Clinicopathologic Features of Breast Cancers that Develop in Women with Previous Benign Breast Disease

Abstract

Background

Women with benign breast disease (BBD) have an increased risk for breast cancer (BC). Almost 30% of all BCs develop in women with prior BBD. Information about features of the expected BCs after BBD would enhance individualized surveillance and prevention strategies for these women. We sought to characterize BCs developing in a large cohort of women with BBD.

Methods

Our cohort includes 13,485 women who underwent breast biopsy for mammographic or palpable concern between 1967-2001. Biopsy slides were reviewed and classified as nonproliferative (NP), proliferative disease without atypia (PDWA), or atypical hyperplasia (AH). BCs were identified by follow-up questionnaires, medical records, and Tumor Registry data. BC tissues were obtained and reviewed.

Results

With median follow-up of 15.8 years, 1273 women developed BC. Most BCs were invasive (81%), among which 61% were ductal, 13% mixed ductal/lobular, and 14% lobular. Two-thirds were intermediate or high-grade, and 29% were node positive. Cancer characteristics were similar across the three histologic categories of BBD, with similar frequency of DCIS, invasive disease, tumor size, time to invasive BC, histologic type of BC, nodal positivity and HER2 positivity. Women with AH had a higher frequency of ER+ BC (91%) than women with PDWA (80%) or NP (85%), p=0.02.

Conclusions

A substantial proportion of all BCs develop in women with prior BBD. The majority of BCs after BBD are invasive tumors of ductal type, with a substantial proportion node positive. Of all BCs, 84% were ER-positive. Prevention therapy should be strongly encouraged in higher-risk women with BBD.

Introduction

When a woman has either a palpable or mammographic breast concern, a biopsy is often necessary to rule out the presence of malignancy. Approximately 75% of such biopsies will have benign findings, and this population of women is said to have benign breast disease (BBD). It is estimated that over a million women in the United States undergo a benign breast biopsy annually.¹ This population has long been recognized as having an increased risk for subsequent breast cancer.^2-5

A substantial proportion of breast cancers diagnosed in the United States develop in women with a history of prior benign breast biopsy. The Breast Cancer Surveillance Consortium followed over one million women undergoing screening mammography at community centers and found that almost 15,000 women had developed invasive breast cancer over 5.3 years of follow-up, Of these women, 29% had at least one prior benign biopsy.⁶ In an HMO-based study, 20-25% of women with a newly diagnosed breast cancer reported having previous BBD.⁷ Given efforts to predict the absolute risk of later breast cancer for women following a benign breast biopsy,⁸ it would be clinically advantageous to identify the likely features of the expected breast cancers so as to individualize prevention and surveillance strategies. Moreover, specific information about the breast cancers – time from BBD to cancer, side of cancer vs side of benign, nodal status, histology – can facilitate understanding of possible progression pathways in breast carcinogenesis.

We have addressed these questions in a cohort of 13,485 women who had either excisional or core needle breast biopsies with benign findings between 1967 and 2001. With current follow-up, 1,273 have developed breast cancer. Here we present the clinical and pathologic features of these breast cancers, comparing the findings by histologic category of original benign breast disease and by patient age at original BBD.

Methods

Study Population

Our initial Benign Breast Disease cohort included all women who had a benign breast biopsy from January 1, 1967 to December 31, 1991 at Mayo Clinic, Rochester, MN.⁵ We have subsequently extended the cohort through December 31, 2001. After meeting prospectively established inclusion and exclusion criteria,⁵ the extended cohort now includes 13,485 women with 231,806 person years of follow up.⁹ A minimum interval of 6 months follow-up information was required for inclusion in the cohort; among all women with a benign biopsy according to institutional records, 6.4% were excluded from eligibility due to inadequate follow-up (either none or less than six months). Women could have had core and/or excisional biopsies. All women found to have atypical hyperplasia on core needle biopsy also underwent surgical excision; if malignancy was found then the patient was deemed to have a malignant biopsy and was ineligible/excluded from the cohort at the time of initial cohort ascertainment. In addition, breast cancer that occurred within 6 months of a benign breast biopsy was considered to be a diagnosis of malignancy associated with the initial biopsy (i.e. an undetected primary) and was also an original criterion for exclusion from the cohort.⁵ The definitions for family history categories are as previously described.⁵

Histology

Stored hematoxylin-and-eosin–stained sections from each participant were evaluated by a breast pathologist (DWV) who was unaware of the initial histologic diagnoses and subsequent patient outcomes. Benign biopsy findings were classified according to the criteria of Page et al.: nonproliferative changes (NP), proliferative disease without atypia (PDWA), and proliferative disease with atypia (AH; includes atypical ductal hyperplasia, atypical lobular hyperplasia, or both).^{2, 5, 10} Atypical lobular hyperplasia was distinguished from lobular carcinoma in situ by lack of significant expansion of involved acini or effaced lobular architecture. Biopsy specimens were designated as having proliferative fibrocystic changes if they contained any of the following: ductal hyperplasia (greater than mild), papilloma, radial scar, sclerosing adenosis, columnar cell hyperplasia, or flat epithelial atypia. Cysts, fibroadenoma, or columnar changes without hyperplasia were considered nonproliferative, unless they also contained one of the proliferative lesions denoted above. The extent of lobular involution in normal background lobules was categorized as previously described (none, partial involution, or complete involution).¹¹

For the breast cancers, medical records and tissue slides were obtained for review. Our study pathologist verified the histology of the breast cancers. The definitions of laterality were: ipsilateral, cancer occurring in the same breast as the benign biopsy; contralateral, cancer in the opposite breast; and bilateral, the occurrence of cancer in both breasts within 6 months of each other. Invasive breast carcinomas were graded according to the Nottingham criteria of Elston and Ellis.¹²

Follow-up procedures

Follow-up for breast cancer events, including both invasive cancer and ductal carcinoma in situ (DCIS), and risk factor information were obtained for all women using the Mayo medical record, Mayo Tumor Registry, and a study-specific questionnaire to capture diagnoses made outside the Mayo Clinic system. Follow-up was defined as the number of days from benign biopsy to date of breast cancer diagnosis, death, or last contact. The study protocol, including patient contact and follow up methods, was approved by the Mayo Clinic Institutional Review Board.^{5, 9}

Statistical analyses

Variables were summarized using frequencies and percents for categorical variables and medians and ranges for continuous variables. We compared cancer-related attributes across levels of other variables of interest using chi-square tests. Following this, age-adjusted comparisons were made using multi-categorical nominal logistic regression analysis.

Cumulative incidence of breast cancer by levels of benign histology was calculated using Kaplan-Meier curves and accounting for death as a competing risk. We estimated relative risks with standardized incidence ratios (SIR) and 95% confidence intervals (CI), dividing observed numbers of incident breast cancers by expected counts. We used the Iowa Surveillance Epidemiology and End Results (SEER) registry as the reference population because of its proximity to the Mayo Clinic catchment area and similar population demographics compared with our cohort.⁵ Tests for heterogeneity in SIRs across levels of demographic and clinical variables were carried out using a Poisson regression analysis that accounted for the population based expected event rate for each individual using an offset term. All statistical sets were two-sided, and all analyses were conducted using the SAS version 9.3(SAS Institute, Inc., Cary, NC) software system.

Results

The 13,485 women in this cohort had a benign breast biopsy (either core or excisional) at the Mayo Clinic, Rochester MN, between 1967 and 2001. Over a median follow-up of 15.8 years, 1,273 women developed either invasive breast cancer or DCIS. The clinical and histologic characteristics of this cohort, with associated risks for later breast cancer, are displayed in Table 1. For the BBD cohort overall, there is an increased risk for breast cancer (SIR 1.64 [95% CI 1.56-1.74]). Further increases in risk are associated with several features, including proliferative breast disease with or without atypia, family history of breast cancer, younger age at BBD, and lesser extent of lobular involution, as previously described.^{5, 11}

Table 1. Mayo Benign Breast Disease (BBD) Cohort: Clinical and Histologic Characteristics and Risk of Breast Cancer.

Characteristic	No. Women	Person Years	Observed Events	Expected Events	SIR (95% CI)	p-value1
Overall	13485	231806	1273	774	1.64 (1.56, 1.74)
Histology of BBD						<.0001
NP	8463	151521	627	483	1.30 (1.20, 1.40)
PDWA	4311	71041	497	256	1.94 (1.78, 2.12)
AH	711	9244	149	36	4.18 (3.53, 4.90)
Age of BBD						0.0105
<45	4402	87469	343	189	1.81 (1.63, 2.02)
45-55	3960	74743	460	268	1.72 (1.56, 1.88)
55+	5123	69595	470	317	1.48 (1.35, 1.62)
Involution						<.0001
None	2245	45136	242	118	2.05 (1.80, 2.33)
Partial	6650	119639	690	407	1.70 (1.57, 1.83)
Complete	3308	45979	208	191	1.09 (0.95, 1.25)
Family history of breast cancer						<.0001
None	7761	132712	614	444	1.38 (1.27, 1.50)
Weak	3198	59615	385	196	1.97 (1.78, 2.18)
Strong	1673	31482	247	107	2.32 (2.04, 2.63)

Standardized incidence ratios (SIRs) compare the observed number of breast cancer events with the number expected on the basis of Iowa SEER Data. All analyses account for the effects of age and calendar period.

¹Test of heterogeneity across SIRs across levels of a given characteristic.

The cumulative incidence of breast cancer in this cohort is shown in Figure 1. Women with AH had an approximate 1% per year risk of breast cancer, and women with PDWA had an approximate 0.5% risk per year. Women with NP findings had a mild increase in relative risk (SIR 1.3) that persisted for approximately 20 years after benign biopsy.

Figure 1.

Cumulative Incidence of Breast Cancer by Histologic Category of Benign Breast Disease

The characteristics of the breast cancers, overall and by category of original benign findings, are featured in Table 2. 81% of cancers were invasive and 19% in situ, and these proportions did not vary across benign histologic categories (Figure 2A). There was a predominance of ipsilateral cancer events relative to the side of benign biopsy, which was most marked in women with AH (90 ipsilateral vs 48 contralateral, p = 0.070) comparing women with AH to those without (Figure 2B). The median time to breast cancer was 8.8 years in women with AH, vs 10.9 years for PDWA and 11.2 years for women with NP findings (p= 0.053). In women with atypical hyperplasia, the time to DCIS was considerably shorter (median 4.5 years) than for invasive cancer (median 9.1 years) consistent with a possible progression pathway. This relationship was not seen, however, in women with other types of proliferative disease or those with non-proliferative findings. Comparing the subgroups of ADH and ALH, the interval to any cancer (DCIS or invasive) was similar for women with ADH and ALH (9.16 years vs 10.65 years, respectively, p=0.21). The interval to DCIS (without invasive cancer) appeared shorter among women with ADH (5.12 years) versus ALH (9.20 years, p=0.10), although this was not statistically different, possibly due to small numbers of events within these subgroups.

Table 2. Characteristics of Breast Cancers by Histologic Category of Benign Breast Disease.

Characteristic	Total (N=1273)	NP (N=627)	PDWA (N=497)	AH (N=149)	p-value1	p-value2
Cancer Type					0.3046	0.2851
Missing	79	39	33	7
Invasive	963 (80.7%)	482 (82.0%)	364 (78.4%)	117 (82.4%)
In situ	231 (19.3%)	106 (18.0%)	100 (21.6%)	25 (17.6%)
Breast Cancer Sidedness3					0.0712	0.0696
Missing	23	13	6	4
Bilateral	44 (3.5%)	18 (2.9%)	19 (3.9%)	7 (4.8%)
Ipsilateral	685 (54.8%)	319 (52.0%)	276 (56.2%)	90 (62.1%)
Contralateral	521 (41.7%)	277 (45.1%)	196 (39.9%)	48 (33.1%)
Time to Breast Cancer, All					0.0020	0.0530
N	1273	627	497	149
Mean (SD)	12.59 (9.13)	13.17 (9.57)	12.64 (8.93)	9.98 (7.24)
Median	10.73	11.24	10.86	8.80
Time to Breast Cancer, Invasive					0.1026	0.2474
N	963	482	364	117
Mean (SD)	12.50 (9.24)	12.61 (9.56)	12.97 (9.28)	10.62 (7.49)
Median	10.63	10.48	11.28	9.07
Time to Breast Cancer, In situ					0.0001	0.0148
N	231	106	100	25
Mean (SD)	13.05 (9.11)	15.50 (10.15)	11.93 (7.77)	7.11 (5.27)
Median	11.39	12.97	10.80	4.51
Invasive Cancers Only					0.3680	0.3937
Histology
Missing	12	8	3	1
Ductal	580 (61.0%)	293 (61.8%)	224 (62.0%)	63 (54.3%)
Lobular	130 (13.7%)	57 (12.0%)	55 (15.2%)	18 (15.5%)
Mixed Ductal/Lobular	122 (12.8%)	67 (14.1%)	37 (10.2%)	18 (15.5%)
Other	119 (12.5%)	57 (12.0%)	45 (12.5%)	17 (14.7%)
Tumor Grade					0.0071	0.0096
Missing	123	59	46	18
Grade 1	275 (32.7%)	128 (30.3%)	105 (33.0%)	42 (42.4%)
Grade 2	352 (41.9%)	191 (45.2%)	118 (37.1%)	43 (43.4%)
Grade 3	213 (25.4%)	104 (24.6%)	95 (29.9%)	14 (14.1%)
Nodal status					0.2699	0.2681
Missing	49	27	20	2
Negative	653 (71.4%)	330 (72.5%)	236 (68.6%)	87 (75.7%)
Positive	261 (28.6%)	125 (27.5%)	108 (31.4%)	28 (24.3%)
Tumor size					0.0167	0.2692
N	887	443	335	109
Mean (SD)	20.01 (16.43)	19.86 (17.19)	20.94 (15.73)	17.75 (15.27)
Median	15.00	15.00	17.00	14.00
ER status					0.0145	0.0191
Missing	61	33	23	5
Negative	145 (16.1%)	67 (14.9%)	68 (19.9%)	10 (8.9%)
Positive	757 (83.9%)	382 (85.1%)	273 (80.1%)	102 (91.1%)
PR status					0.0556	0.0572
Missing	116	70	37	9
Negative	212 (25.0%)	107 (26.0%)	88 (26.9%)	17 (15.7%)
Positive	635 (75.0%)	305 (74.0%)	239 (73.1%)	91 (84.3%)
HER2 intensity					0.2920	0.2792
Missing	542	285	195	62
None	202 (48.0%)	94 (47.7%)	85 (50.3%)	23 (41.8%)
weak (1+)	123 (29.2%)	55 (27.9%)	44 (26.0%)	24 (43.6%)
moderate (2+)	52 (12.4%)	27 (13.7%)	21 (12.4%)	4 (7.3%)
strong (3+)	44 (10.5%)	21 (10.7%)	19 (11.2%)	4 (7.3%)

¹unadjusted p-value from chi-square test

²p-value adjusted for age at initial BBD diagnosis using multi-categorical nominal logistic regression

³Side of breast cancer relative to the side of the original benign biopsy.

Figure 2.

Features of Breast Cancers by Histologic Category of Previous BBD. (A) Side and type of breast cancers (invasive, in situ);* (B) Side and timing of breast cancers; (C) Histology of invasive breast cancers; (D) Grade of invasive breast cancers

The majority of invasive cancers were of ductal histology (61%), observed for all categories of BBD. Another 14% were lobular, and 13% were mixed ductal-lobular (Figure 2C). Among invasive cancers, 33% were grade 1 (Nottingham), 42% grade 2, and 25% grade 3 (Figure 2D). The proportion of grade 1 breast cancer was higher in women with AH compared to women with NP or PDWA (p=0.009). Nodal status was positive in 28.6% of women with invasive BC and did not differ with regard to original benign histology (p=0.27). Tumor size did not differ significantly across the benign histologic categories. Women with atypical hyperplasia were more likely to have ER and PR positive disease. HER-2 overexpression was seen in 10.5% of BCs, which did not differ by original BBD histology (p=0.279).

We asked whether or not a woman's age at BBD (<45, 45-55, 55+) was associated with later BC phenotype (Table 3). The proportion of invasive and in situ cancers did not differ by age grouping, nor did laterality or histologic type of invasive cancer. However, women < age 45, compared to those age 55+, were more likely to have a grade 3 cancer (33% vs 22%, p=0.009), larger tumor size (median 17mm vs 15mm, p=0.003) and node-positive disease (28% vs 21%, p=0.078). ER, PR, and HER-2 status did not differ across these age groups. Comparing breast cancer features by original category of lobular involution, those women with no lobular involution at the time of benign biopsy were more likely to have node-positive disease (36%) compared to those with partial involution (28%) and those with complete involution (20%) (p=0.036, data controlled for age at benign biopsy but not shown).

Table 3. Characteristics of breast cancers by age at initial BBD diagnosis.

Characteristic	Total (N=1273)	<45 (N=343)	45-55 (N=460)	55+ (N=470)	p-value1
Cancer Type					0.8307
Missing	79	18	26	35
Invasive	963 (80.7%)	260 (80.0%)	354 (81.6%)	349 (80.2%)
In situ	231 (19.3%)	65 (20.0%)	80 (18.4%)	86 (19.8%)
Breast Cancer Sidedness (all cancers)2					0.5317
Missing	23	8	7	8
Bilateral	44 (3.5%)	15 (4.5%)	13 (2.9%)	16 (3.5%)
Ipsilateral	685 (54.8%)	176 (52.5%)	245 (54.1%)	264 (57.1%)
Contralateral	521 (41.7%)	144 (43.0%)	195 (43.0%)	182 (39.4%)
Time to Breast Cancer, All					<0.0001
N	1273	343	460	470
Mean (SD)	12.59 (9.13)	17.09 (10.70)	13.51 (8.67)	8.40 (5.96)
Median	10.73	14.55	12.11	7.30
Time to Breast Cancer, Invasive					<0.0001
N	963	260	354	349
Mean (SD)	12.50 (9.24)	16.76 (10.85)	13.32 (8.95)	8.50 (6.08)
Median	10.63	14.02	11.83	7.31
Time to Breast Cancer, In situ					<0.0001
N	231	65	80	86
Mean (SD)	13.05 (9.11)	19.02 (10.34)	14.03 (8.04)	7.62 (5.12)
Median	11.39	18.38	13.62	6.18
Invasive Cancers Only
Histology					0.9262
Missing	12	5	2	5
Ductal	580 (61.0%)	156 (61.2%)	216 (61.4%)	208 (60.5%)
Lobular	130 (13.7%)	32 (12.5%)	53 (15.1%)	45 (13.1%)
Mixed Ductal/Lobular	122 (12.8%)	32 (12.5%)	42 (11.9%)	48 (14.0%)
Other	119 (12.5%)	35 (13.7%)	41 (11.6%)	43 (12.5%)
Tumor Grade					0.0095
Missing	123	33	46	44
Grade 1	275 (32.7%)	71 (31.3%)	90 (29.2%)	114 (37.4%)
Grade 2	352 (41.9%)	86 (37.9%)	132 (42.9%)	134 (43.9%)
Grade 3	213 (25.4%)	70 (30.8%)	86 (27.9%)	57 (18.7%)
Nodal status					0.0782
Missing	49	9	16	24
Negative	653 (71.4%)	166 (66.1%)	245 (72.5%)	242 (74.5%)
Positive	261 (28.6%)	85 (33.9%)	93 (27.5%)	83 (25.5%)
Tumor size					0.0034
N	887	239	326	322
Mean (SD)	20.01 (16.43)	23.24 (21.32)	20.17 (15.22)	17.46 (12.62)
Median	15.00	17.00	16.00	15.00
ER status					0.2505
Missing	61	14	21	26
Negative	145 (16.1%)	38 (15.4%)	62 (18.6%)	45 (13.9%)
Positive	757 (83.9%)	208 (84.6%)	271 (81.4%)	278 (86.1%)
PR status					0.2373
Missing	116	24	46	46
Negative	212 (25.0%)	57 (24.2%)	87 (28.2%)	68 (22.4%)
Positive	635 (75.0%)	179 (75.8%)	221 (71.8%)	235 (77.6%)
HER2 intensity					0.4698
Missing	542	162	180	200
None	202 (48.0%)	45 (45.9%)	93 (53.4%)	64 (43.0%)
weak (1+)	123 (29.2%)	31 (31.6%)	44 (25.3%)	48 (32.2%)
moderate (2+)	52 (12.4%)	11 (11.2%)	18 (10.3%)	23 (15.4%)
strong (3+)	44 (10.5%)	11 (11.2%)	19 (10.9%)	14 (9.4%)

¹chi-square p-value.

²Side of breast cancer relative to the side of the original benign biopsy.

Discussion

Women with a history of a previous benign breast biopsy, especially those with proliferative findings, are known to be at increased risk for a later breast cancer. While almost 30% of breast cancers develop in women with prior BBD, there has been no comprehensive description of the breast cancers these women develop. With the risk prediction models available for this population, knowledge of the pathologic features and timeframe of expected breast cancers could enable fine tuning of the clinical management of these women.^{3, 8} Here we report on 1,273 women who had either excisional or core breast biopsy at the Mayo Clinic for palpable or mammographic concerns, with benign results, who later developed breast cancer.

To our knowledge, this is the first report detailing subsequent breast cancers that developed in a cohort of women with previous benign breast disease. We show that 81% of cancers are invasive, with 29% having spread to regional lymph nodes, comparable to incidence and nodal staging in the general US population.^13-15 The histologic subtypes of breast cancers do not vary by original benign pathology, nor does nodal status or tumor size. Although women with AH have a higher proportion of ER positive and grade 1 cancers compared to those with PDWA or NP findings, the majority of invasive cancers occurring after AH are intermediate or high grade.

Younger women requiring a benign breast biopsy have been shown to have an increased relative risk of later breast cancer.^{4, 5} Here we expand on that finding, showing that the breast cancers developing in women with BBD < age 45 have more aggressive features, including higher grade, and nodal positivity.

In previous work, Jacobs et al described the pathologic features of 169 invasive breast cancers that had developed in women with prior BBD. Similar to our findings, they reported no significant difference in the cancer features (including tumor size, nodal status, grade or histologic subtype) based on the histologic categories of the prior benign biopsy.¹⁶ McLaren et al reported on 48 invasive breast cancers that developed in a group of 252 women with previous atypical lobular hyperplasia. Almost half of the cancers (42%) were categorized as special or variant types, and they concluded that women with atypical lobular hyperplasia who developed breast cancer would be likely to have a good prognosis with low mortality, although mortality data were not provided.¹⁷ As described above for our cohort, a higher proportion of grade 1 cancers developed in the AH subgroup; however, these women were just as likely to develop node-positive invasive disease as those with other types of BBD. Moreover, we highlight that the high cumulative risk of breast cancer in women with AH, at slightly more than 1% per year, is a finding that has been appreciated only recently.¹⁸

Whether some of these benign lesions represent actual precursors, as opposed to indicators of generalized risk in breast tissue, remains the subject of ongoing research.^{9, 19} If a benign lesion is a direct precursor, subsequent cancers would be expected to occur in the same breast, in a shorter time interval, and perhaps with a higher proportion of carcinoma in situ. Interestingly, the proportion of ipsilateral cancers is higher in women with an original proliferative lesion, especially those with AH. Also, the median time to breast cancer is shortest in this group (8.8 years for AH vs 10.9 years for PDWA and 11.2 years for NP, p=0.05). Furthermore, the interval to DCIS appeared possibly shorter after a diagnosis of ADH (5.1 years) compared to ALH (9.2 years), although not statistically significant with small numbers of events in these subgroups. However, this adds evidence that ADH may be a precursor lesion to DCIS. Atypical hyperplasia represents a proliferation of dysplastic, monotonous epithelial cells, including clonal subpopulations, and thus is considered a premalignant condition.^{18, 20}

The high cumulative risk of breast cancer in women with AH, coupled with prospective, placebo-controlled data showing effective risk reduction with estrogen antagonists in this population, should compel stronger recommendations for prevention therapy in these women.^{18, 21} The findings of a higher likelihood of more aggressive breast cancers developing in younger women with BBD are, to our knowledge, unique and should direct more careful attention being paid to this group.

The strengths of this study are the long and thorough follow-up of this large cohort of women who had benign breast biopsies at the Mayo Clinic, and the detailed information on cancer phenotypes they subsequently developed. The benign biopsies were done to evaluate either palpable or mammographic concerns. Of note, women are not generally referred to the Mayo Clinic for a benign biopsy, so this population represents women receiving general medical care at our institution. We followed this population for breast cancer events, whether diagnosed at Mayo or elsewhere. Our breast pathologist (DWV) categorized the breast cancers using standardized criteria. Importantly, the size of the cohort provides suitable power to examine findings in relevant subgroups, such as women with atypical hyperplasia and those at younger age. A limitation is that this cohort is comprised of primarily non-Hispanic white women, consistent with our surrounding population demographic.

In summary, we provide details regarding the specific clinical and histologic features of breast cancers developing in women with a prior history of benign breast disease. Over 80% of cancers are invasive, with 29% having spread to regional lymph nodes, regardless of category of benign histology. Women who were younger at benign biopsy (< 45) developed breast cancers with some higher risk features than older women. Women with atypical hyperplasia have a higher cumulative risk of breast cancer (1% per year) than has been previously recognized, and the ER positive nature of these cancers should prompt more widespread utilization of prevention therapy in this group. Overall, these findings refute a simplistic “low grade – favorable biology” hypothesis of breast cancers that develop in women after benign biopsy.