SEARCH

SEARCH BY CITATION

Keywords:

  • atypical hyperplasia;
  • benign breast disease;
  • breast cancer;
  • cohort study;
  • family history

Abstract

  1. Top of page
  2. Abstract
  3. MATERIAL AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

A major risk factor for breast cancer is having a first-degree family history of the disease. Benign breast disease (BBD), particularly atypical hyperplasia, is also associated with an increased risk of breast cancer. However, the relationship between family history of breast cancer and BBD is unclear. From 1989 through 1997, 80,995 participants in the Nurses' Health Study II were followed; 16,849 reported a first diagnosis of BBD. Pathology slides were reviewed for 1,465 women who reported having a tissue biopsy, and these were classified as nonproliferative BBD, proliferative BBD without atypia or atypical hyperplasia. Women with a family history of breast cancer were more likely to report a physician diagnosis of BBD [rate ratio (RR) = 1.38, 95% confidence interval (CI) 1.29–1.46]. The magnitude of this association declined with age from RR = 1.96 (95% CI 1.55–2.47) at 25–29 years to RR = 1.20 (95% CI 0.95–1.52) at age 45–50 years. Among women with proliferative disease, those with a family history of breast cancer were almost 3 times as likely to have atypia (prevalence odds ratio = 2.72, 95% CI 1.23–5.89) than those with no family history. In conclusion, women with a family history of breast cancer appear to be at increased risk of being diagnosed with BBD, in particular the high-risk types of BBD associated with a greatly increased risk of breast cancer. This link adds weight to the belief that BBD with atypia is a precursor or marker lesion for breast cancer. © 2002 Wiley-Liss, Inc.

Women with a history of benign breast disease (BBD) are at increased risk of developing breast cancer.1 Although nonproliferative disease does not appear to be associated with increased risk,2, 3 proliferative disease without atypia and proliferative disease with atypia have been associated with approximately a 1.5-fold and a 4-fold increased risk of cancer, respectively.3, 4, 5 This risk appears to be approximately equal for cancer developing in the same breast as the BBD or in the opposite breast6, 7, 8 and it is unclear whether the BBD is a precursor lesion or simply a marker for a generalized increased risk of cancer. Better understanding of the risk factors for BBD, especially the relation between BBD and known risk factors for breast cancer, will help to clarify the relation between BBD and breast cancer and increase understanding of the carcinogenic process.

A major risk factor for breast cancer is having a history of breast cancer in a first-degree relative,9 though this effect may diminish with increasing age.10, 11 The relation between breast cancer family history and BBD is less clear. Some studies have reported no association,12, 13 while others have reported an increased risk of BBD among women with a family history of breast cancer.14, 15, 16 These studies have included relatively small numbers of women with biopsy-confirmed BBD and have employed different definitions of BBD. Some have compared women with biopsy-confirmed BBD to unscreened hospital or population controls13, 14, 16 and it is possible that women with a family history of breast cancer are more likely to consult a physician and/or undergo biopsy and thus be diagnosed with BBD, falsely elevating measures of association. Conversely, some studies have included women with a prior history of BBD in the control group,12, 14 potentially biasing estimates of association toward the null.

Our present aim was to evaluate the association between family history of breast cancer and incidence of BBD in a cohort of 80,995 U.S. nurses and to evaluate variation in this association by the histologic type of BBD and age.

MATERIAL AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIAL AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

The Nurses' Health Study II is a prospective cohort study that began in 1989 when 116,671 female nurses aged 25–44 years completed a mailed questionnaire. Subsequent questionnaires have been mailed to the women every 2 years, and follow-up of this cohort in each 2-year interval is >90%. The protocol for the study was approved by the Harvard School of Public Health (Boston, MA).

For the present study, all women who reported a prior physician diagnosis of fibrocystic disease or other BBD (n = 33,375) or cancer (n = 691) in 1989 were excluded, as were an additional 1,610 women (1.4% of the original cohort) with no follow-up information. This left a baseline population of 80,995 women.

Self-reported BBD

In 1991, 1993, 1995 and 1997, the women were asked if, since the previous questionnaire, a physician had told them they had fibrocystic disease or other BBD and, if so, if this had been confirmed by biopsy. Two definitions of self-reported BBD were considered: firstly, all new reports of BBD and, secondly, a subgroup of cases reported as confirmed by biopsy. The date of diagnosis of BBD was estimated as the mid-point between the date of return of the first questionnaire reporting BBD and the return date of the most recent questionnaire prior to that. Fifty-one women reported BBD and breast cancer on the same questionnaire and were excluded from the case groups and further follow-up.

