The influence of family history on breast cancer risk in women with biopsy-confirmed benign breast disease

Results from the Nurses' Health Study

Authors

  • Laura C. Collins MD,

    Corresponding author
    1. Department of Pathology, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, Massachusetts
    • Department of Pathology, Beth Israel Deaconess Medical Center, 330 Brookline Avenue, Boston, MA 02215
    Search for more papers by this author
    • Fax: (617) 667 7120.

  • Heather J. Baer ScD,

    1. Channing Laboratory, Department of Medicine, Brigham and Women's Hospital and Harvard Medical School, Boston, Massachusetts
    2. Department of Epidemiology, Harvard School of Public Health, Boston, Massachusetts
    Search for more papers by this author
  • Rulla M. Tamimi ScD,

    1. Channing Laboratory, Department of Medicine, Brigham and Women's Hospital and Harvard Medical School, Boston, Massachusetts
    2. Department of Epidemiology, Harvard School of Public Health, Boston, Massachusetts
    Search for more papers by this author
  • James L. Connolly MD,

    1. Department of Pathology, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, Massachusetts
    Search for more papers by this author
  • Graham A. Colditz MD, DrPh,

    1. Channing Laboratory, Department of Medicine, Brigham and Women's Hospital and Harvard Medical School, Boston, Massachusetts
    2. Harvard Center for Cancer Prevention, Boston, Massachusetts
    Search for more papers by this author
  • Stuart J. Schnitt MD

    1. Department of Pathology, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, Massachusetts
    Search for more papers by this author

Abstract

BACKGROUND

An association between histologic category of benign breast disease (BBD) and breast cancer risk has been well documented. However, the influence of a positive family history (FH) on breast cancer risk among women with biopsy-confirmed BBD is less certain.

METHODS

The authors conducted a nested case–control study of BBD and breast cancer risk among 2005 women who were enrolled in the Nurses' Health Study. Cases were women with breast cancer who had a previous benign breast biopsy (n = 395 women). Controls were women who also had previous biopsy-confirmed BBD but were free from breast cancer at the time the corresponding case was diagnosed (n = 1610 women). BBD slides were reviewed and categorized as either nonproliferative lesions, proliferative lesions without atypia, or atypical hyperplasia (AH).

RESULTS

Compared with women who had nonproliferative lesions and no FH, women who had proliferative lesions without atypia and a positive FH had a higher breast cancer risk (odds ratio [OR], 2.45; 95% confidence interval [95% CI], 1.61–3.70) than women with no FH (OR, 1.51; 95% CI, 1.12–2.06; P = .07). Among women who had AH, the OR for the development of breast cancer was 4.38 (95% CI, 2.93–6.55) for those with no FH and 5.37 (95% CI, 3.01–9.58) for those with a positive FH (P = .57). There was no significant interaction between the type of BBD and FH (P = .74).

CONCLUSIONS

A positive FH of breast cancer slightly increased the breast cancer risk among women who had proliferative lesions without atypia. The increase in risk of breast cancer associated with FH was not significant among women who had AH. Cancer 2006. © 2006 American Cancer Society.

Previous clinical and epidemiologic studies of women with benign breast disease (BBD) have established a relation between histologic category of BBD and risk of subsequent breast cancer.1–13 The results of both retrospective cohort studies and case–control studies have indicated that women who have proliferative lesions without atypia in a benign breast biopsy have a 1.5-fold to 2-fold increase in breast cancer risk compared with women who have nonproliferative lesions, whereas the risk of breast cancer associated with the presence of atypical hyperplasia (AH) is 3-fold to 5-fold.1–15 Some of those studies indicated further that the risk associated with these categories of BBD may be influenced by other factors, including the time since the benign breast biopsy,6, 11, 13, 16 menopausal status,6, 11, 13 and histologic type of AH (i.e., atypical ductal hyperplasia vs. atypical lobular hyperplasia).11, 15

