer control in the brain.

AC

There have been limited studies solely focusing on the prognosis of breast cancer brain metastasis in the last decade. Moreover, most studies utilized slightly different subsets of variables in their analyses while some lacked one or more essential clinical and pathological parameters, such as the patient’s gender and ethnicity, or histologic type and grade of the tumor. To that end, we have performed a systemic analysis to identify clinicopathologic predictors of outcomes in patients with breast cancer brain metastasis. The study has resulted in a number of interesting observations. Among the pathologic characteristics of breast cancer, the histologic grade is arguably the most important prognostic factor in early-stage disease [21]. The significance of grading has

8

ACCEPTED MANUSCRIPT been further stressed by the most recent edition of AJCC Cancer Staging Manual, in which the histologic grade has emerged as a component of Prognostic Stage Groups in breast cancer along

T

with pathologic stage and ER, PR and HER2 statuses [22]. In the current study, tumor grade was

RI P

not significantly associated with brain relapse-free survival, in keeping with two recent studies,[10, 16] and likely reflecting the fact that high grade tumors constitute a much higher

SC

proportion of breast cancers that subsequently develop brain metastases. Interestingly, the

NU

histologic grade was a significant predictor for survival after brain metastasis as well as an independent factor for overall survival despite the fact of most metastatic tumors in the brain

MA

were poorly differentiated in this cohort, albeit in a limited number of cases. This observation is also in harmony with the findings in a large cohort of patients with stage IV breast cancer

ED

regardless of brain relapse [23]. Another novel observation was that the lobular histologic type

PT

was found to be a significantly worse factor for brain relapse-free survival, a finding not observed in the previous studies with brain metastasis but in line with a most recent cohort of

CE

metastatic breast cancer, in which the clinical stage was significantly higher in lobular tumors

AC

than that in ductal carcinomas [24]. Again, this reflects the highly infiltrative nature of this histologic variant due to loss of adhesion molecules thus resulting in epithelial-mesenchymal transition.

First established in 2000, the breast cancer molecular subtypes have demonstrated prognostic significance, and thereby largely influenced patient management over the past decade [25]. Given the difficulties of gene profiling in clinical practice, the expression profile of ER/PR/HER2 has been extensively used as a surrogate marker, although classification of breast cancer into subtypes based on protein expression is somewhat subjective and controversial, and the optimal cutoffs are still being debated and have changed over time [5, 26]. There have also

9

ACCEPTED MANUSCRIPT been controversies in regard to the prognostic significance of breast cancer subtypes in patients with brain metastasis in the very few previously published studies. Brain relapse-free survival

T

was found to differ significantly between breast cancer subtypes in two studies [16, 17], of which

RI P

the first utilized different cutoffs for ER/PR (10%) in contrast to the current practice in the US (1%). However, the same observation was not found in another cohort and, interestingly, both

SC

the latter study and ours demonstrated prognostic significance of subtypes in survival after brain

NU

metastasis [15]. This finding is also largely in accordance with the aforementioned study of advanced breast cancers regardless of brain involvement [23]. Moreover, the discrepancies

MA

regarding receptor status between primary and metastatic tumors were of further interest. While the changes from a positive to negative status may have resulted from targeted therapy or

ED

heterogeneity of the tumor, those from a negative to positive status, likely representing a

PT

sampling issue in the initial breast biopsy, resulting in an alteration of non-luminal primary to luminal metastatic tumor in a subset of cases (4/25, 16% in this cohort). The finding also

CE

emphasized the importance of performing receptor studies in the metastatic deposits.

AC

It is not surprising that the patients who presented with higher clinical stages (III and IV) had shorter time to brain relapse as these tumors tend to respond poorly to systemic therapies, as also demonstrated by a previous study [13]. It is of substantial interest that the presence of extracranial metastasis demonstrated a potential prognostic value. When compared to brain only metastasis, the lack of coexisting bone lesions was associated with a shorter brain relapse-free interval, an observation in keeping with an early study [10]. A potential explanation for this was that the subset of breast carcinomas with bone relapse was significantly associated with the luminal subtypes [5], and the ER/PR-positive tumors had a prolonged disease free survival [21]. Moreover, the lack of visceral involvement was a poor prognostic factor after brain metastasis.

10

ACCEPTED MANUSCRIPT The mechanisms by which the prognostic factors differ before and after brain metastases largely remain unclear thus requiring further investigation. Of note, there were only 12 patients in each

T

of the groups of brain/bone and brain/bone/viscera, thus the power of this significance

RI P

necessitates further validation in larger cohorts.