Histologically Confirmed BBD

Women who in 1993, 1995 or 1997 reported a diagnosis of biopsy-confirmed BBD were contacted to seek confirmation of the diagnosis and permission to obtain pathology records. (Reports prior to 1993 were not followed up). Of these 2,198 women, 1,761 (80%) confirmed the diagnosis and granted permission for review of the pathology slides and records from the biopsy. The main reasons for nonresponse were that the woman could not be contacted (6%), did not confirm the biopsy (6%) or did not give permission for her slides to be reviewed (7%). Pathology material was obtained and reviewed for 1,577 women. This represents 92% of those who had given permission and approximately 77% of those who actually had a biopsy (after excluding women who originally reported a biopsy but subsequently denied having one). Of these, 1,465 (93%) were confirmed to be eligible cases and a valid diagnosis was obtained. The main reasons for exclusion were that the pathology specimen did not contain breast tissue (n = 41) or that the biopsy date was before the start of follow-up (n = 38). In addition, women were excluded if their biopsy date was after the date they reported BBD (n = 16) or if they had a prior cancer or BBD (n = 9), a diagnosis of breast cancer within 1 year of diagnosis of BBD (n = 3) or a diagnosis of carcinoma in situ (n = 5). Women with a confirmed histologic diagnosis of BBD did not differ from women for whom valid pathology material was not available with respect to family history of breast cancer (7.6% vs. 8.4%, p = 0.4).

Benign breast lesions were classified by 1 of 4 pathologists (S.J.S., J.L.C., T.J. or G.P.) as normal or nonproliferative, proliferative without atypia or atypical hyperplasia. This classification was performed after standard training and according to standard criteria.17 A proportion of biopsies were jointly reviewed by 2 pathologists to ensure consistency, and any biopsies that showed atypia or questionable atypia were re-reviewed by one or more of the other pathologists so that a consensus diagnosis could be reached. Biopsy tissue with intraductal papilloma, radial scar, sclerosing adenosis, fibroadenoma, fibroadenomatous change or moderate to florid ductal hyperplasia in the absence of atypical hyperplasia was classified as proliferative without atypia.

Statistical Analysis

Each participant contributed person-time from when she returned the baseline questionnaire in 1989 until the return of the 1997 questionnaire. Women who did not return the 1997 questionnaire were censored prior to this at the earliest of the following: death, report of BBD or cancer other than nonmelanoma skin cancer or the date they last returned a questionnaire. A woman was classified as having a family history of breast cancer if she reported breast cancer in any first-degree relative on the baseline questionnaire. Analyses were stratified by age (in 5-year categories) and other potential risk factors to control for confounding using Mantel-Haenszel methods.

Logistic regression was used to conduct case–case comparisons among the women with histologically confirmed BBD. We calculated prevalence odds ratios (PORs) and 95% confidence intervals (CIs) to compare women with proliferative BBD with or without atypia and women with nonproliferative BBD. For these analyses, covariate information was taken from the most recent questionnaire prior to the date of the biopsy.

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIAL AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

A total of 16,849 women self-reported a physician diagnosis of BBD during 545,596 person-years (py) of follow-up, and of these, 3,165 (18.8%) reported that the diagnosis was confirmed by tissue biopsy. The overall incidence of self-reported BBD was 30.9/1,000 py; however, this varied significantly with age, increasing from 22.6/1,000 py at age 25–29 years to 35.6/1,000 py at age 40–44 years, when the incidence rate leveled out (Table I). Older women were significantly more likely to report that their BBD was confirmed by biopsy (13%, 16%, 19%, 21% and 25% at ages 25–29, 30–34, 35–39, 40–44 and 45–50 years, respectively; p for trend < 0.001). Thus, the incidence of biopsy-confirmed BBD increased almost 3-fold over the age range from 2.9/1,000 py at age 25–29 years to 7.9/1,000 py at age 45–50 years.

Table I. RR And 95% CI For the Association between Incidence of Self-Reported BBD, Age and Family History of Breast Cancer (follow-up 1989–1997)
 Person-YearsAny BBDBBD confirmed by biopsy
CasesRate per 1,000 pyRR95% CICasesRate per 1,000 pyRR95% CI
  • 1

    RR adjusted for age group; rates shown are crude rates.

