Clinical and Pathologic Characteristics of Patients With BRCA-Positive and BRCA-Negative Breast Cancer

Abstract

Purpose

Mutations in the BRCA1 and BRCA2 genes confer greater risk of developing breast cancer. We determined whether tumor pathologic features and clinical features differ in patients with and without BRCA mutations.

Patients and Methods

Tumor pathologic features and clinical characteristics were examined in 491 women with breast cancer who underwent genetic testing for BRCA mutations between 1997 and 2006. A retrospective review of medical records was conducted to determine clinical characteristics including ethnicity, age and clinical stage at diagnosis, age at parity, number of full-term pregnancies, use of oral contraceptives and hormone replacement therapy, and BRCA mutation status. Tumor pathology was reviewed to determine histologic type, tumor grade, and estrogen receptor, progesterone receptor, and HER-2/neu status.

Results

Of the 491 patients with identified breast cancers, 391 patients were BRCA negative, and 86 patients were BRCA positive. Triple-negative breast cancer (ie, those with negative estrogen receptor, progesterone receptor, and HER-2/neu status) was diagnosed in 57.1% of the BRCA1-positive patients, 23.3% of the BRCA2-positive patients, and 13.8% of the BRCA-negative patients. BRCA1 mutation carriers had higher nuclear grade tumors than the other two groups (P < .001). Of the triple-negative cancer patients, BRCA2 mutation carriers were older when diagnosed than BRCA1 mutation carriers and noncarriers (P < .01).

Conclusion

These results suggest that tumors associated with BRCA1 mutations may be divided into two distinct groups, triple-negative and non–triple-negative groups. Future studies should seek to determine whether patients with BRCA1 mutations and triple-negative breast cancer respond to treatment better than BRCA-negative patients with similar tumor pathology.

INTRODUCTION

Mutations in the tumor suppressor genes BRCA1 and BRCA2 are believed to be responsible for the majority of hereditary breast cancer cases. It is estimated that women with BRCA1 and BRCA2 mutations have a lifetime risk of developing breast cancer as high as 87%.^1,2 However, evidence indicating whether overall prognosis is poorer for women who have BRCA-related cancers has not been conclusive.^3,4 Currently, treatment recommendations for BRCA-related cancers are similar to those for sporadic breast cancers. It is possible that as treatment regimens become more tumor specific, future patients with BRCA mutations will be treated differently. Thus, it is important to determine clinical characteristics and tumor pathologic features in BRCA carriers that may affect treatment recommendations.

Breast cancer patients display diverse pathologic and clinical features, some of which have prognostic significance. Recent research has defined distinct subtypes of breast cancer using gene expression patterns.^5-7 Based on the molecular profiling of tumors, breast cancers have been divided into those with high expression of the estrogen receptor (ER) gene (luminal A and luminal B subtypes), and those that do not express ER.⁵ Within the ER-negative group, tumors that overexpress the HER2/neu oncogene are named the HER-2/neu–positive subtype.⁵ ER-negative tumors that express genes found in basal epithelial cells and can be stained with antibodies to keratin 5/6 have been identified as basal-like tumors.⁵ A majority of these basal-like tumors are believed to consist of tumors that do not express ER, progesterone receptor (PR), or HER-2/neu (ie, triple-negative tumors).⁸ Several studies have demonstrated that BRCA1-mutation carriers are more likely to be diagnosed with triple-negative breast cancer than noncarriers.^9-11 In contrast, carriers of BRCA2 mutations seem to share similar pathologic characteristics with noncarriers.^12,13 However, previous studies have been limited by relatively small sample sizes. In addition, important clinical information that may influence tumor development has been largely ignored by previous studies. Thus, the purpose of this study is to determine the pathologic characteristics of breast cancers in patients with and without a BRCA mutation, and to describe the clinical features of this population.