Family history also may interact with BBD in establishing the level of breast cancer risk. Dupont and Page reported in 1985 that a first-degree family history of breast cancer more than doubled the breast cancer risk in women who had AH, but it had little effect on the risk of subsequent breast cancer in women who had nonproliferative lesions or in women who had proliferative lesions without atypia.2, 17 Two subsequent studies that attempted to address the influence of family history on breast cancer risk in women with biopsy-confirmed BBD8, 9 were limited by small numbers of participants, resulting in un stable risk estimates for most subgroups. Most recently, in a large, retrospective cohort study of women with biopsy-confirmed BBD, Hartmann et al.13 observed no significant interaction between family history and histologic findings in benign breast biopsies in determining breast cancer risk. Thus, currently, the relation between family history, type of BBD, and breast cancer risk remains poorly defined.

BBD is a condition that affects large numbers of women and causes a substantial amount of anxiety because of its association with breast cancer. Nearly 20% of women who have annual screening mammograms over a 10-year period will undergo a biopsy,18 and many of those biopsies will show BBD. Given the large numbers of women who undergo screening mammography and the frequent occurrence of BBD among those women, a better understanding of the relation between BBD and family history clearly would be of value to assist in risk assessment and clinical management. Therefore, we conducted a nested case–control study among women in the Nurses' Health Study (NHS) and the Nurses' Health Study II (NHS II) who had a previous diagnosis of BBD to evaluate further the interaction between family history of breast cancer and type of BBD in determining breast cancer risk.

MATERIALS AND METHODS

Study Population

The NHS is an ongoing, prospective cohort study that began in 1976, when 121,700 female registered nurses between ages 30 years and 55 years completed a mailed, self-administered questionnaire regarding their health behaviors, lifestyle factors, and medical histories. Follow-up questionnaires have been sent to participants every 2 years to obtain updated information. The biennial questionnaires have assessed a variety of known and suspected risk factors for breast cancer, including history of BBD. On the 1976, 1978, and 1980 questionnaires, participants were asked whether they ever had been diagnosed with fibrocystic or other BBD and whether this diagnosis had required hospitalization; from 1982 onward, the questionnaires have inquired specifically regarding BBD confirmed by biopsy. Deaths are reported by family members and the postal service, and regular searches of the computerized National Death Index also are conducted.19 The NHS II is a separate cohort study that consists of 116,671 female registered nurses who were between ages 25 years and 42 years when the study began in 1989. The follow-up methods used in that cohort are very similar to those used in the original NHS cohort.20 On each biennial questionnaire, participants have been asked whether they ever had been diagnosed with BBD and, if so, then whether the diagnosis was confirmed by biopsy. In general, the response rate to each questionnaire has been very similar among women who reported a previous diagnosis of BBD than among those who did not. For example, 85% of women who had been diagnosed with BBD as of 1986 completed the 1996 questionnaire, compared with 82% of those who did not report a diagnosis of BBD.

Design of Nested Case–Control Study

We conducted a nested case–control study of BBD and breast cancer risk among participants in the NHS and NHS II who had reported a previous diagnosis of BBD that either required hospitalization or was confirmed by biopsy. Within this subcohort, eligible cases were women who reported a first diagnosis of breast cancer between 1976 and return of the 1996 questionnaire (NHS) or between 1989 and return of the 1995 questionnaire (NHS II). Self-reported breast cancers were confirmed by a review of medical records, and both invasive breast cancer and carcinoma in situ were included in the study. Eligible controls were women who were free from breast cancer at the time the matching case was diagnosed and who also had a previous diagnosis of biopsy-confirmed BBD; they were matched to cases on exact year of birth and exact year of diagnosis of BBD. We attempted to identify 4 matched controls for each case, but this was not always possible for logistical reasons. The study was approved by the Human Research Committee of Brigham and Women's Hospital (Boston, MA).

Collection and Review of Benign Breast Biopsy Specimens

Cases and controls were contacted for permission to obtain their BBD pathology records and biopsy specimens. Greater than 70% of the 1310 cases and 5273 controls originally identified for the study confirmed the diagnosis and granted permission, and histologic specimens subsequently were obtained for 52% of those who granted permission (465 cases and 1939 controls). The primary reason given by hospital pathology departments for not sending specimens was that they had been destroyed or were no longer available (35%). Approximately 98% of pathology specimens that were obtained were considered to be of good quality and were evaluated by the study pathologists (431 cases and 1869 controls). An additional 262 participants were excluded, because they reported bilateral BBD, but the specimen did not identify the breast side (n = 2 women), the specimen did not contain breast tissue (n = 76 women), the breast cancer diagnosis was prior to the beginning of the study or after the 1996 or 1995 follow-up cycle (n = 10 women), or the date of BBD diagnosis was within 6 months of the date of the breast cancer diagnosis or index date for controls (n = 174 women).

Biopsy slides were reviewed independently by 1 of 2 pathologists (S.J.S. or J.L.C.) who were blinded to participants' case or control status. The pathologists completed a detailed worksheet with information on the morphologic features of each specimen; then, lesions were classified into 1 of 3 main histologic categories of BBD, according to the criteria developed by Page et al.,15 as nonproliferative, proliferative without atypia, or AH. All biopsies, including bilateral biopsies, were classified according to the most severe changes present, and specimens with possible or definite AH were reviewed jointly by both pathologists. During the pathology review, it was determined that 33 women who had an original diagnosis of BBD had evidence of carcinoma in situ (n = 27 women) or invasive carcinoma (n = 6 women). These women were excluded from further analyses, leaving 395 breast cancer cases and 1610 controls with available benign pathology specimens for the current study.

Assessment of Family History of Breast Cancer

On the baseline questionnaire in 1976, NHS participants were asked whether their mother or any of their sisters had ever been diagnosed with breast cancer, and this information was updated again in 1982, 1988, and 1992. Similar information was collected from NHS II participants on the baseline questionnaire in 1989. Women were classified as having a positive family history of breast cancer in a first-degree relative if they reported a diagnosis of breast cancer in either a mother or sister. To maintain the prospective nature of the study and to minimize bias in recall or reporting, women were assigned to a family history category based on information reported prior to the cycle of the breast cancer diagnosis (cases) or the index date (controls). The proportion with a positive family history of breast cancer did not differ significantly between women whose pathology specimens were obtained (18.4%) and women whose specimens were not available (17.0%; P = .16).

Statistical Analysis

We examined the distribution of breast cancer risk factors among controls according to histologic category of BBD. Unconditional logistic regression was used to estimate odds ratios (ORs) and 95% confidence intervals (95% CIs) for breast cancer risk according to histologic category of BBD, using nonproliferative disease as the reference group. Unconditional logistic regression was used rather than conditional logistic regression, because this analytic approach allowed us to include all women in the case and control groups for whom we had histologic information. First, we adjusted only for the matching factors, which were age at breast cancer diagnosis or index date (younger than 45 years, 45–49 years, 50–54 years, 55–59 years, and 60 years or older) and year of benign breast biopsy (before 1970, 1970–1979, 1980–1989, and 1990 or later), and follow-up (years from BBD diagnosis to breast cancer diagnosis or index date). Then, adjustments were made for potential confounding factors, classifying women according to their responses for the questionnaire cycle prior to the breast cancer diagnosis or index date. In addition to the matching factors and length of follow-up, the following factors were included in the multivariate models: first-degree family history of breast cancer (yes or no), age at menarche (ages younger than 12 years, 12 years, 13 years, or 14 years and older), parity/age at first birth (nulliparous, 1 or 2 children with age at first birth younger than 25 years, 1 or 2 children with age at first birth from 25 to 29 years, 1 or 2 children with age at first birth 30 years or older, ≥3 children with age at first birth younger than 25 years, or ≥3 children with age at first birth 25 years or older), body mass index (<21 kg/m2, 21–22.9 kg/m2, 23–24.9 kg/m2, 25–29.9 kg/m2, or ≥30 kg/m2), menopausal status/type of menopause (premenopausal, natural menopause, bilateral oophorectomy, or other or unknown type of menopause), and status/duration of postmenopausal hormone use (never, past, current <5 years, or current ≥5 years). Indicator variables were created to represent categories of the matching factors and other covariates, and the length of follow-up was modeled as a continuous variable.

We examined the combined effect of first-degree family history of breast cancer and histologic category of BBD by classifying participants jointly according to both variables and computing ORs and 95% CIs for different subgroups, and we used women with nonproliferative disease and no family history of breast cancer as the reference group for these comparisons. In addition, we computed ORs for breast cancer risk according to family history of breast cancer separately among women with nonproliferative lesions, proliferative lesions without atypia, and AH, and we used women with no family history as the reference group within each BBD category. Finally, we assessed whether there was a significant interaction between family history and histologic category of BBD on the multiplicative scale by conducting likelihood-ratio tests to compare models that included only the main effects of family history and type of BBD with models that included the jointly classified variable. All statistical analyses were performed using the SAS software package (version 8.2; SAS Institute, Cary, NC).

RESULTS

The mean age at the time of benign breast biopsy was 43.5 years for cases (range, 17–68 years) and 43.8 years for controls (range, 15–71 years), and the median time since benign biopsy was 9.1 years for cases and 8.0 years for controls. Descriptive characteristics of controls according to histologic category of BBD are presented in Table 1. Women who had AH were older at the time of the benign biopsy, had a later year and less time since the benign biopsy, were less likely to be premenopausal at benign biopsy, were more likely to have a first-degree family history of breast cancer, and had slightly greater alcohol consumption compared with women who had nonproliferative lesions or proliferative lesions without atypia. The women with AH also were slightly more likely to have undergone a bilateral oophorectomy and to have used postmenopausal hormones for ≥5 years.

Table 1. Characteristics of the Control Group (n = 1610 Women) According to Histologic Subtype of Benign Breast Disease among Participants in the Nurses' Health Study (1976–1996) and Nurses' Health Study II (1989–1995)
CharacteristicHistologic subtype of Benign breast disease
NonproliferativeProliferative without atypiaAtypical hyperplasia
  • *

    Median values.

  • Among postmenopausal women.

  • Among parous women.

No. of controls (%)611 (38.0)839 (52.1)160 (9.9)
Mean values
 Age at benign biopsy, y42.543.749.4
 Year of benign biopsy*197919801982
 Time since benign biopsy, y*8.68.26.2
 Age at menarche, y12.612.612.8
 Age at menopause, y47.947.447.7
 Body mass index, kg/m225.024.624.7
 Alcohol intake, g/day4.85.46.5
 Parity2.93.03.0
 Age at first birth, y24.924.825.2
Frequencies (%)
 First-degree family history of breast cancer17.417.422.5
 Premenopausal at benign biopsy72.371.456.3
 Nulliparous6.48.66.3
 Natural menopause56.359.655.6
 Bilateral oophorectomy21.420.124.1
 Current postmenopausal hormone use ≥5 y16.820.322.2

Compared with women who had nonproliferative lesions, the multivariate OR for the development of breast cancer (either invasive or in situ) among women who had proliferative lesions without atypia was 1.52 (95% CI, 1.17–1.99), and the OR for women who had AH was 4.11 (95% CI, 2.90–5.83) (Table 2). These ORs were very similar when only the development of invasive cancer was used as an endpoint; therefore, in the case group, both women with invasive carcinoma and women with in situ breast cancer were included in all subsequent analyses.

Table 2. Odds Ratios and 95% Confidence Intervals for Breast Cancer Risk According to Histologic Subtype of Benign Breast Disease among Participants in the Nurses' Health Study (1976–1996) and Nurses' Health Study II (1989–1995)
Histologic subtype of Benign breast diseaseControls(No.)All breast cancers (n = 395 Women)Invasive breast cancers (n = 316 Women)
Cases(No.)OR(95% CI)*Adjusted OR (95% CI)Cases (No.)OR(95% CI)*Adjusted OR (95% CI)
  • OR indicates odds ratio; 95% CI, 95% confidence interval; Ref, reference group.

  • *

    Adjusted for age at breast cancer diagnosis, year of benign breast biopsy, and years of follow-up.

  • Adjusted for the same factors mentioned above and for family history of breast cancer, age at menarche, parity/age at first birth, body mass index, menopausal status/type of menopause, and recency/duration of postmenopausal hormone use.

Nonproliferative611991.0 (Ref)1.0 (Ref)771.0 (Ref)1.0 (Ref)
Proliferative without atypia8392001.50 (1.15–1.95)1.52 (1.17–1.99)1641.58 (1.18–2.12)1.62 (1.21–2.18)
Atypical hyperplasia160964.09 (2.90–5.76)4.11 (2.90–5.83)754.04 (2.78–5.87)4.04 (2.76–5.92)

Cases were more likely than controls to have a first-degree family history of breast cancer (25.1% of cases vs. 17.9% of controls). Among all participants, the multivariate OR for the development of breast cancer among women who had a positive family history was 1.51 (95% CI, 1.15–1.99) compared with women who had no family history, also adjusting for histologic category of BBD. When participants were classified jointly according to both family history and category of BBD, women who had a first-degree family history of breast cancer were at somewhat greater risk compared with women who had no family history for each type of BBD (Table 3). Compared with women who had nonproliferative lesions and no family history of breast cancer, the multivariate OR for women who had proliferative lesions without atypia and no family history was 1.51 (95% CI, 1.12–2.06), and the multivariate OR for women who had proliferative lesions without atypia and a first-degree family history of breast cancer was 2.45 (95% CI, 1.61–3.70); the difference between these estimates approached statistical significance (P = .07). The multivariate OR for women who had AH and no family history of breast cancer was 4.38 (95% CI, 2.93–6.55), and the multivariate OR for women who had AH and a first-degree family history of breast cancer was 5.37 (95% CI, 3.01–9.58) compared with women who had nonproliferative lesions and no family history. These estimates did not differ significantly from one another (P = .57), although this could be due in part to the small number of cases with AH.

Table 3. Odds Ratios and 95% Confidence Intervals for Breast Cancer Risk According to Jointly Classified Histologic Subtype of Benign Breast Disease and First-Degree Family History of Breast Cancer among Participants in the Nurses' Health Study (1976–1996) and Nurses' Health Study II (1989–1995)
Histologic subtype of Benign breast diseaseCases (No.)Controls (No.)OR (95% CI)*Adjusted OR (95% CI)
  • OR indicates odds ratio; 95% CI, 95% confidence interval; Ref, reference group.

  • *

    Adjusted for age at breast cancer diagnosis, year of benign breast biopsy, and years of follow-up.

  • Adjusted for the same factors mentioned above and for age at menarche, parity/age at first birth, body mass index, menopausal status/type of menopause, and recency/duration of postmenopausal hormone use.

Nonproliferative
 No family history755051.0 (Ref)1.0 (Ref)
 Family history241061.56 (0.94–2.59)1.57 (0.94–2.62)
Proliferative without atypia
 No family history1516931.50 (1.11–2.03)1.51 (1.12–2.06)
 Family history491462.34 (1.56–3.52)2.45 (1.61–3.70)
Atypical hyperplasia
 No family history701244.31 (2.91–6.40)4.38 (2.93–6.55)
 Family history26365.19 (2.94–9.18)5.37 (3.01–9.58)

When we stratified by type of BBD and examined the association between family history and risk of breast cancer separately within each category, the increase in risk associated with a first-degree family history of breast cancer appeared to be greater among women who had nonproliferative lesions and proliferative lesions without atypia than among women who had AH (Table 4). Among women who had proliferative lesions without atypia, the multivariate OR for women who had a first-degree family history of breast cancer was 1.56 (95% CI, 1.06–2.30) compared with women who had no family history. Among women who had AH, the multivariate OR was 1.20 (95% CI, 0.61–2.36) for women who had a family history compared with women who had no family history. However, the interaction between type of BBD and family history of breast cancer on the multiplicative scale was nonsignificant according to a likelihood-ratio test (P = .74).

Table 4. Odds Ratios and 95% Confidence Intervals for Breast Cancer Risk According to First-Degree Family History of Breast Cancer for Different Histologic Subtypes of Benign Breast Disease among Participants in the Nurses' Health Study (1976–1996) and Nurses' Health Study II (1989–1995)
Histologic subtype of Benign breast diseaseCases (No.)Controls (No.)OR (95% CI)*Adjusted OR(95% CI)
  • OR indicates odds ratio; 95% CI, 95% confidence interval; Ref, reference group.

  • *

    Adjusted for age at breast cancer diagnosis, year of benign breast biopsy, and years of follow-up.

  • Adjusted for the same factors mentioned above and for age at menarche, parity/age at first birth, body mass index, menopausal status/type of menopause, and recency/duration of postmenopausal hormone use.

Nonproliferative
 No family history755051.0 (Ref)1.0 (Ref)
 Family history241061.63 (0.97–2.73)1.65 (0.96–2.84)
Proliferative without atypia
 No family history1516931.0 (Ref)1.0 (Ref)
 Family history491461.51 (1.04–2.20)1.56 (1.06–2.30)
Atypical hyperplasia
 No family history701241.0 (Ref)1.0 (Ref)
 Family history26361.17 (0.63–2.18)1.20 (0.61–2.36)

DISCUSSION

There is general agreement that a family history of breast cancer in a first-degree relative is associated with an increased risk of breast cancer, even after accounting for the effects of mutations in known breast cancer-susceptibility genes.21 There also is consensus, that among women who have had a benign breast biopsy, the subsequent breast cancer risk varies according to the histologic category of BBD: it is increased moderately in women who have proliferative lesions without atypia and is increased substantially among women who have AH.1–15 Despite potentially important clinical implications because of the large number of women diagnosed with BBD, the interactions between family history and histologic category of BBD in defining the magnitude of breast cancer risk are less well understood.

We observed that among, all women who had biopsy-confirmed BBD, a family history increased the risk of breast cancer by approximately 50%. Although some increase in risk was apparent for all histologic categories of BBD, there was no statistically significant interaction between family history of breast cancer and histologic category of BBD.

Four prior studies evaluated the possible interaction between family history of breast cancer, histologic category of BBD, and breast cancer risk.2, 8, 9, 13 In the first and most widely cited of those studies, Dupont and Page2 observed that, compared with women who had nonproliferative lesions and no family history of breast cancer, the relative risk (RR) of breast cancer was 1.9 (95% CI, 1.2–3.0) for women who had proliferative lesions without atypia and no family history and 2.7 (95% CI, 1.4–5.3) for women who had proliferative lesions without atypia and a positive family history, results similar to those observed in the current study. The RR for women who had AH without a family history, compared with women who had nonproliferative lesions and no family history (RR, 4.3; 95% CI, 2.4–7.8) in the study by Dupont and Page2 also was similar to that observed in the current study (OR, 4.2; 95% CI, 2.8–6.2) (Table 3). However, in contrast to our study, in which a positive family history had little effect on the risk of breast cancer among women who had AH, Dupont and Page reported that a positive family history substantially increased the breast cancer risk among women who had AH (RR, 11.0; 95% CI, 5.5–24.0) and that there was a significant interaction between type of BBD and family history.2

McDivitt et al.8 noted that, for women who had proliferative lesions without atypia, the RR for subsequent breast cancer was 1.8 (95% CI, 1.2–2.6) for women who had no family history and 2.3 (95% CI, 0.8–6.2) for women who had a family history, compared with women who had no BBD. Those authors also observed that, for women who had AH, the RR of breast cancer was 3.0 (95% CI, 1.7–5.3) for women who had no family history and 1.8 (95% CI, 0.3–9.3) for women who had a family history. The results of that study suggested that a positive family history does not influence the risk of breast cancer either among women who have proliferative lesions without atypia or among women who have AH. However, those results are difficult to compare directly with the results from our study, because McDivitt et al. used women without BBD as the referent group8, whereas for our referent group we used women who had nonproliferative lesions and no family history of breast cancer. In addition, the number of women in some of the subgroups in their study was small, particularly in the subgroup of women who had AH and a positive family history, in which there were only 6 cases. In the third study, Dupont et al.9 observed that, among women who had proliferative lesions without atypia, the RR for breast cancer for women who had no family history was 1.7 (95% CI, 0.9–3.2) and, for women who had a family history, the RR was 2.6 (95% CI, 1.0–6.4). Those authors also noted that, among women who had AH, the RR was 4.2 (95% CI, 1.4–12) for women who had no family history and 22.0 (95% CI, 2.4–203.0) for women who had a family history. Although the referent group for their analysis was comparable to the referent group used in our study (i.e., women who had nonproliferative lesions without a family history), like in the study by McDivitt et al.,8 the number of women in some of the subgroups in the study by Dupont et al. was small, particularly in the subgroup of women who had AH and a positive family history, in which there were only 5 cases. Most recently, in a retrospective cohort study of >9000 women with biopsy-confirmed BBD, Hartmann et al.13 did not observe a significant interaction between family history and histologic findings from benign breast biopsies in determining breast cancer risk.

It should be emphasized that all 4 of the studies described above that addressed the possible interaction among family history, BBD, and breast cancer risk were characterized by a relatively small number of women who had the combination of AH and a positive family history,2, 8, 9, 13 and the same was true for the current study. Although there were only 26 cases and 36 controls in our study who had both a positive family history of breast cancer and AH, there were only 39 women who had the combination of a positive family history and AH in the widely cited study by Dupont and Page, in which a large, additive effect of family history to risk was reported among women who had AH.2 Given the fairly broad 95% CIs of the breast cancer risk estimates for women with AH and a positive family history in all of those studies, in fact, the results may not be inconsistent with each other, and any apparent differences simply may reflect the manner in which the results of those studies have been presented and interpreted.

Several other potential limitations of our current study should be noted. First, we were unable to obtain pathology material for review from a proportion of eligible women in the case and control groups who had given permission. However, the percentages with a positive family history of breast cancer were similar for participants with and without available specimens, and the primary reason for not being able to obtain specimens was the routine disposal of biopsy material by the hospitals; therefore, it is unlikely that this introduced a selection bias. In addition, we did not examine the possible role of strength of family history of breast cancer because of the small numbers in the subset of women who had a first-degree relative diagnosed with breast cancer before age 50 years (10 cases and 34 controls) and the subset of women who had ≥2 first-degree relatives with breast cancer (4 cases and 14 controls) for whom information was available. Furthermore, we did not collect information on family history other than for first-degree relatives. However, in the recent study by Hartmann et al., strength of family history did not influence the breast cancer risk among women with AH.13

In conclusion, we observed that a positive family history of breast cancer increased the risk of breast cancer slightly among women who had proliferative lesions without atypia. However, in this study, the increase in risk associated with family history was not significant among women who had AH, and these factors did not modify the effects of each other. Thus, the current results, in conjunction with recently published results from the Mayo Clinic,13 raise questions concerning the certainty of the prevailing view that a positive family history more than doubles the breast cancer risk associated with AH,2 and we believe that the influence of family history on the risk of breast cancer in women who have AH remains an unresolved issue. Therefore, risk assessment and management recommendations for these women should be based on the presence of AH and do not necessarily need to be altered in women who also have a positive family history of breast cancer.

Ancillary