  • 2

    Test for linear trend across groups.

Age (years)         
 25–2950,9241,15222.60.670.63–0.721452.90.450.38–0.54
 30–34147,9443,84126.00.770.74–0.806314.30.680.62–0.75
 35–39183,7726,17333.61.0Reference1,1566.31.0Reference
 40–44130,3004,63635.61.061.02–1.109757.51.191.09–1.30
 45–5032,6561,04732.10.950.89–1.022587.91.261.10–1.44
     p < 0.0012   p < 0.0012
Family history1         
 No519,60715,74230.31.0Reference2,9115.61.0Reference
 Yes25,9891,10742.61.381.29–1.462549.81.671.47–1.90
Total545,59616,84930.9  3,1655.8  

Women with a family history of breast cancer in a first-degree relative were significantly more likely to report any diagnosis of BBD or a diagnosis of biopsy-confirmed BBD than women with no family history (Table I). The association was somewhat stronger if the relative was diagnosed before age 50 years (rate ratio [RR] = 1.81, 95% CI 1.51–2.16 vs. RR = 1.56, 95% CI 1.31–1.86 for relatives diagnosed after the age of 50).

The effect of family history on risk of BBD decreased with increasing age. The RR for any diagnosis of BBD fell from 1.96 (95% CI 1.55–2.47) at age 25–29 years to 1.60 (95% CI 1.41–1.82) at age 30–34 years, 1.36 (95% CI 1.23–1.50) at age 35–39 years, 1.21 (95% CI 1.08–1.36) at age 40–44 years and 1.20 (95% CI 0.95–1.52) at age 45–50 years. A similar pattern was observed for the association with biopsy-confirmed BBD, which fell from 2.08 (95% CI 1.09–3.96) at age 25–29 years to 1.31 (95% CI 0.83–2.06) at age 45–50 years.

These results were unchanged when the analysis was restricted to women who reported having a mammogram or breast examination by a physician for screening in the previous 2 years. Similarly, adjustment for potential confounders, including body mass index, smoking history, age at menarche, age at first birth, parity and oral contraceptive use, did not alter the results.

Figure 1 shows age-incidence rate curves separately for women with and without a family history of breast cancer. Among women with a family history of breast cancer, the incidence rate of self-reported BBD was high at all ages. In contrast, among women with no family history, the incidence rate increased with age, approaching that for women with a family history by age 45–50 years. The pattern was somewhat similar for biopsy-confirmed BBD, with the relative difference again falling at the older ages.

thumbnail image

Figure 1. Incidence rates (and 95% CIs) of self-reported BBD by age and family history of breast cancer. Solid lines represent women with a family history of breast cancer; broken lines represent women without a family history. The upper 2 lines represent any self-reported BBD; the lower 2 lines represent self-reported BBD confirmed by tissue biopsy.

Download figure to PowerPoint

Of the 1,465 women with a postreview diagnosis of BBD, 532 were classified as having nonproliferative BBD, 863 as having proliferative BBD without atypia and 70 as having atypical hyperplasia. Women with proliferative BBD, in particular proliferative BBD with atypia, were significantly older than those with nonproliferative disease (Table II). Women with atypia were more likely to have a family history of breast cancer than those with nonproliferative disease, though family history was less common among women with proliferative BBD without atypia. Atypia was significantly associated with family history in comparison to proliferative disease without atypia (age-adjusted POR = 2.76, 95% CI 1.33–5.74) or to any histologically confirmed BBD without atypia (POR = 2.16, 95% CI 1.05–4.35). Adjustment for potential confounders did not alter these results. There were too few women with atypical hyperplasia to evaluate variation in this association with age at diagnosis.

Table II. POR And 95% CI For the Association between Histologically Confirmed BBD, Age and Family History of Breast Cancer (follow-up 1991–1997)
 Nonproliferative BBD1 NumberProliferative BBD, no atypiaAtypical hyperplasia
NumberPOR(95% CI)NumberPOR95% CI
  • 1

    Reference group.

  • 2

    Test for linear trend across groups.

  • 3

    POR adjusted for age group.