PATIENTS AND METHODS

Study Population

Between 1997 and 2006, 1,510 women were seen at the Department of Clinical Cancer Genetics at The University of Texas M. D. Anderson Cancer Center, and underwent BRCA genetic testing; 913 of these women were diagnosed with breast cancer. Of these, 491 women had pathology reports with complete ER, PR, and HER-2/neu status available for review and were thus included in the study. For these 491 women, electronic medical records were reviewed to extract data on clinical characteristics, including ethnicity, age, and clinical stage at diagnosis; age at parity; number of full-term pregnancies; use of oral contraceptive pills and hormone replacement therapy; and genetic-test results for mutations in the BRCA1 and BRCA2 genes. This retrospective study was approved by The University of Texas M. D. Anderson Cancer Center review board.

Tumor Pathology

Tumor pathology for 491 patients with breast cancer was reviewed by one of our designated breast pathologists. Information regarding the histologic type of breast cancer; tumor grade using the modified Black's nuclear grading system; and ER, PR, and HER-2/neu status of breast cancer samples were obtained from the patients’ institutional pathology reports. All invasive breast cancer specimens were routinely evaluated for ER, PR, and HER-2/neu status using immunohistochemistry (IHC). Cases with HER-2/neu staining of 1+, 2+ or 3+ on IHC analysis were further evaluated by fluorescent in situ hybridization for amplification of the HER2/neu gene.

Statistical Analysis

The Kruskal-Wallis exact test was used to compare the number of full-term pregnancies and the median age at diagnosis, menarche, and first full-term pregnancy across the three patient groups (BRCA-negative, BRCA1 mutation carriers, and BRCA2 mutation carriers). A P value less than .05 was considered significant for accepting the hypothesis that at least two of the medians were significantly different from each other. A Wilcoxon rank sum test was used to compare median age at diagnosis, median age at menarche, median age at parity, and number of full-term pregnancies with receptor status in the BRCA1 mutation group.

Fisher's exact test was used to assess the association between type of receptor (ER, PR, or HER-2/neu), fluorescent in situ hybridization for HER-2/neu, nuclear grade, and clinical stage across the BRCA groups. The same test was used when menopause status, use of birth control, use of hormone replacement therapy, and ethnicity were compared across the BRCA groups and by receptor status in the BRCA1 group. A P value less than .05 using the two-tailed Fisher's exact test was considered statistically significant. All of the statistical analyses were performed using SAS version 9.1 for Windows (SAS Institute, Cary, NC).

RESULTS

Of the 491 women included in this study, 391 women (79.6%) tested negative for mutations in the BRCA1 or BRCA2 genes, 56 women (11.4%) had a BRCA1 mutation, 30 women (6.1%) had a BRCA2 mutation, 13 women (2.6%) had variants of uncertain significance in BRCA1 or BRCA2, and one woman (0.2%) had mutations in both BRCA1 and BRCA2. Because the number of women who had variants of uncertain significance or mutations in both genes was too small to produce statistically valid results, these women were excluded from further analysis.

Triple-negative breast cancer was detected in 93 women. Of the women diagnosed with triple-negative breast cancer, 54 women (58.1%) were BRCA negative, 32 women (34.4%) were BRCA1 positive, and seven women (7.5%) were BRCA2 positive. The remainder of the study population was positive for the expression of at least one hormone receptor, or had HER-2/neu–positive disease. BRCA1 carriers were more likely to be diagnosed with triple-negative breast cancer than noncarriers or BRCA2 carriers (Table 1; P < .001). When each receptor was examined alone, BRCA1-associated cancers were more frequently ER and PR negative than were BRCA-negative and BRCA2-associated cancers (P < .001). HER-2/neu expression, as detected by either IHC or fluorescent in situ hybridization, did not differ significantly between mutation carriers and noncarriers (P = .06 and .23, respectively). BRCA1 mutation carriers had higher nuclear grade tumors than did BRCA-negative women or BRCA2 mutation carriers (P < .001). The ER, PR, and HER-2/neu status and nuclear grade of BRCA2-associated cancers were similar to those of BRCA-negative cancers (Table 1). The clinical stages at diagnosis of mutation carriers and noncarriers were similar (P = .58).

Table 1.

Association Between Receptor, ER, PR, HER-2/neu, HER-2/neu (FISH), Nuclear Grade, and Clinical Stage and BRCA Status

Covariate	BRCA Status											P*
		Noncarriers (n = 391)			BRCA1 Mutation Carriers (n = 56)			BRCA2 Mutation Carriers (n = 30)
		No. of Patients	%	No. of Patients	%	No. of Patients	%
Receptor
Nontriple negative	337	86.2	24	42.9	23	76.7	< .001
Triple negative	54	13.8	32	57.1	7	23.33
ER
Negative	90	23.0	38	69.1	8	27.6	< .001
Positive	301	76.9	17	30.9	21	72.4
PR
Negative	143	37.2	37	71.2	10	34.5	< .001
Positive	241	62.7	15	28.9	19	65.5
HER-2/neu (IHC)
Negative	229	85.8	37	97.4	23	95.8	.06
Positive	38	14.2	1	2.6	1	4.2
HER-2/neu (FISH)
Negative	152	74.5	22	88.0	15	88.2	.23
Positive	52	25.5	3	12.0	2	11.8
Nuclear grade
1	43	12.6	2	4.2	1	4.4	< .001
2	167	49.0	5	10.4	9	39.1
3	131	38.4	41	85.4	13	56.5
Clinical stage
1	85	28.4	16	34.0	12	44.5	.58
2	137	45.8	18	38.3	11	44.7
3	74	24.8	13	27.7	4	14.8
4	3	1.0	0	0	0	0	0

Abbreviations: ER, estrogen receptor; PR, progesterone receptor; HER-2, human epidermal growth factor 2; IHC, immunohistochemistry; FISH, fluorescent in situ hybridization.

^*Using Fisher's exact test.

Within the BRCA1 mutation carriers, 32 carriers (57.1%) had triple-negative breast cancer and 24 carriers (42.9%) had non–triple-negative breast cancer. There was a trend for BRCA1 carriers with triple-negative disease to have higher nuclear grade tumors than non–triple-negative BRCA1 carriers (Table 2; P = .06). However, the age at diagnosis and clinical stage did not differ between the two groups (P = .44 and .80, respectively).

Table 2.

Association Between Age of Diagnosis, Nuclear Grade, and Clinical Stage, and Triple-Negative Status in BRCA1 Mutation Carriers

Covariate	Receptor Status							P
		Triple Negative (n = 32)			Nontriple Negative (n = 24)
		No. of Patients	%	No. of Patients	%
Age at diagnosis, years
No. of observations	32			24
Median	41.5			42.5			.44*
Range	27-71			25-61
Nuclear grade
1	0	0.0	2	10.0	.06†
2	2	6.5	3	15.0
3	29	93.5	15	75.0
Clinical stage
1	9	30.0	7	41.2	.80†
2	12	40.0	6	35.3
3	9	30.0	4	23.5
4	0	0.0	0	0.0

^*Using Wilcoxon rank sum test.

^†Using Fisher's exact test.

Breast cancer risk factors did not differ between women with or without BRCA mutations. Ethnic background was similar in mutation carriers and noncarriers (P = .78). Although there was a trend for BRCA1 carriers to be diagnosed at a younger age, this did not reach statistical significance (Table 3; P = .07). The median age of menarche ranged from 12 to 13 years and was similar between groups (P = .43). The median age of first full-term pregnancy and number of full-term pregnancies did not differ significantly between BRCA-positive and BRCA-negative women (P = .31 and .87, respectively). Women in the three mutation status groups were equally likely to have a history of using oral contraceptive pills (Table 3; P = .94). Menopause status at the time of diagnosis did not differ between BRCA mutation carriers and noncarriers. In contrast, the history of hormone replacement therapy differed between the three groups (P < .05); more women who were BRCA negative or who were BRCA2 mutation carriers had received hormone replacement therapy than women who were BRCA1 mutation carriers.

Table 3.

Association Between Age at Diagnosis, Age at Menarche, Age at Parity, Number of Full-Term Pregnancies, Menopause Status, Use of Oral Contraceptives, Hormone Replacement Therapy, Ethnicity, and BRCA Group

Variable	BRCA											P
		Noncarriers (n = 391)			BRCA1 Mutation Carriers (n = 56)			BRCA2 Mutation Carriers (n = 30)
		No. of Patients	%	No. of Patients	%	No. of Patients	%
Age at diagnosis, years							.07*
No. of observations	391			56			30
Median	44			42			43
Range	21-75			25-71			30-67
Age at menarche, years							.43*
No. of observations	367			54			30
Median	12			13			13
Range	8-18			9-18			11-15
Age at parity, years							.31*
No. of observations	303			45			27
Median	26			25			25
Range	15-42			16-40			19-43
No. of full-term pregnancies							.87*
No. of observations	383			55			30
Median, years	2			2			2
Range, years	0-9			0-10			0-4
Menopause status
Premenopausal	226	60.7	39	69.6	14	46.7	.12†
Postmenopausal	146	39.3	17	30.4	16	53.3
History of oral contraceptive use
No	69	18.9	11	20.4	5	16.7	.94†
Yes	296	81.1	43	79.6	25	83.3
History of hormone replacement therapy use
No	285	78.1	49	90.7	21	70.0	.04†
Yes	80	21.9	5	9.3	9	30.0
Ethnicity
White	280	71.9	41	73.2	20	66.7	.78†
Hispanic	49	12.6	5	8.9	4	13.3
Black	16	4.1	4	7.1	1	3.3
Asian	9	2.3	1	1.83	2	6.7
Ashkenazi Jew	35	9.0	5	8.9	3	10.0

^*P value from Kruskall-Wallis exact test.

^†P value from Fisher's exact test.

The breast cancer risk factors in triple-negative carriers and noncarriers of BRCA1 and BRCA2 mutations were also compared. Among the patients with triple-negative breast cancer, BRCA2-associated cancers were diagnosed at a later age than were BRCA1- and BRCA-negative cancers (Table 4; P < .01). As in the general study population, age of menarche, age at parity, number of full-term pregnancies, and use of oral contraception were similar between BRCA carriers and noncarriers (Table 4). Although BRCA2 mutation carriers were more likely than noncarriers to be diagnosed postmenopausal and to have a higher frequency of prior hormone replacement therapy, these did not reach statistical significance.

Table 4.

Association Between Age at Diagnosis, Age at Menarche, Age at Parity, Number of Full-Term Pregnancies, Menopause, Oral Contraceptive and Hormone Replacement Therapy Use, and BRCA Group in the Triple-Negative Group of Patients Only

Variable	BRCA Status											P
		No. of Noncarriers (n = 54)			No. of BRCA1 Mutation Carriers (%), n = 32			No. of BRCA2 Mutation Carriers (%), n = 7
		No. of Patients	%	No. of Patients	%	No. of Patients	%
Age at diagnosis, years							.01*
No. of observations	54			32			7
Median	42			41.5			52
Range	24-69			27-71			50-54
Age at menarche, years							.30*
No. of observations	49			32			7
Median	12			13			13
Range	8-16			10-17			11-14
Age at parity, years							.36*
No. of observations	38			27			7
Median	26			25			21
Range	15-36			17-40			19-43
No. of full-term pregnancies							.50*
No. of observations	54			32			7
Median, years	2			2			2
Range, years	0-9			0-10			1-3
Menopause status
Premenopausal	34	68.0	23	71.9	2	28.6	.10†
Postmenopausal	16	32.0	9	28.1	5	71.4
History of oral contraceptive use
No	6	11.8	7	22.6	1	14.3	.39†
Yes	45	88.2	24	77.4	6	85.7
History of hormone replacement therapy use
No	40	78.4	28	90.3	4	57.1	.09†
Yes	11	21.6	3	9.7	3	42.9

^*P value from Kruskall-Wallis exact test.

^†P value from Fisher's exact test.

Within the BRCA1 carriers, the majority of clinical characteristics were similar between women with triple-negative cancer and women with non–triple-negative cancer. Although age at menarche was younger in women with triple-negative cancer (Table 5; P < .05), age at parity, number of full-term pregnancies, menopause status, history of oral contraceptive use, and history of hormone replacement therapy use was similar between the two groups.

Table 5.

Association Between Age at Menarche, Age at Parity, Number of Full-Term Pregnancies, Menopause Status, History of Oral Contraceptive Use, and History of Hormone Replacement Therapy Use and Triple-Negative Status in BRCA1 Mutation Carriers

Variable	Receptor Status							P
		Triple Negative (n = 32)			Nontriple Negative (n = 24)
		No. of Patients	%	No. of Patients	%
Age at menarche, years
No. of observations	32			22
Median	13			12			.04*
Range	10-17			9-18
Age at parity, years
No. of observations	27			17
Median	25.5			25			.82*
Range	16-40			17-40
No. of full-term pregnancies
No. of observations	32			22			.13*
Median, years	2			2
Range, years	0-10			0-3
Menopause status
Premenopausal	23	71.9	16	66.6	.77†
Postmenopausal	9	28.1	8	33.3
History of oral contraceptive use
No	7	22.6	4	17.4	.74†
Yes	24	77.4	19	82.6
History of hormone replacement therapy use
No	28	90.3	21	91.3	1.00†
Yes	3	9.6	2	8.7

^*Using Wilcoxon rank sum test.

^†Using Fisher's exact test.

DISCUSSION

In this study, we identified clinical and pathologic characteristics of tumors in a cohort of women with BRCA-positive and BRCA-negative breast cancer. Significantly more of the BRCA1-related breast cancers were triple negative than were the BRCA-negative and BRCA2-related breast cancers. In addition, significantly more BRCA1-related cancers were poorly differentiated and had a higher modified Black's nuclear grade. Although the median age at diagnosis was similar in BRCA1 mutation carriers, BRCA2 mutation carriers, and BRCA-negative women when receptor status was not considered, women with triple-negative breast cancer who were BRCA2 mutation carriers were diagnosed at a later age than were BRCA1 mutation carriers and noncarriers. It is possible that referral bias reduced the ability to detect differences in age and menopausal status between BRCA mutation carriers and noncarriers. Within our institution, physicians are likely to refer patients to genetic counseling and testing if they are younger than age 50 years when diagnosed with breast cancer, but only refer women diagnosed at an older age if they have a substantial family history of breast or ovarian cancer. Ideally, it would be beneficial to compare BRCA1 and BRCA2 mutation carriers with the general breast cancer population; however, this is difficult because family history information is often incomplete and not always accurate in patients who have not been seen by a genetic counselor.

Several studies have evaluated hormone receptor status in BRCA1 mutation carriers. In 39 women, those with a BRCA1 mutation or a known familial BRCA1 mutation more frequently had ER- and PR-negative tumors than control participants.¹⁴ Studies of women in several different ethnic groups, including Ashkenazi Jew, Japanese, and Swede, have demonstrated that BRCA1 mutation carriers were more likely to have ER-negative breast cancer.^12,13,15,16 PR-negative breast cancer also has been associated with BRCA1 mutation carriers.^12,15,16 However, the relationship between BRCA mutation status and HER-2/neu positivity has been inconsistent. In one study of six BRCA1 mutation carriers, no association between HER-2/neu status and BRCA1 status was found, although there was a trend for BRCA1 carriers to have ER- or PR-negative disease more frequently than noncarriers.¹⁷ However, in other studies, BRCA1 status was associated with HER-2/neu–negative tumors.^13,16 Here, 38 of 56 BRCA1-related tumors were ER negative and 37 of 56 BRCA1-related tumors were PR negative, which is consistent with previous reports. In addition, we provide evidence from a larger cohort that HER-2/neu expression status was similar between BRCA1 mutation carriers and noncarriers.

Previous reports describing the distribution of ER, PR, and HER-2/neu positivity in BRCA carriers have been inconclusive mainly because, in these studies, BRCA1- and BRCA2-mutation carriers were grouped together instead of being examined as two distinct groups.^18,19 In 58 Ashkenazi Jewish women, significantly fewer BRCA1 and BRCA2 mutation carriers had ER-, PR-, or HER-2/neu–positive disease than noncarriers.¹⁸ In another study,¹⁹ ER-positive disease was less common in 39 BRCA1 and BRCA2 mutation carriers than noncarriers, but no differences were observed between the two groups in PR or HER-2/neu positivity. Although these authors reported finding no difference between BRCA1 and BRCA2 tumor pathology, the majority of participants (30 of 39) were BRCA2 mutation carriers. Thus, it is possible that if more BRCA1-associated cancers had been examined, differences would have emerged.¹⁹

Several previous studies have examined the relationship between BRCA mutation status and triple-negative tumor pathology. ^9-11,20 In these studies, 50% to 88% of BRCA1 carriers were diagnosed with triple-negative breast cancer compared with 14.6% to 34% of BRCA noncarriers. In addition, because triple-negative tumors are believed to constitute a majority of the basal-like tumors reported,⁸ it is relevant to note that in a series of studies by Foulkes et al,^21,22 BRCA1-related cancers were more likely to be basal-like than sporadic cancers. However, these studies have been limited by the number of BRCA1-related breast cancers examined. In this study, we expanded on these findings in a much larger cohort, in which we found 57.1% of BRCA1 carriers (32 of 56) had triple-negative tumors, compared with 13.8% of BRCA1 carriers in BRCA noncarriers. Together, this evidence indicates that a significant number of BRCA1 patients are triple negative.

In our study, nuclear grade at diagnosis was higher for BRCA1-related breast cancers than for other breast cancers. This result is consistent with previous reports. In a study of Ashkenazi Jewish women, 76.5% of BRCA1-positive tumors had a high nuclear grade compared with only 27.3% of BRCA-negative tumors.¹⁵ In a Swedish report, BRCA1-related tumors were more likely to have a nuclear grade of 3 than non-BRCA related tumors.¹⁶ Several previous reports also have shown that BRCA1-related cancers were of a higher histologic grade.^{9,10,13,16,18} Although another study¹⁸ reported that nuclear grade did not differ in sporadic and BRCA-related tumors, this study analyzed BRCA1 and BRCA2 mutation carriers as one group. As our results have indicated, because BRCA1- and BRCA2-related tumors seem to have different pathologic characteristics, it is possible that combining these two groups obscured the differences in nuclear grade between BRCA1 and sporadic cases.¹⁸

In this study, the pathology of BRCA2-related breast cancer was similar to that of BRCA-negative breast cancers. One previous study suggested that BRCA2-related cancers have tumor pathology that is between that of BRCA1 and sporadic cancers; however, this study was limited by the small number of BRCA2 patients available for examination.¹⁴ Other research has demonstrated that BRCA2-associated breast cancers and sporadic breast cancers are equally likely to be triple negative⁹ or to have tumor pathology similar to sporadic cases.^12,13

Our study, like previous reports, suggests that BRCA1 mutation carriers have tumor pathology that differs from BRCA-negative patients and BRCA2 mutation carriers. Because of these differences, researchers have examined whether tumor biology can predict BRCA mutation status. In a study of 207 families, the sensitivity of BRCAPRO, a traditional model of predicting mutations in the BRCA genes using family history information, was increased by including the ER and PR receptor status and pathologic grade of the tumor.²³ Additional pathologic variables that may predict BRCA1 mutation status include Ki67 and epidermal growth factor receptor.¹⁴ In young (age younger than 54 years) women with breast cancer, high levels of Ki67 expression predicted a chance of having a BRCA1 mutation as high as 75%.¹⁴ Based on these studies and an increasing amount of evidence suggesting that BRCA1 tumors have unique pathologic features, clinicians should perhaps consider using pathology results along with family history information when deciding whether a patient is at an increased risk for hereditary breast cancer. This may be particularly useful when family history information results in an intermediate concern for hereditary cancer.¹⁴ However, because it appears that BRCA2-related cancers have pathology similar to that of non-BRCA carriers, it is currently unclear whether pathologic results may be used in predicting BRCA2 mutation status. Additional research should be conducted to determine how much emphasis should be placed on tumor pathology and how this information can be included in already established models.

Our study demonstrated that BRCA2 mutation carriers with triple-negative breast cancer were older at diagnosis than BRCA1 carriers and noncarriers with triple-negative cancer. However, the general study population of BRCA2 mutation carriers showed no significant difference in the age at diagnosis compared with BRCA1 mutation carriers and noncarriers. Why BRCA2 carriers develop triple-negative cancer at a later age than the other groups has yet to be determined. Referral bias and the small number of BRCA2 mutation carriers may have influenced these results. In our study, we found that more BRCA2 carriers had triple-negative cancer diagnosed postmenopausal, and that more of these carriers had undergone hormone replacement therapy than BRCA1 carriers. Therefore, it is possible that menopausal status or the use of hormone replacement therapy may be involved. Additional research with a larger number of BRCA2 carriers with triple-negative disease should be conducted to test this hypothesis.

One limitation of this study was the small number of women who were BRCA2 mutation carriers and had triple-negative breast cancer. Because this combination of mutation status and tumor pathology appears to be uncommon, it would be interesting to examine this group in further detail. Future studies that explore this population should be conducted. Another limitation of our study was the possible referral bias in our study population. Because the BRCA-negative women were referred to our genetics program, they may not fully represent sporadic breast cancer cases, resulting in a bias in our control group. However, BRCA status and family history information is often unknown in unselected breast cancer patients, thus making comparisons between BRCA-positive and unselected breast cancer patients difficult. Future studies that screen prospectively for BRCA mutations in an unselected breast cancer population should be conducted to examine differences in tumor pathologic features and clinical characteristics in BRCA-positive and BRCA-negative patients in the general breast cancer population.

In conclusion, our results suggest that BRCA1-related breast cancers may be divided into two subgroups: one group consisting of triple-negative, high-grade tumors, and the other group consisting of pathology more consistent with breast cancer observed in BRCA noncarriers. Future studies should determine whether treatment outcomes differ for BRCA1 mutation carriers depending on different tumor pathology (ie, triple negative v non–triple negative). With the identification of novel targets for BRCA1- and BRCA2-related tumors, such as the poly(adenosine diphosphate-ribose) polymerase-1 pathway,²⁴ the efficacy of agents that target these pathways, such as poly(adenosine diphosphate-ribose) polymerase-1 inhibitors, should be examined.

AUTHORS’ DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

Although all authors completed the disclosure declaration, the following author(s) indicated a financial or other interest that is relevant to the subject matter under consideration in this article. Certain relationships marked with a “U” are those for which no compensation was received; those relationships marked with a “C” were compensated. For a detailed description of the disclosure categories, or for more information about ASCO's conflict of interest policy, please refer to the Author Disclosure Declaration and the Disclosures of Potential Conflicts of Interest section in Information for Contributors.

Employment or Leadership Position: None Consultant or Advisory Role: Banu K. Arun, Pfizer Inc (C) Stock Ownership: None Honoraria: Banu K. Arun, Pfizer Inc, AstraZeneca Research Funding: Banu K. Arun, National Cancer Institute, Pfizer Inc, AstraZeneca Expert Testimony: None Other Remuneration: None

AUTHOR CONTRIBUTIONS

Conception and design: Deann P. Atchley, Vicente Valero, Banu K. Arun

Financial support: Banu K. Arun

Administrative support: Deann P. Atchley, Gabriel N. Hortobagyi, Banu K. Arun

Provision of study materials or patients: Deann P. Atchley, Vicente Valero, Banu K. Arun

Collection and assembly of data: Deann P. Atchley, Christopher I. Amos, Banu K. Arun

Data analysis and interpretation: Deann P. Atchley, Constance T. Albarracin, Adriana Lopez, Christopher I. Amos, Ana Maria Gonzalez-Angulo, Gabriel N. Hortobagyi, Banu K. Arun

Manuscript writing: Deann P. Atchley, Constance T. Albarracin, Ana Maria Gonzalez-Angulo, Gabriel N. Hortobagyi, Banu K. Arun

Final approval of manuscript: Deann P. Atchley, Constance T. Albarracin, Vicente Valero, Christopher I. Amos, Ana Maria Gonzalez-Angulo, Gabriel N. Hortobagyi, Banu K. Arun