We should acknowledge that there are some limitations inherit to this retrospective study.

SC

This was a single institutional study that constituted patients seeking care at our regional medical

NU

center, thus may not entirely represent the general population of breast cancer patients with brain metastases. However, utilizing this patient group from a southern state gave us an opportunity to

MA

better characterize the racial disparities between Caucasians and African Americans. Further, the optimal cutoff values to define the ER, PR and HER2 statuses have changed over time. However,

ED

any changes in the positive rates of the receptors in advanced breast cancer were thought

PT

unlikely to be significant. Furthermore, the classification of breast cancer subtypes based on protein expression remains subjective. In this study, we did not separate the luminal A and B

CE

subtypes in ER/PR-positive tumors or basal-like and non-basal-like in triple-negative breast

AC

cancers due to the controversies in classification and, more importantly, no clear clinical indication for the routine identification of these subtypes. Lastly, the staging criteria and therapeutic modalities have changed over time during the study period. To minimize potential confounding factors, we used the AJCC Cancer Staging Manual 7th edition [18]. and all patients received standard of care treatment at the time of diagnosis. In summary, the previously established clinicopathologic prognostic factors in early-stage or advanced breast cancers may not entirely apply to patients with brain metastases. In addition, some clinical and pathologic parameters, such as histologic type, tumor grade, molecular subtypes, and the patterns of metastases, may have different prognostic values before and after a

11

ACCEPTED MANUSCRIPT patient develops brain metastasis, While these findings necessitate further large-scaled studies to confirm and to unravel the potential biologic mechanisms, this knowledge might help in

T

establishing an algorithm to further stratify breast cancer patients into prognostically significant

RI P

categories for optimal prevention, screening and treatment of brain metastasis in the pursuit of

SC

precision medicine.

MA

The authors declare no conflicts of interest.

NU

Acknowledgment:

Reference:

Siegel RL, Miller KD, Jemal A: Cancer statistics, 2016. CA Cancer J Clin 2016, 66(1):730.

2.

Rugo HS: The importance of distant metastases in hormone-sensitive breast cancer. Breast 2008, 17 Suppl 1:S3-8.

3.

Chang J, Clark GM, Allred DC, Mohsin S, Chamness G, Elledge RM: Survival of patients with metastatic breast carcinoma: importance of prognostic markers of the primary tumor. Cancer 2003, 97(3):545-553.

4.

Hess KR, Varadhachary GR, Taylor SH, Wei W, Raber MN, Lenzi R, Abbruzzese JL: Metastatic patterns in adenocarcinoma. Cancer 2006, 106(7):1624-1633.

5.

Soni A, Ren Z, Hameed O, Chanda D, Morgan CJ, Siegal GP, Wei S: Breast cancer subtypes predispose the site of distant metastases. Am J Clin Pathol 2015, 143(4):471478.

6.

Weil RJ, Palmieri DC, Bronder JL, Stark AM, Steeg PS: Breast cancer metastasis to the central nervous system. Am J Pathol 2005, 167(4):913-920.

7.

Wei S, Siegal GP: Metastatic Organotropism: An Intrinsic Property of Breast Cancer Molecular Subtypes. Adv Anat Pathol 2017.

8.

Lin NU, Bellon JR, Winer EP: CNS metastases in breast cancer. J Clin Oncol 2004, 22(17):3608-3617.

9.

Rostami R, Mittal S, Rostami P, Tavassoli F, Jabbari B: Brain metastasis in breast cancer: a comprehensive literature review. J Neurooncol 2016, 127(3):407-414.

AC

CE

PT

ED

1.

12

ACCEPTED MANUSCRIPT Lee SS, Ahn JH, Kim MK, Sym SJ, Gong G, Ahn SD, Kim SB, Kim WK: Brain metastases in breast cancer: prognostic factors and management. Breast Cancer Res Treat 2008, 111(3):523-530.

11.

Ogawa K, Yoshii Y, Nishimaki T, Tamaki N, Miyaguni T, Tsuchida Y, Kamada Y, Toita T, Kakinohana Y, Tamaki W et al: Treatment and prognosis of brain metastases from breast cancer. J Neurooncol 2008, 86(2):231-238.

12.

Sperduto PW, Kased N, Roberge D, Xu Z, Shanley R, Luo X, Sneed PK, Chao ST, Weil RJ, Suh J et al: Effect of tumor subtype on survival and the graded prognostic assessment for patients with breast cancer and brain metastases. Int J Radiat Oncol Biol Phys 2012, 82(5):2111-2117.

13.

Leone JP, Lee AV, Brufsky AM: Prognostic factors and survival of patients with brain metastasis from breast cancer who underwent craniotomy. Cancer Med 2015, 4(7):989994.

14.

Largillier R, Ferrero JM, Doyen J, Barriere J, Namer M, Mari V, Courdi A, HannounLevi JM, Ettore F, Birtwisle-Peyrottes I et al: Prognostic factors in 1,038 women with metastatic breast cancer. Ann Oncol 2008, 19(12):2012-2019.

15.

Niwinska A, Murawska M, Pogoda K: Breast cancer brain metastases: differences in survival depending on biological subtype, RPA RTOG prognostic class and systemic treatment after whole-brain radiotherapy (WBRT). Ann Oncol 2010, 21(5):942-948.

16.

Berghoff A, Bago-Horvath Z, De Vries C, Dubsky P, Pluschnig U, Rudas M, Rottenfusser A, Knauer M, Eiter H, Fitzal F et al: Brain metastases free survival differs between breast cancer subtypes. Br J Cancer 2012, 106(3):440-446.

17.

Sperduto PW, Kased N, Roberge D, Chao ST, Shanley R, Luo X, Sneed PK, Suh J, Weil RJ, Jensen AW et al: The effect of tumor subtype on the time from primary diagnosis to development of brain metastases and survival in patients with breast cancer. J Neurooncol 2013, 112(3):467-472.

18.

Edge SB, American Joint Committee on Cancer.: AJCC cancer staging manual, 7th edn. New York: Springer; 2010.

19.

Ren Z, Li Y, Hameed O, Siegal GP, Wei S: Prognostic factors in patients with metastatic breast cancer at the time of diagnosis. Pathol Res Pract 2014, 210(5):301-306.

20.

Frisk G, Svensson T, Backlund LM, Lidbrink E, Blomqvist P, Smedby KE: Incidence and time trends of brain metastases admissions among breast cancer patients in Sweden. Br J Cancer 2012, 106(11):1850-1853.

21.

Cianfrocca M, Goldstein LJ: Prognostic and predictive factors in early-stage breast cancer. Oncologist 2004, 9(6):606-616.

AC

CE

PT

ED

MA

NU

SC

RI P

T

10.

13

ACCEPTED MANUSCRIPT Amin MB, Edge S, Greene F, Byrd DR, Brookland RT, Washington MK, Gershenwald JE, Compton CC, Hess KR, Sullivan DC et al: AJCC Cancer Staging Manual 8th edn. New York: Springer; 2017.

23.

Ren Z, Li Y, Shen T, Hameed O, Siegal GP, Wei S: Prognostic factors in advanced breast cancer: Race and receptor status are significant after development of metastasis. Pathol Res Pract 2016, 212(1):24-30.

24.

Inoue M, Nakagomi H, Nakada H, Furuya K, Ikegame K, Watanabe H, Omata M, Oyama T: Specific sites of metastases in invasive lobular carcinoma: a retrospective cohort study of metastatic breast cancer. Breast Cancer 2017.

25.

Perou CM, Sorlie T, Eisen MB, van de Rijn M, Jeffrey SS, Rees CA, Pollack JR, Ross DT, Johnsen H, Akslen LA et al: Molecular portraits of human breast tumours. Nature 2000, 406(6797):747-752.

26.

Schwartz GF, Bartelink H, Burstein HJ, Cady B, Cataliotti L, Fentiman IS, Holland R, Hughes KS, Masood S, McCormick B et al: Adjuvant therapy in stage I carcinoma of the breast: the influence of multigene analyses and molecular phenotyping. Breast J 2012, 18(4):303-311.

AC

CE

PT

ED

MA

NU

SC

RI P

T

22.

14

ACCEPTED MANUSCRIPT Figure legends: Figure 1 Representative images of metastatic breast cancer to the brain. A metastatic carcinoma

T

from a mixed ductal and lobular carcinoma demonstrating both ductal and lobular features (A

RI P

and B, respectively; hematoxylin and eosin), with weak ER expression (C). A metastatic carcinoma from an ER-/PR- invasive ductal carcinoma (D, hematoxylin and eosin) showing

SC

strong ER expression (E) and focal PR expression (F). A metastatic lobular carcinoma exhibiting

NU

pleomorphic features (G, hematoxylin and eosin) and demonstrating HER2 overexpression (H) and amplification (I). A-H, original magnification x 200; I, original magnification x 400.

MA

Figure 2 Brain relapse-free survival stratified by (A) histologic type, (B) the pattern of distant metastases at the time of brain metastasis, (C) clinical stage at the time of diagnosis, and (D)

ED

breast cancer subtypes. BM, brain metastasis; BoM, bone metastasis; VM, visceral metastasis.

PT

Figure 3 Survival after brain metastasis stratified by (A) histologic grade, (B) the pattern of distant metastases, and (C) breast cancer subtypes. BM, brain metastasis; BoM, bone metastasis;

CE

VM, visceral metastasis.

AC

Figure 4 Overall survival stratified by (A) histologic grade, (B) the pattern of distant metastases, (C) clinical stage at the time of diagnosis, and (D) breast cancer subtypes. BM, brain metastasis; BoM, bone metastasis; VM, visceral metastasis.

15

ACCEPTED MANUSCRIPT Table 1 Patients’ demographic information and pathologic characteristics of the primary tumor No. (%)

RI P

T

19 (17.6) 36 (33.3) 35 (32.4) 18 (16.7)

NU

MA

95 (88) 10 (9.2) 2 (1.9) 1 (0.9)

SC

77 (71.3) 30 (27.8) 1 (0.9)

ED

23 (21.3) 64 (59.3) 21 (19.4) 51 (47.2) 27 (25) 28 (25.9) 2 (1.9)

a

AC

CE

PT

Clinicopathologic characteristics Age <40 40-49 50-59 ≥60 Race Caucasian African American Asian Tumor type Ductala Lobular Ductal and lobular Unknown Histologic grade Grade II Grade III Unknown Breast cancer subtype Luminal HER2 Triple negative Unknown Stage at diagnosis I-IIIb IV Pattern of metastasis Brain only Brain + Bone Brain + Viscera Brain + Bone + Viscera

91 (84) 17 (16)

62 (57.4) 12 (11.1) 22 (20.4) 12 (11.1)

Of which 7 patients presented with inflammatory carcinoma Of which 12 patients presented with stage I, 21 with stage II, 33 with stage III, and 25 with unknown stage but no known distant metastasis at diagnosis. b

16

ACCEPTED MANUSCRIPT Table 2 Univariate and multivariate analysis for brain relapse-free survival

African American

1.023 (0.649, 1.615)

0.921

0.885 (0.444, 1.764)

Ductal

2.572 (1.165, 5.677)

0.0194

6.911 (1.790, 26.69)

Grade II

1.174 (0.702, 1.965)

0.541

1.484 (0.696, 3.162)

0.31

Brain only

0.722 (0.356, 1.464) 0.936 (0.475, 1.841) 1.976 (1.148, 3.400)

0.366 0.847 0.0139

1.209 (0.410, 3.567) 0.650 (0.231, 1.830) 1.863 (0.763, 4.552)

0.73 0.41 0.17

Stage I

1.868 (0.867, 4.023) 2.626 (1.265, 5.455) 9.788 (3.507, 27.32)

0.110 0.0096 <0.0001

2.149 (0.692, 6.673) 3.389 (1.261, 9.106) 6.643 (1.267, 34.84)

0.19 0.016 0.025

Luminal

1.443 (0.9026, 2.608) 1.059 (0.6571, 1.712)

0.737 (0.347, 1.565) 1.149 (0.506, 2.606)

0.43 0.74

RI P

T

Multivariate Analysis HR (95% CI) P-value 0.978 (0.949, 1.008) 0.14

SC

Age Race Caucasian Histologic type Lobular Histologic grade Grade III Pattern of metastasis Brain/bone Brain/bone/viscera Brain/viscera Stage at diagnosis II III IV Subtype HER2 TNBC

Univariate Analysis HR (95% CI) P-value 0.994 (0.974, 1.014) 0.539

Reference Level N/A

NU

Covariate

0.1279 0.8117

AC

CE

PT

ED

MA

HR, hazard ratio; CI, confidence interval; TNBC, triple negative breast cancer

17

0.73 0.005

ACCEPTED MANUSCRIPT Table 3 Univariate and multivariate analysis for survival after brain metastasis

0.925 (0.560, 1.525)

0.759

0.945 (0.419, 2.130)

0.5295 (0.3028, 1.215)

0.1598

0.338 (0.039, 2.954)

Grade II

2.202 (1.144, 4.239)

0.018

1.835 (0.725, 4.641)

0.20

Brain only

2.792 (1.279, 6.12) 2.047 (0.911, 4.603) 1.728 (0.99, 3.017)

0.010 0.083 0.054

3.229 (1.015, 10.28) 1.296 (0.365, 4.610) 1.214 (0.468, 3.149)

0.047 0.69 0.69

Stage I

0.671 (0.272, 1.651) 0.820 (0.351, 1.919) 0.857 (0.269, 2.731)

0.385 0.648 0.794

1.510 (0.456, 4.998) 0.569 (0.087, 3.734) 0.538 (0.231, 1.253)

0.50 0.56 0.150

Luminal

0.7913 (0.4716, 1.316) 1.729 (1.060, 3.420)

1.336 (0.558, 3.199) 1.005 (0.973, 10.38)

0.520 0.74

Ductal

RI P

African American

T

Multivariate Analysis HR (95% CI) P-value 1.005 (0.973, 10.38) 0.74

SC

Age Race Caucasian Histologic type Lobular Histologic grade Grade III Pattern of metastasis Brain/bone Brain/bone/viscera Brain/viscera Stage at diagnosis II III IV Subtype HER2 TNBC

Univariate Analysis HR (95% CI) P-value 1.009 (0.988, 1.032) 0.392

Reference Level N/A

NU

Covariate

0.371 0.0357

AC

CE

PT

ED

MA

HR, hazard ratio; CI, confidence interval; TNBC, triple negative breast cancer

18

0.89 0.33

ACCEPTED MANUSCRIPT

Multivariate Analysis HR (95% CI) P-value 0.991 (0.961, 1.022) 0.580

African American

0.969 (0.603, 1.553)

0.893

0.602 (0.291, 1.243)

0.170

Ductal

1.078 (0.433, 2.685)

0.872

4.055 (0.979, 16.79)

0.053

Grade II

2.044 (1.098, 3.086)

0.021

2.704 (1.117, 6.547)

0.027

Brain only

1.276 (0.599, 2.848) 1.391 (0.651, 3.255) 2.355 (1.648, 6.235)

0.503 0.362 0.0008

2.633 (0.967, 7.173) 1.261 (0.432, 3.680) 2.269 (0.943, 5.460)

0.058 0.670 0.067

Stage I

1.551 (0.640, 3.759) 2.177 (0.945, 5.015) 6.166 (2.382, 15.97)

0.331 0.068 0.00018

2.121 (0.654, 6.882) 3.829 (1.307, 11.22) 6.109 (1.501, 24.86)

0.210 0.014 0.011

Luminal

1.368 (0.8145, 2.436) 1.302 (0.7615, 2.297)

0.23 0.324

0.587 (0.273, 1.262) 1.694 (0.723, 3.968)

0.170 0.220

Caucasian

Histologic type Lobular Histologic Grade Grade III Pattern of metastasis Brain/bone Brain/bone/viscera Brain/viscera Stage at diagnosis II III IV Subtype HER2 TNBC

SC

Age Race

NU

Reference Level N/A

Covariate

RI P

Univariate Analysis HR (95% CI) P-value 1.004 (0.984, 1.025) 0.677

T

Table 4 Univariate and multivariate analysis for overall survival

AC

CE

PT

ED

MA

HR, hazard ratio; CI, confidence interval; TNBC, triple negative breast cancer

19

ACCEPTED MANUSCRIPT

AC

CE

PT

ED

MA

NU

SC

RI P

T

Figure 1

20

ACCEPTED MANUSCRIPT

AC

CE

PT

ED

MA

NU

SC

RI P

T

Figure 2

21

ACCEPTED MANUSCRIPT

AC

CE

PT

ED

MA

NU

SC

RI P

T

Figure 3

22

ACCEPTED MANUSCRIPT

AC

CE

PT

ED

MA

NU

SC

RI P

T

Figure 4

23

ACCEPTED MANUSCRIPT Highlights: Analysis of 108 consecutive patients with breast cancer brain metastases.



The established prognostic factors may not apply to patients with brain relapse.



The prognostic factors may differ before and after brain metastasis.



The presence of extracranial metastasis is of prognostic value.



Incorporation of this knowledge may help in stratifying advanced breast cancers.

AC

CE

PT

ED

MA

NU

SC

RI P

T



24