Age (years)       
 25–34971180.770.56–1.0770.770.31–1.92
 35–391812851.0Reference171.0Reference
 40–441873051.040.80–1.34301.710.91–3.20
 45–50671551.471.04–2.07162.541.22–5.32
   p = 0.0032  p = 0.0022
Family history3       
 No4808141.0 601.0 
 Yes52490.560.37–0.84101.590.76–3.33
Total532863  70  

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIAL AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

In this population, women with a history of breast cancer in a first-degree relative were significantly more likely to report a diagnosis of BBD than those with no family history. This difference was greatest for younger women and almost disappeared by age 45–50 years. It is possible that this association simply resulted from an increased tendency for women who have a relative with breast cancer to seek medical advice for a breast lump. The results were, however, unchanged when the analysis was restricted to women who had undergone some form of breast screening and who therefore had similar opportunity for diagnosis of BBD, suggesting that this association might be real. It is, however, possible that the even stronger association with biopsy-confirmed BBD is due, at least in part, to an increased tendency for a physician to order a tissue biopsy in a woman with a family history of breast cancer.

Women with atypical hyperplasia were significantly more likely to have a family history of breast cancer than either women with proliferative disease without atypia or a combined group of all women with biopsy-confirmed BBD without atypia. The apparently low prevalence of family history among women with proliferative compared to nonproliferative disease was surprising though consistent with data from a case-control study nested within a cohort of women attending a breast-screening program in Canada (OR = 0.59, 95% CI 0.24–1.48).18 This could be an artifact if physicians are more likely to order a biopsy for women with a family history just to “be on the safe side”, thus leading to a higher proportion of normal/nonproliferative diagnoses in this group. It is, however, hard to imagine how selection bias could lead to the selective inclusion of women with a family history if they also had atypia relative to those with proliferative disease without atypia.

The present study included more than twice as many women with histologically confirmed BBD as previous studies that have evaluated the association between breast cancer family history and type of BBD. In one study, 173 women with histologically confirmed BBD (121 proliferative, 52 atypical hyperplasia) were compared to 403 control women who had sought evaluation for breast symptoms but had a diagnosis of nonproliferative BBD or did not undergo biopsy.19 There was no association with proliferative BBD without atypia (OR = 0.9, 95% CI 0.5–1.6), but women with a family history of breast cancer had a slightly increased risk of atypia (OR = 1.5, 95% CI 0.7–3.0) and breast cancer (OR = 1.8, 95% CI 1.1–2.9). In that study, women with a prior history of BBD were not excluded from the control group, thereby potentially biasing the OR toward the null. A second study, conducted among participants in a breast-screening program in Japan,15 compared 382 women with histologically confirmed BBD (130 proliferative, 252 nonproliferative) to 1,489 women who did not undergo biopsy. A family history of breast cancer was associated with an increased risk of both nonproliferative (OR = 1.95, 95% CI 0.96–3.94) and proliferative BBD (OR = 4.13, 95% CI 1.46–11.71). Only 17 women had atypical hyperplasia, and the authors did not look at this group separately. Although the results of these studies are not totally consistent, in each, women with a family history of breast cancer were at greater risk of developing the high-risk types of BBD.

In the present study, family history was ascertained at baseline, when the women were relatively young (age 25–45 years). As a result, some women who were classified as not having a family history in 1989, will since have had a first-degree relative diagnosed with breast cancer. The likely effect of such misclassification would be to underestimate the true association between family history and BBD. However, the first-degree relatives, especially the sisters, of young women will also be relatively young and cancers diagnosed in this group may be more likely to be genetic than sporadic. This could potentially account for the stronger association between family history and diagnosis of BBD seen among younger women. Inclusion of additional information collected in 1997, when the women were next asked about their family history of cancer, did not alter the results.

These data support the hypothesis that women with a family history of breast cancer are at increased risk of BBD and particularly of the high-risk types of BBD (proliferative changes with atypia), which are associated with a greatly increased risk of breast cancer. They also add support to the hypothesis that the adverse effect of a positive family history is greatest for younger women and that the effect diminishes with increasing age. This link between a positive family history of breast cancer and development of both atypical hyperplasia and breast cancer together with the known association between atypical hyperplasia and the subsequent development of breast cancer adds weight to the belief that BBD with atypia is a precursor or marker lesion for breast cancer. Better understanding of the risk factors for the different types of BBD, defined according to standardized criteria, will help to clarify the pathway leading to breast cancer, thereby increasing options for prevention.

REFERENCES

  1. Top of page
  2. Abstract
  3. MATERIAL AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES