Bilateral Risk for Subsequent Breast Cancer After Lobular Carcinoma-In-Situ: Analysis of Surveillance, Epidemiology, and End Results Data
http://www.100md.com
《临床肿瘤学》
the Department of Radiation Oncology, St John Hospital Van Elslander Cancer Center and Webber Cancer Center, Grosse Pointe
Department of Pediatrics, Children's Hospital of Michigan
Department of Family Practice, Wayne State University School of Medicine, Detroit
Department of Pathology, University of Michigan, Ann Arbor, MI
Department of Radiation Oncology, and Cedars-Sinai Comprehensive Cancer Center, Los Angeles, CA
ABSTRACT
PURPOSE: Noninvasive lesions involving the lobules of the breast are increasingly diagnosed as incidental microscopic findings at the time of lumpectomy or core-needle biopsy. We investigated the incidence rates of invasive breast cancer (IBC) after a diagnosis of lobular carcinoma-in-situ (LCIS) by using Surveillance, Epidemiology, and End Results (SEER) data.
PATIENTS AND METHODS: Patients (N = 4,853) having a diagnosis of primary LCIS in the time period of 1973 to 1998 were identified using the SEER Public Use CD-ROM data. The database was then searched for patients with subsequent primary IBC occurrences (n = 350). The clinical and pathologic characteristics of patients with subsequent primary IBCs were compared with the characteristics of patients with primary IBCs attained during the same time period (N = 255,114).
RESULTS: The incidence of IBC increased over time from diagnosis of LCIS, with 7.1% ± 0.5% incidence of IBC at 10 years. IBCs detected after partial mastectomy occurred in either breast (46% ipsilateral and 54% contralateral); however, after mastectomy, most IBCs were contralateral (94.7%). IBCs occurring after LCIS more often represented invasive lobular histology (23.1%) compared with primary IBCs (6.5%). The standardized incidence ratio (the ratio of observed to expected cases) for developing IBC was 2.4 (95% CI, 2.1 to 2.6) adjusted for age and year of diagnosis.
CONCLUSION: LCIS is associated with increased risk of subsequent invasive disease, with equal predisposition in either breast. The minimum risk of developing IBC after LCIS is 7.1% at 10 years.
INTRODUCTION
Lobular carcinoma-in-situ (LCIS) is a clinically undetectable form of noninvasive breast cancer.1,2 Unlike ductal carcinoma-in-situ (DCIS), LCIS is typically confined to the lobules and terminal ducts of the breast and is almost exclusively diagnosed as an incidental finding when biopsy is performed for another reason.2-5 LCIS differs from atypical lobular hyperplasia (ALH) only in the degree of neoplasia,6-9 and these entities have been studied both together7,10 and separately.9,11 Because of increased rates of biopsy for mammographic findings, the diagnosis of LCIS has increased.5,9,12 For many years, LCIS has been considered a predisposing risk factor for the subsequent development of malignant disease in either breast,3,8,12-14 which may be greater for high-grade or more extensive lesions.10,15 In addition, some series10,16 have suggested preferential occurrence of invasive cancer in the ipsilateral breast at the site of biopsied LCIS (ie, suggestive of an actual precursor lesion or preinvasive proliferation). Progression to invasive disease seems to be inhibited by the use of tamoxifen.17 For the current study, Surveillance, Epidemiology, and End Results (SEER) data18 were analyzed to quantify and examine the laterality of invasive breast cancer (IBC) occurring after LCIS.
PATIENTS AND METHODS
Women having a diagnosis of LCIS were identified (n = 4,853) in the SEER Cancer Incidence Public-Use Database, 1973 to 199818 using the International Classification of Disease for Oncology codes (WHO 2000).19 Excluded from the analysis were patients initially identified based on autopsy reports or death certificates, patients with simultaneous occurrence of LCIS and IBC, and male breast cancers. The database was queried to identify patients subsequently diagnosed with an IBC at 1 or more years after diagnosis of primary breast cancer. IBC patients diagnosed at less than 1 year from LCIS diagnosis were excluded to eliminate the possibility of concurrent diagnosis. Of 49 IBCs occurring at less than 1 year, 34 were contralateral, and 15 were ipsilateral. Thirty-three mastectomies were performed for IBCs found in the first year. Patients with IBC after LCIS were grouped by laterality, age, the type of initial surgery, and the International Classification of Disease for Oncology morphology codes. The surgical categories were lesser surgery (ie, needle biopsy or aspirate; n = 178), partial mastectomy (segmental mastectomy, lumpectomy, quadrantectomy, tylectomy, wedge resection, nipple resection, or excisional biopsy; n = 3,141), and mastectomy (simple mastectomy, modified radical mastectomy, and radical mastectomy; n = 1,281).
To compare patients with IBC occurring after LCIS with IBC patients in the population at large, data were collected for all SEER patients with primary IBC who were diagnosed from 1973 to 1998. Data included the primary site, laterality, histology, and extent of disease for SEER IBC patients. The standardized incidence ratio (SIR) was calculated by comparing observed rates with expected rates as follows. First, the incidence of all primary malignant breast cancer in females within the SEER database (N = 255,114) was determined by year of diagnosis and by 5-year age groups. Next, the number of individuals at risk (ie, after diagnosis of LCIS) was determined by year of diagnosis and age group. Finally, the expected number of females developing IBC after LCIS was calculated by year of diagnosis and age group. These numbers were summed to give total expected patients. Ninety-five percent CI for the SIR were determined based on a Poisson distribution. Cumulative incidence was determined by the Kaplan-Meier method,20 with differences between groups assessed by the log-rank method.21,22 Contingency tables were evaluated by the 2 test.21,22
RESULTS
a total of 4,853 patients with LCIS identified in the SEER database between 1973 and 1998, 350 patients were found to have later developed IBC. In Table 1, the location within the breast, the laterality, and the histology for these patients are listed together with data from 255,114 patients with primary IBC identified within the SEER database over the same time period. Location within the breast (primary subsite) and laterality for the IBC after LCIS patients closely paralleled the data for primary IBC patients (Table 1). Approximately one third of IBCs occurred in the upper outer quadrant of the breast, and numbers of right-sided and left-sided IBCs were roughly equal. The histology for IBC after LCIS was more often lobular (23.1%) compared with primary IBC (6.5%), and this difference was mostly accounted for by a difference in the percentage of invasive ductal carcinomas (71.0% for primary IBC v 49.7% for IBC after LCIS).
In Table 2, the cumulative incidence of IBC is shown according to the age distribution of the LCIS patients. For all age groups, the rates of IBC continually increased over time. At 5, 10, and 15 years, 2,959 (60.9%), 1,664 (34.2%), and 724 (14.9%) LCIS patients remained at risk for IBC, respectively, and the incidence of IBC increased from 4.1% ± 0.3% (at 5 years) to 7.1% ± 0.5% (at 10 years) and to 11.3% ± 0.8% (at 15 years). As expected, the incidence of IBC was greater with increasing age (Table 2). For patients with LCIS under age 40 years, the incidence of IBC 10 years later was 5.6% ± 1.5% compared with 10.4% ± 1.4% for patients 60 to 69 years of age at LCIS diagnosis and 13.9% ± 2.5% for patients with LCIS older than age 70 years (P < .001). The age of developing IBC after LCIS seemed to parallel the age of developing IBC in the general population, and the SIRs for the five different age groups examined in Table 2 ranged from 2.1 to 3.3, with CIs overlapping. Regarding race, no statistically significant differences were detected when considering IBC incidence after LCIS for white, black, Hispanic, and other race categories (P = .18; Table 3). As expected, IBCs were detected with approximately equal frequency overall in patients having right-sided (6.8%) versus left-sided LCIS (5.5%). The SIR (comparing observed to expected numbers of patients in the general population) for developing IBC after LCIS was 2.4 (95% CI, 2.1 to 2.6) considering all patients.
To investigate whether the time period of diagnosis of LCIS influenced the rate of detection of second IBC, the data set was arbitrarily divided into patients accrued in the 12-year time period between 1973 and 1985 (n = 1,293) versus patients accrued in the 12-year time period between 1986 and 1998 (n = 3,560; Table 4). Using the observation period of 5 years, there was a 2.9% ± 0.5% incidence of IBC in the first time period (1973 to 1985) compared with 4.7% ± 0.4% in the second time period (1986 to 1998). Using the 10-year observation period, this difference persisted with 5.0% ± 0.6% v 8.4% ± 0.7% IBCs observed in the first and second time periods, respectively (P < .0001; Table 4).
In Table 5, the incidence rates for IBCs are listed and categorized according to the type of surgery performed for the preceding LCIS. Most patients (n = 3,141; 68%) had partial mastectomy with or without axillary lymph node dissection. One thousand two hundred eighty-one patients (28%) had mastectomy (including total mastectomy, modified radical mastectomy, or radical mastectomy). Few patients were coded as having surgery less than partial mastectomy (n = 178). Rates of IBC occurrence for patients with partial mastectomy were greater than with mastectomy, and the magnitude of this difference remained constant over time. At 10 years, the incidence rates were 8.8% for partial mastectomy compared with 5.7% for mastectomy (ratio, 1.5); at 15 and 20 years, the rates were 12.6% v 9.5% and 16.7% v 11.0%, respectively (ratio, 1.5; Table 5).
Next, the incidence of ipsilateral versus contralateral IBC occurrences was compared (Table 6). For patients having partial mastectomy, the frequency of subsequent ipsilateral and contralateral IBC was nearly identical 5 to 25 years after LCIS. A substantial proportion of patients had mastectomy (n = 1,281), and as expected, this intervention was associated with dramatic (although not complete) reduction in the rate of ipsilateral occurrences of IBC, with less than 0.5% to 1.0% ipsilateral IBC 15 to 25 years after mastectomy (P < .001).
IBC detected after LCIS represented less advanced disease compared with primary IBC detected during the same time period (1973 to 1998). In Figure 1, the extent of disease for patients with primary IBC versus IBC after LCIS is compared in terms of the primary tumor size and the lymph node status. Patients with a primary tumor size of less than 1 cm represented 21.4% of primary IBC patients compared with 42.9% of patients with IBC detected after LCIS. Similarly, extent of disease with respect to lymph node positivity was greater in patients with primary IBC compared with patients with IBC after LCIS. Patients with negative lymph nodes made up 52.8% of primary IBC patients compared with 59.7% of patients with IBC after LCIS. Correspondingly, smaller proportions of patients with IBC after LCIS had either one to three or more than three positive lymph nodes. Approximately 20% of patients in each group had either no lymph node dissection or no information collected regarding lymph node dissection.
DISCUSSION
The subsequent breast cancer risk that is conferred on a patient having a primary diagnosis of LCIS has the unique attribute of not only affecting the primary site of disease but also the contralateral breast.8,13,14,23 LCIS is uncommon,24 and over several decades, changing risk estimates for secondary IBC have led to differing clinical recommendations.14,25 On the basis of results for LCIS in patients treated at Memorial Hospital between 1952 and 1965, Rosen et al25,26 recommended ipsilateral mastectomy and contralateral breast biopsy. If LCIS was found in the contralateral biopsy, a mastectomy was then recommended for this breast as well. Currently, the most common clinical recommendation for women with LCIS is close follow-up as proposed by Haagensen et al,8,14 together with chemoprevention.17 Prophylactic bilateral mastectomy has been used in selected patients. Regarding core-needle biopsies showing ALH or LCIS, Arpino et al9 reported that 10% to 20% of specimens from follow-up surgical excision contained DCIS or IBC and suggested that complete excision could be of benefit in this situation. Unfortunately, few studies have had sufficient power to adequately investigate the long-term subsequent IBC risk associated with LCIS.
The SEER Program,18 which was initiated in 1973, collects cancer-related data for 11 (formerly nine) geographic registries representing 14% of the US population. Importantly for the current study, each registry tracks information on invasive and selected in situ cancers. Previously, we used population-based SEER data for the study of the occurrence of second malignancy among breast cancer survivors because of the relatively large numbers of events.27-29 Using SEER data and a similar methodology, we looked at progression to invasive disease in women with LCIS.
The current population-based study of 4,853 women with LCIS has the advantage of examining many more IBC events compared with previously published patient series (IBC occurrences = 350). At 10 years of follow-up, the incidence of IBC after LCIS was 7.1% for all patients. Considering patients who had partial mastectomy versus complete mastectomy, the incidence rates were 8.8% and 5.7%, respectively (Table 5).
Although the median age group for developing LCIS was 45 to 49 years, older females had higher relative risk for developing secondary IBC (Table 2). The SIR (ratio of observed incidence to the expected incidence) was 2.4. The cumulative risk of developing IBC continued to increase 15 to 25 years after LCIS diagnosis. Comparison of the risk level calculated here with previous single-institution data reveals a large range of incidence rates. Haagensen et al14 reported 10 patients with IBC in 53 patients with LCIS, for an 18% crude incidence rate with 16 years of minimum follow-up. Rosen et al26 showed a 29% rate, with 29 IBCs in 99 LCIS patients. Other smaller studies (32 to 80 LCIS patients) had no more than four to nine IBC occurrences each, with incidence rates ranging from 6% to 12% at 5 years and 12% to 23% at 18 years of follow-up.4,23,25,30,31
Among the most informative data regarding IBC after LCIS available to date are those derived from the National Surgical Adjuvant Breast and Bowel Project (NSABP) studies.11,15,17 A report of 182 patients with LCIS treated with local excision alone as part of the NSABP B-17 study15 was recently updated with 12 years of follow-up.11 All of these patients were initially thought to have had DCIS. At 5 years, a total of 13 ipsilateral breast tumor recurrences (IBTRs) and four contralateral breast tumor recurrences (CBTRs; including one bilateral recurrence) were observed. In the updated study, a total of 26 IBTRs (14.4%) and 14 CBTRs (7.8%) were reported. Nine of 26 IBTRs (5.0% of the total cohort) and 10 of 14 CBTRs (5.6% of the total cohort) were invasive. Pathologic grade of 2 to 3 seemed to be predictive for invasive ipsilateral recurrence. Interestingly, a high proportion of invasive IBTRs (89%) and invasive CBTRs (75%) were of lobular histology. This contrasts with the 23.1% proportion found in our study.
In NSABP P-1,17 a total of 13,338 women at increased risk for developing breast cancer were randomly assigned to receive placebo or tamoxifen 20 mg/d for 5 years. A 49% decrease in the occurrence of IBC for women receiving tamoxifen compared with women receiving placebo was shown. Most relevant to our analysis is that 826 women were accrued to NSABP P-1 because of a prior diagnosis of LCIS (411 patients on the placebo arm and 415 patients on the tamoxifen arm). A total of 26 IBCs were observed among these patients (18 patients receiving placebo and eight patients receiving tamoxifen). The average annual rate per 1,000 women was reported as 12.99 and 5.69 for the placebo and tamoxifen arms, respectively.17 Thus, among 411 LCIS patients treated with placebo on NSABP P-1, an estimated 6.5% could be expected to have developed IBC at 5 years (approximately 4.7% considering both the treatment and placebo arms). These numbers compare with the SEER database at 5 years, which shows a 4.1% ± 0.3% 5-year IBC incidence rate detected in 2,959 LCIS patients at risk (Table 2). Considering only patients diagnosed in the later time period (1986 to 1998), the 5-year SEER rate was identical to the 4.7% rate from NSABP P-1. Considering only patients treated with partial mastectomy, the SEER rate was 4.8%.
The question of whether LCIS may represent a true precursor lesion7,10 versus a prognostic marker4,8,14,32 for the subsequent development of IBC in either breast has been addressed by investigating the laterality of the second event (ie, a true precursor lesion would be associated with higher frequency of later IBC in the same breast). The SEER data did not demonstrate a difference between the incidence rates of secondary IBC in the unilateral versus contralateral breast (Table 6). Thus, from these data, LCIS would seem to be a marker of risk for which close follow-up may be considered adequate.
This result agrees with most previous studies,4,14,32 with some exceptions. First, as part of the Nashville Breast Cancer Studies,7 a retrospective cohort of 252 women with ALH (including LCIS) found in benign biopsies that had been performed between 1950 and 1985 mostly for palpable abnormalities was followed. Fifty of these 252 patients later developed IBC. The IBC was ipsilateral in 34 patients (68%) and contralateral in 12 patients (24%). The authors concluded that IBC after a diagnosis of ALH was three times more likely to be diagnosed in the ipsilateral breast.7 Second, the Danish Breast Cancer Cooperative Group16,30 collected follow-up data for 275 women treated with excision alone for DCIS (n = 142), LCIS (n = 100), or DCIS plus LCIS (n = 26) between 1982 and 1989. With a median follow-up of 10 years, 40 patients experienced recurrence as invasive carcinomas, and no statistical difference was found regarding development of IBC between the three groups. Considering only the LCIS patients, 18 recurrences were detected (13 IBC, one DCIS, and four DCIS plus LCIS). Sixteen of 18 recurrences were in the ipsilateral breast. Third, Marshall et al33 examined the subsequent breast cancer risk for ALH among participants in the Nurse's Health Study.34 Ipsilateral occurrence (in the same breast as the biopsy) was demonstrated in seven of 11 women who developed IBC. Interestingly, ALH was more strongly associated with premenopausal than postmenopausal status.33 Why the results of these studies differ from the SEER and other data is not clear, but the reason may be related to pathologic factors (ie, differences in grade or extent of disease), unknown patient selection factors, or numbers of events analyzed. One factor could be that, in our study, patients with IBC diagnosed less than 1 year after LCIS were excluded to eliminate the possibility of simultaneous occurrences. The age ranges for different study populations could also be important because, in our study, the rates of IBC after LCIS were greater in older individuals.
For the SEER data, the IBC detected after LCIS most often was of invasive ductal or related histology (49.7%); nonetheless, an unusually large proportion (23.1%) showed lobular histology (Table 1). This is a considerably greater proportion of lobular disease than is observed in the general breast cancer population (6.5%). Previous investigators have also pointed out increased incidence of invasive lobular carcinoma in patients with a history of LCIS and, in most of these studies, 25% to 35% of secondary IBCs were invasive lobular.4,14,26 Studies of the influence of LCIS on the risk of local recurrence in breast cancer have also suggested LCIS to be much more likely to be associated with invasive lobular carcinoma. In a study reported by Sasson et al,35 LCIS was present in 65 (5%) of 1,274 patients with stage I and II breast cancer. LCIS was more likely to be associated with an invasive lobular carcinoma (30 of 59 patients; 51%) than with invasive ductal carcinoma (26 of 1,125 patients; 2%).
The implication would be that underlying genetic (somatic or germline) defects that manifest as lobular in situ disease later lead to invasive disease that is fated to preferentially involve lobular tissue. Candidate genes include germline mutations of MLH1 and MSH2.35 Another molecular correlate for LCIS (and for ALH) is the altered expression of the E-cadherin gene located at chromosome 16q22.36-40 Comparative genomic hybridization analysis performed after microdissection37-39 has confirmed a high rate of losses from chromosome 16q in LCIS. Most patients with 16q loss also show altered E-cadherin expression. Such molecular changes are similar to those found in invasive lobular carcinomas.
Mixed results regarding the recurrence rates of IBCs that are associated with LCIS have been reported. Sasson et al35 found that IBTR occurred in 57 (5%) of 1,209 patients without LCIS compared with 10 (15%) of 65 patients with associated LCIS (P = .001). Abner et al41 found an 8-year recurrence rates of 13% for moderate or marked LCIS adjacent to the tumor in 137 patients and 12% for 1,062 patients without LCIS. They concluded that neither the presence nor extent of LCIS associated with IBC should influence management decisions.41 In the current study, the subsequent outcome for patients with IBC detected after LCIS was not evaluated.
Although the SEER data analyzed in this study reflects the largest study of this topic to date, a number of possible limitations need to be addressed. First is the potential for lead-time bias. The data show that having an LCIS diagnosis in the later years of the study increased the risk of being diagnosed with IBC (Table 4; P = .04). For example, at 5 years, there was a 2.9% ± 0.5% incidence of IBC in the first 12 years of the study (1973 to 1985) compared with 4.7% ± 0.4% in the second 12 years (1986 to 1998). Two main factors are likely to account for this difference. First, patients with LCIS may have been more diligently screened for breast cancer compared with the general population in the later time period. Second, mammographic screening increased for the general population during the study period. Both of these factors increase the potential for early detection of breast cancer. The effect of lead-time bias is also manifested in the comparison of extent of disease at diagnosis for IBC detected in the general population versus IBC detected after LCIS (Fig 1). Observed extent of disease would be expected to be lower in cohorts having greater screening. Breast cancers detected in patients who had prior LCIS in the current study were smaller and less likely to be associated with lymph node metastasis (ie, consistent with increased screening). The slightly higher rate of IBC after LCIS observed in the later time period of the study (eg, 8.4% at 10 years; Table 4) may, therefore, be closer to the true rate in the current era.
A general limitation for population-based data is the possibility of under-reporting or imperfect ascertainment, and reported rates are generally considered underestimates. Nonetheless, ascertainment levels for the SEER database are high. A case-finding study performed in six different SEER areas (63% of the total SEER caseload) assessed the completeness of reporting for 6,342 patients diagnosed and/or treated in 74 hospitals.42 In each hospital, 2 study months were selected at random, and patients in hospital department files were then matched against lists of patients entered in the SEER system. From the 6,342 reportable cancer diagnoses, 149 had not been entered into the SEER program, giving an estimated completeness of reporting for 1987 of 97.7%. Patients with breast malignancy accounted for nine of the 149 patients not entered into SEER.42 Incompleteness was highest for hospitals with the smallest caseload.42
Other limitations regarding the current study include the fact that both practice patterns and pathologic definitions for LCIS changed during the study time period. Because pathologic review is impossible, we cannot make any definite statement regarding the extent of LCIS or nuclear grade, both of which may have prognostic value. The reported data reflect the common usage of the criteria for the LCIS category by pathologists within the SEER reporting areas during the study time period. Similarly, current practice with chemoprevention for LCIS is not reflected in the SEER data presented here. Although second malignancy in patients with breast primaries may be influenced by radiotherapy,29 this factor is largely absent for LCIS. Despite these limitations, the SEER data indicate clearly that the risk of IBC after LCIS is equal in both breasts. This finding is exemplified by the fact that LCIS patients who had partial mastectomy had equal rates of ipsilateral versus contralateral IBC, whereas patients who had mastectomy had few secondary ipsilateral IBCs (Table 6). It should be noted that possible drawbacks of population-based studies (ie, lead-time bias and underascertainment) do not influence the laterality of the subsequent cancers detected.
In conclusion, LCIS may be considered a predisposing determinant of risk for subsequent invasive disease in either breast. Risk reduction with unilateral mastectomy is only modest, and complete excision after lumpectomy is unlikely to result in significant risk reduction. Further work is necessary to determine whether ALH or LCIS diagnosed on core-needle biopsy may require excision to fully evaluate the extent of disease.9 Prophylactic bilateral mastectomy may be an option for selected patients. The minimum cumulative risk of developing IBC after LCIS was 7.1% at 10 years, with risk equally distributed between both breasts.
Authors' Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
NOTES
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Foote FW Jr, Stewart FW: Lobular carcinoma in situ: A rare form of mammary cancer. Am J Pathol 17:491-496, 1941
Walt A, Simon M, Swanson GM: The continuing dilemma of lobular carcinoma in situ. Arch Surg 127:904-907, 1992
Goldschmidt RA, Victor TA: Lobular carcinoma in situ of the breast. Semin Surg Oncol 12:314-320, 1996
Anderson JA: Lobular carcinoma in situ of the breast: An approach to rational treatment. Cancer 39:2597-2602, 1977
Li CI, Anderson BO, Daling JR, et al: Changing incidence of lobular carcinoma in situ of the breast. Breast Cancer Res Treat 5:259-268, 2002
Yarbro JW, Page DL, Fielding LP, et al: American Joint Committee on Cancer Prognostic Factors Consensus Conference. Cancer 86:2436-2446, 1999
Page DL, Schuyler PA, Dupont WD, et al: Atypical lobular hyperplasia as a unilateral predictor of breast cancer risk: A retrospective cohort study. Lancet 361:125-129, 2003
Haagensen CD, Lane N, Lattes R, et al: Lobular neoplasia (so-called lobular carcinoma in situ) of the breast. Cancer 42:737-769, 1978
Arpino G, Allred DC, Mohsin SK, et al: Lobular neoplasia on core-needle biopsy: Clinical significance. Cancer 101:242-250, 2004
Page DL, Kidd TE Jr, Dupont WD, et al: Lobular neoplasia of the breast: Higher risk for subsequent invasive cancer predicted by more extensive disease. Human Pathol 22:1232-1239, 1991
Fisher ER, Land SR, Fisher B, et al: Pathologic findings from the National Surgical Adjuvant Breast and Bowel Project: Twelve-year observations concerning lobular carcinoma in situ. Cancer 100:238-244, 2004
Levi F, Te VC, Randimbison L, et al: Trends of in situ carcinoma of the breast in Vaud, Switzerland. Eur J Cancer 33:903-906, 1997
Franceschi S, Levi F, La Vecchia C, et al: Second cancers following in situ carcinoma of the breast. Int J Cancer 77:392-395, 1998
Haagensen CD, Bodian C, Haagensen DE: Lobular Neoplasia (Lobular Carcinoma In Situ) Breast Carcinoma: Risk and Detection. Philadelphia, PA, WB Saunders, 1981, p 238
Fisher ER, Costantino J, Fisher B, et al: Pathologic findings from the National Surgical Adjuvant Breast Project (NSABP) Protocol B-17: Five year observations concerning lobular carcinoma in situ. Cancer 78:1403-1416, 1996
Ottesen GL, Graversen HP, Blichert-Toft M, et al: Carcinoma in situ of the female breast. 10 year follow-up results of a prospective nationwide study. Breast Cancer Res Treat 62:197-210, 2000
Fisher B, Constantino JP, Wickerman DL, et al: Tamoxifen for prevention of breast cancer: Report of the National Surgical Adjuvant Breast and Bowel Project P-1 Study. J Natl Cancer Inst 90:1371-1388, 1998
National Cancer Institute: Surveillance, Epidemiology, and End Results (SEER) Program Public-Use CD-ROM (1973-1998). Bethesda, MD, National Cancer Institute, 1999
Percy C, Fritz A, Jack A, et al: (eds): International Classification of Diseases for Oncology (ed 3). Geneva, Switzerland, World Health Organization, 2000
Kaplan EL, Meier P: Nonparametric estimation from incomplete observations. J Am Statist Assoc 53:457-481, 1958
Lee ET: Statistical Methods for Survival Analysis. Belmont, CA, Lifetime Learning Publications, 1980
Snedecor GW, Cochran WG: Statistical Methods (ed 7). Ames, IA, The Iowa State University Press, 1980
Salvadori B, Bartoli C, Zurrida S, et al: Risk of invasive cancer in women with lobular carcinoma in situ of the breast. Eur J Cancer Clin Oncol 27:35-37, 1991
Akashi-Tanaka S, Fukutomi T, Nanasawa T, et al: Treatment of non-invasive carcinoma: Fifteen year results at the National Cancer Center Hospital in Tokyo. Breast Cancer 7:341-344, 2000
Rosen PP, Braun DW, Lynggholm B, et al: Lobular carcinoma in situ of the breast: Preliminary results of treatment by ipsilateral mastectomy and contralateral breast biopsy. Cancer 47:813-819, 1981
Rosen PP, Lieberman PH, Braun DW Jr, et al: Lobular carcinoma in situ of the breast. Am J Surg Pathol 2:225-251, 1978
Devarahally SR, Severson RK, Chuba P, et al: Second malignant neoplasms after primary central nervous system malignancies of childhood and adolescence. Pediatr Hematol Oncol 20:617-625, 2003
Hamre MR, Severson RK, Chuba P, et al: Osteosarcoma as a second malignant neoplasm. Radiother Oncol 65:153-157, 2002
Yap J, Chuba PJ, Thomas R, et al: Sarcoma as a second malignancy after treatment for breast cancer. Int J Radiat Oncol Biol Phys 52:1231-1237, 2002
Ottesen GL, Graversen HP, Blichert-Toft M, et al: Lobular carcinoma in situ of the female breast: Short-term results of a prospective nationwide study. Am J Surg Pathol 17:14-21, 1993
Wheeler JE, Enterline HT, Roseman JM, et al: Lobular carcinoma in situ of the breast: Long-term follow-up. Cancer 34:554-563, 1974
Bodian CA, Perzin KH, Latter R: Lobular neoplasia: Long term risk of breast cancer in relation to other factors. Cancer 78:1024-1034, 1996
Marshall LM, Hunter DJ, Connolly JL, et al: Risk of breast cancer associated with atypical hyperplasia of lobular and ductal types. Cancer Epidemiol Biomarkers Prev 6:297-301, 1997
Colditz GA, Rosner B: Cumulative risk of breast cancer to age 70 years according ot risk factor status: Data from the Nurses' Health Study. Am J Epidemiol 152:950-964, 2000
Sasson AR, Fowble B, Hanlon AL, et al: Lobular carcinoma in situ increases the risk of local recurrence in selected patients with stages I and II breast carcinoma treated with conservative surgery and radiation. Cancer 91:1862-1869, 2001
Stone JG, Coleman G, Gusterson B, et al: Contribution of germline MLH1 and MSH2 mutations to lobular carcinoma in situ of the breast. Cancer Lett 167:171-174, 2001
Goldstein NS, Kestin LL, Vicini FA: Clinicopathologic implications of E-cadherin reactivity in patients with lobular carcinoma in situ of the breast. Cancer 92:738-747, 2001
Rieger-Chris KM, Pezza JA, Dugan JM, et al: Disparate E-cadherin mutations in LCIS and associated invasive breast carcinomas. Mol Pathol 54:91-97, 2001
Lu YJ, Osin P, Lakhani SR, et al: Comparative genomic hybridization analysis of lobular carcinoma in situ and atypical lobular hyperplasia and potential roles for gains and losses of genetic material breast neoplasia. Cancer Res 58:4721-4727, 1998
Etzell JE, Devries S, Chew K, et al: Loss of chromosome 16q in lobular carcinoma in situ. Hum Pathol 32:292-296, 2001
Abner AL, Connolly JL, Recht A, et al: The relation between the presence and extent of lobular carcinoma in situ and the risk of local recurrence for patients with invasive carcinoma of the breast treated with conservative surgery and radiation therapy. Cancer 88:1072-1077, 2000
Zippin C, Lum BS, Hankey BF: Completeness of hospital cancer case reporting from the SEER program of the National Cancer Institute. Cancer 76:2343-2350, 1995(Paul J. Chuba, Merlin R. )
Department of Pediatrics, Children's Hospital of Michigan
Department of Family Practice, Wayne State University School of Medicine, Detroit
Department of Pathology, University of Michigan, Ann Arbor, MI
Department of Radiation Oncology, and Cedars-Sinai Comprehensive Cancer Center, Los Angeles, CA
ABSTRACT
PURPOSE: Noninvasive lesions involving the lobules of the breast are increasingly diagnosed as incidental microscopic findings at the time of lumpectomy or core-needle biopsy. We investigated the incidence rates of invasive breast cancer (IBC) after a diagnosis of lobular carcinoma-in-situ (LCIS) by using Surveillance, Epidemiology, and End Results (SEER) data.
PATIENTS AND METHODS: Patients (N = 4,853) having a diagnosis of primary LCIS in the time period of 1973 to 1998 were identified using the SEER Public Use CD-ROM data. The database was then searched for patients with subsequent primary IBC occurrences (n = 350). The clinical and pathologic characteristics of patients with subsequent primary IBCs were compared with the characteristics of patients with primary IBCs attained during the same time period (N = 255,114).
RESULTS: The incidence of IBC increased over time from diagnosis of LCIS, with 7.1% ± 0.5% incidence of IBC at 10 years. IBCs detected after partial mastectomy occurred in either breast (46% ipsilateral and 54% contralateral); however, after mastectomy, most IBCs were contralateral (94.7%). IBCs occurring after LCIS more often represented invasive lobular histology (23.1%) compared with primary IBCs (6.5%). The standardized incidence ratio (the ratio of observed to expected cases) for developing IBC was 2.4 (95% CI, 2.1 to 2.6) adjusted for age and year of diagnosis.
CONCLUSION: LCIS is associated with increased risk of subsequent invasive disease, with equal predisposition in either breast. The minimum risk of developing IBC after LCIS is 7.1% at 10 years.
INTRODUCTION
Lobular carcinoma-in-situ (LCIS) is a clinically undetectable form of noninvasive breast cancer.1,2 Unlike ductal carcinoma-in-situ (DCIS), LCIS is typically confined to the lobules and terminal ducts of the breast and is almost exclusively diagnosed as an incidental finding when biopsy is performed for another reason.2-5 LCIS differs from atypical lobular hyperplasia (ALH) only in the degree of neoplasia,6-9 and these entities have been studied both together7,10 and separately.9,11 Because of increased rates of biopsy for mammographic findings, the diagnosis of LCIS has increased.5,9,12 For many years, LCIS has been considered a predisposing risk factor for the subsequent development of malignant disease in either breast,3,8,12-14 which may be greater for high-grade or more extensive lesions.10,15 In addition, some series10,16 have suggested preferential occurrence of invasive cancer in the ipsilateral breast at the site of biopsied LCIS (ie, suggestive of an actual precursor lesion or preinvasive proliferation). Progression to invasive disease seems to be inhibited by the use of tamoxifen.17 For the current study, Surveillance, Epidemiology, and End Results (SEER) data18 were analyzed to quantify and examine the laterality of invasive breast cancer (IBC) occurring after LCIS.
PATIENTS AND METHODS
Women having a diagnosis of LCIS were identified (n = 4,853) in the SEER Cancer Incidence Public-Use Database, 1973 to 199818 using the International Classification of Disease for Oncology codes (WHO 2000).19 Excluded from the analysis were patients initially identified based on autopsy reports or death certificates, patients with simultaneous occurrence of LCIS and IBC, and male breast cancers. The database was queried to identify patients subsequently diagnosed with an IBC at 1 or more years after diagnosis of primary breast cancer. IBC patients diagnosed at less than 1 year from LCIS diagnosis were excluded to eliminate the possibility of concurrent diagnosis. Of 49 IBCs occurring at less than 1 year, 34 were contralateral, and 15 were ipsilateral. Thirty-three mastectomies were performed for IBCs found in the first year. Patients with IBC after LCIS were grouped by laterality, age, the type of initial surgery, and the International Classification of Disease for Oncology morphology codes. The surgical categories were lesser surgery (ie, needle biopsy or aspirate; n = 178), partial mastectomy (segmental mastectomy, lumpectomy, quadrantectomy, tylectomy, wedge resection, nipple resection, or excisional biopsy; n = 3,141), and mastectomy (simple mastectomy, modified radical mastectomy, and radical mastectomy; n = 1,281).
To compare patients with IBC occurring after LCIS with IBC patients in the population at large, data were collected for all SEER patients with primary IBC who were diagnosed from 1973 to 1998. Data included the primary site, laterality, histology, and extent of disease for SEER IBC patients. The standardized incidence ratio (SIR) was calculated by comparing observed rates with expected rates as follows. First, the incidence of all primary malignant breast cancer in females within the SEER database (N = 255,114) was determined by year of diagnosis and by 5-year age groups. Next, the number of individuals at risk (ie, after diagnosis of LCIS) was determined by year of diagnosis and age group. Finally, the expected number of females developing IBC after LCIS was calculated by year of diagnosis and age group. These numbers were summed to give total expected patients. Ninety-five percent CI for the SIR were determined based on a Poisson distribution. Cumulative incidence was determined by the Kaplan-Meier method,20 with differences between groups assessed by the log-rank method.21,22 Contingency tables were evaluated by the 2 test.21,22
RESULTS
a total of 4,853 patients with LCIS identified in the SEER database between 1973 and 1998, 350 patients were found to have later developed IBC. In Table 1, the location within the breast, the laterality, and the histology for these patients are listed together with data from 255,114 patients with primary IBC identified within the SEER database over the same time period. Location within the breast (primary subsite) and laterality for the IBC after LCIS patients closely paralleled the data for primary IBC patients (Table 1). Approximately one third of IBCs occurred in the upper outer quadrant of the breast, and numbers of right-sided and left-sided IBCs were roughly equal. The histology for IBC after LCIS was more often lobular (23.1%) compared with primary IBC (6.5%), and this difference was mostly accounted for by a difference in the percentage of invasive ductal carcinomas (71.0% for primary IBC v 49.7% for IBC after LCIS).
In Table 2, the cumulative incidence of IBC is shown according to the age distribution of the LCIS patients. For all age groups, the rates of IBC continually increased over time. At 5, 10, and 15 years, 2,959 (60.9%), 1,664 (34.2%), and 724 (14.9%) LCIS patients remained at risk for IBC, respectively, and the incidence of IBC increased from 4.1% ± 0.3% (at 5 years) to 7.1% ± 0.5% (at 10 years) and to 11.3% ± 0.8% (at 15 years). As expected, the incidence of IBC was greater with increasing age (Table 2). For patients with LCIS under age 40 years, the incidence of IBC 10 years later was 5.6% ± 1.5% compared with 10.4% ± 1.4% for patients 60 to 69 years of age at LCIS diagnosis and 13.9% ± 2.5% for patients with LCIS older than age 70 years (P < .001). The age of developing IBC after LCIS seemed to parallel the age of developing IBC in the general population, and the SIRs for the five different age groups examined in Table 2 ranged from 2.1 to 3.3, with CIs overlapping. Regarding race, no statistically significant differences were detected when considering IBC incidence after LCIS for white, black, Hispanic, and other race categories (P = .18; Table 3). As expected, IBCs were detected with approximately equal frequency overall in patients having right-sided (6.8%) versus left-sided LCIS (5.5%). The SIR (comparing observed to expected numbers of patients in the general population) for developing IBC after LCIS was 2.4 (95% CI, 2.1 to 2.6) considering all patients.
To investigate whether the time period of diagnosis of LCIS influenced the rate of detection of second IBC, the data set was arbitrarily divided into patients accrued in the 12-year time period between 1973 and 1985 (n = 1,293) versus patients accrued in the 12-year time period between 1986 and 1998 (n = 3,560; Table 4). Using the observation period of 5 years, there was a 2.9% ± 0.5% incidence of IBC in the first time period (1973 to 1985) compared with 4.7% ± 0.4% in the second time period (1986 to 1998). Using the 10-year observation period, this difference persisted with 5.0% ± 0.6% v 8.4% ± 0.7% IBCs observed in the first and second time periods, respectively (P < .0001; Table 4).
In Table 5, the incidence rates for IBCs are listed and categorized according to the type of surgery performed for the preceding LCIS. Most patients (n = 3,141; 68%) had partial mastectomy with or without axillary lymph node dissection. One thousand two hundred eighty-one patients (28%) had mastectomy (including total mastectomy, modified radical mastectomy, or radical mastectomy). Few patients were coded as having surgery less than partial mastectomy (n = 178). Rates of IBC occurrence for patients with partial mastectomy were greater than with mastectomy, and the magnitude of this difference remained constant over time. At 10 years, the incidence rates were 8.8% for partial mastectomy compared with 5.7% for mastectomy (ratio, 1.5); at 15 and 20 years, the rates were 12.6% v 9.5% and 16.7% v 11.0%, respectively (ratio, 1.5; Table 5).
Next, the incidence of ipsilateral versus contralateral IBC occurrences was compared (Table 6). For patients having partial mastectomy, the frequency of subsequent ipsilateral and contralateral IBC was nearly identical 5 to 25 years after LCIS. A substantial proportion of patients had mastectomy (n = 1,281), and as expected, this intervention was associated with dramatic (although not complete) reduction in the rate of ipsilateral occurrences of IBC, with less than 0.5% to 1.0% ipsilateral IBC 15 to 25 years after mastectomy (P < .001).
IBC detected after LCIS represented less advanced disease compared with primary IBC detected during the same time period (1973 to 1998). In Figure 1, the extent of disease for patients with primary IBC versus IBC after LCIS is compared in terms of the primary tumor size and the lymph node status. Patients with a primary tumor size of less than 1 cm represented 21.4% of primary IBC patients compared with 42.9% of patients with IBC detected after LCIS. Similarly, extent of disease with respect to lymph node positivity was greater in patients with primary IBC compared with patients with IBC after LCIS. Patients with negative lymph nodes made up 52.8% of primary IBC patients compared with 59.7% of patients with IBC after LCIS. Correspondingly, smaller proportions of patients with IBC after LCIS had either one to three or more than three positive lymph nodes. Approximately 20% of patients in each group had either no lymph node dissection or no information collected regarding lymph node dissection.
DISCUSSION
The subsequent breast cancer risk that is conferred on a patient having a primary diagnosis of LCIS has the unique attribute of not only affecting the primary site of disease but also the contralateral breast.8,13,14,23 LCIS is uncommon,24 and over several decades, changing risk estimates for secondary IBC have led to differing clinical recommendations.14,25 On the basis of results for LCIS in patients treated at Memorial Hospital between 1952 and 1965, Rosen et al25,26 recommended ipsilateral mastectomy and contralateral breast biopsy. If LCIS was found in the contralateral biopsy, a mastectomy was then recommended for this breast as well. Currently, the most common clinical recommendation for women with LCIS is close follow-up as proposed by Haagensen et al,8,14 together with chemoprevention.17 Prophylactic bilateral mastectomy has been used in selected patients. Regarding core-needle biopsies showing ALH or LCIS, Arpino et al9 reported that 10% to 20% of specimens from follow-up surgical excision contained DCIS or IBC and suggested that complete excision could be of benefit in this situation. Unfortunately, few studies have had sufficient power to adequately investigate the long-term subsequent IBC risk associated with LCIS.
The SEER Program,18 which was initiated in 1973, collects cancer-related data for 11 (formerly nine) geographic registries representing 14% of the US population. Importantly for the current study, each registry tracks information on invasive and selected in situ cancers. Previously, we used population-based SEER data for the study of the occurrence of second malignancy among breast cancer survivors because of the relatively large numbers of events.27-29 Using SEER data and a similar methodology, we looked at progression to invasive disease in women with LCIS.
The current population-based study of 4,853 women with LCIS has the advantage of examining many more IBC events compared with previously published patient series (IBC occurrences = 350). At 10 years of follow-up, the incidence of IBC after LCIS was 7.1% for all patients. Considering patients who had partial mastectomy versus complete mastectomy, the incidence rates were 8.8% and 5.7%, respectively (Table 5).
Although the median age group for developing LCIS was 45 to 49 years, older females had higher relative risk for developing secondary IBC (Table 2). The SIR (ratio of observed incidence to the expected incidence) was 2.4. The cumulative risk of developing IBC continued to increase 15 to 25 years after LCIS diagnosis. Comparison of the risk level calculated here with previous single-institution data reveals a large range of incidence rates. Haagensen et al14 reported 10 patients with IBC in 53 patients with LCIS, for an 18% crude incidence rate with 16 years of minimum follow-up. Rosen et al26 showed a 29% rate, with 29 IBCs in 99 LCIS patients. Other smaller studies (32 to 80 LCIS patients) had no more than four to nine IBC occurrences each, with incidence rates ranging from 6% to 12% at 5 years and 12% to 23% at 18 years of follow-up.4,23,25,30,31
Among the most informative data regarding IBC after LCIS available to date are those derived from the National Surgical Adjuvant Breast and Bowel Project (NSABP) studies.11,15,17 A report of 182 patients with LCIS treated with local excision alone as part of the NSABP B-17 study15 was recently updated with 12 years of follow-up.11 All of these patients were initially thought to have had DCIS. At 5 years, a total of 13 ipsilateral breast tumor recurrences (IBTRs) and four contralateral breast tumor recurrences (CBTRs; including one bilateral recurrence) were observed. In the updated study, a total of 26 IBTRs (14.4%) and 14 CBTRs (7.8%) were reported. Nine of 26 IBTRs (5.0% of the total cohort) and 10 of 14 CBTRs (5.6% of the total cohort) were invasive. Pathologic grade of 2 to 3 seemed to be predictive for invasive ipsilateral recurrence. Interestingly, a high proportion of invasive IBTRs (89%) and invasive CBTRs (75%) were of lobular histology. This contrasts with the 23.1% proportion found in our study.
In NSABP P-1,17 a total of 13,338 women at increased risk for developing breast cancer were randomly assigned to receive placebo or tamoxifen 20 mg/d for 5 years. A 49% decrease in the occurrence of IBC for women receiving tamoxifen compared with women receiving placebo was shown. Most relevant to our analysis is that 826 women were accrued to NSABP P-1 because of a prior diagnosis of LCIS (411 patients on the placebo arm and 415 patients on the tamoxifen arm). A total of 26 IBCs were observed among these patients (18 patients receiving placebo and eight patients receiving tamoxifen). The average annual rate per 1,000 women was reported as 12.99 and 5.69 for the placebo and tamoxifen arms, respectively.17 Thus, among 411 LCIS patients treated with placebo on NSABP P-1, an estimated 6.5% could be expected to have developed IBC at 5 years (approximately 4.7% considering both the treatment and placebo arms). These numbers compare with the SEER database at 5 years, which shows a 4.1% ± 0.3% 5-year IBC incidence rate detected in 2,959 LCIS patients at risk (Table 2). Considering only patients diagnosed in the later time period (1986 to 1998), the 5-year SEER rate was identical to the 4.7% rate from NSABP P-1. Considering only patients treated with partial mastectomy, the SEER rate was 4.8%.
The question of whether LCIS may represent a true precursor lesion7,10 versus a prognostic marker4,8,14,32 for the subsequent development of IBC in either breast has been addressed by investigating the laterality of the second event (ie, a true precursor lesion would be associated with higher frequency of later IBC in the same breast). The SEER data did not demonstrate a difference between the incidence rates of secondary IBC in the unilateral versus contralateral breast (Table 6). Thus, from these data, LCIS would seem to be a marker of risk for which close follow-up may be considered adequate.
This result agrees with most previous studies,4,14,32 with some exceptions. First, as part of the Nashville Breast Cancer Studies,7 a retrospective cohort of 252 women with ALH (including LCIS) found in benign biopsies that had been performed between 1950 and 1985 mostly for palpable abnormalities was followed. Fifty of these 252 patients later developed IBC. The IBC was ipsilateral in 34 patients (68%) and contralateral in 12 patients (24%). The authors concluded that IBC after a diagnosis of ALH was three times more likely to be diagnosed in the ipsilateral breast.7 Second, the Danish Breast Cancer Cooperative Group16,30 collected follow-up data for 275 women treated with excision alone for DCIS (n = 142), LCIS (n = 100), or DCIS plus LCIS (n = 26) between 1982 and 1989. With a median follow-up of 10 years, 40 patients experienced recurrence as invasive carcinomas, and no statistical difference was found regarding development of IBC between the three groups. Considering only the LCIS patients, 18 recurrences were detected (13 IBC, one DCIS, and four DCIS plus LCIS). Sixteen of 18 recurrences were in the ipsilateral breast. Third, Marshall et al33 examined the subsequent breast cancer risk for ALH among participants in the Nurse's Health Study.34 Ipsilateral occurrence (in the same breast as the biopsy) was demonstrated in seven of 11 women who developed IBC. Interestingly, ALH was more strongly associated with premenopausal than postmenopausal status.33 Why the results of these studies differ from the SEER and other data is not clear, but the reason may be related to pathologic factors (ie, differences in grade or extent of disease), unknown patient selection factors, or numbers of events analyzed. One factor could be that, in our study, patients with IBC diagnosed less than 1 year after LCIS were excluded to eliminate the possibility of simultaneous occurrences. The age ranges for different study populations could also be important because, in our study, the rates of IBC after LCIS were greater in older individuals.
For the SEER data, the IBC detected after LCIS most often was of invasive ductal or related histology (49.7%); nonetheless, an unusually large proportion (23.1%) showed lobular histology (Table 1). This is a considerably greater proportion of lobular disease than is observed in the general breast cancer population (6.5%). Previous investigators have also pointed out increased incidence of invasive lobular carcinoma in patients with a history of LCIS and, in most of these studies, 25% to 35% of secondary IBCs were invasive lobular.4,14,26 Studies of the influence of LCIS on the risk of local recurrence in breast cancer have also suggested LCIS to be much more likely to be associated with invasive lobular carcinoma. In a study reported by Sasson et al,35 LCIS was present in 65 (5%) of 1,274 patients with stage I and II breast cancer. LCIS was more likely to be associated with an invasive lobular carcinoma (30 of 59 patients; 51%) than with invasive ductal carcinoma (26 of 1,125 patients; 2%).
The implication would be that underlying genetic (somatic or germline) defects that manifest as lobular in situ disease later lead to invasive disease that is fated to preferentially involve lobular tissue. Candidate genes include germline mutations of MLH1 and MSH2.35 Another molecular correlate for LCIS (and for ALH) is the altered expression of the E-cadherin gene located at chromosome 16q22.36-40 Comparative genomic hybridization analysis performed after microdissection37-39 has confirmed a high rate of losses from chromosome 16q in LCIS. Most patients with 16q loss also show altered E-cadherin expression. Such molecular changes are similar to those found in invasive lobular carcinomas.
Mixed results regarding the recurrence rates of IBCs that are associated with LCIS have been reported. Sasson et al35 found that IBTR occurred in 57 (5%) of 1,209 patients without LCIS compared with 10 (15%) of 65 patients with associated LCIS (P = .001). Abner et al41 found an 8-year recurrence rates of 13% for moderate or marked LCIS adjacent to the tumor in 137 patients and 12% for 1,062 patients without LCIS. They concluded that neither the presence nor extent of LCIS associated with IBC should influence management decisions.41 In the current study, the subsequent outcome for patients with IBC detected after LCIS was not evaluated.
Although the SEER data analyzed in this study reflects the largest study of this topic to date, a number of possible limitations need to be addressed. First is the potential for lead-time bias. The data show that having an LCIS diagnosis in the later years of the study increased the risk of being diagnosed with IBC (Table 4; P = .04). For example, at 5 years, there was a 2.9% ± 0.5% incidence of IBC in the first 12 years of the study (1973 to 1985) compared with 4.7% ± 0.4% in the second 12 years (1986 to 1998). Two main factors are likely to account for this difference. First, patients with LCIS may have been more diligently screened for breast cancer compared with the general population in the later time period. Second, mammographic screening increased for the general population during the study period. Both of these factors increase the potential for early detection of breast cancer. The effect of lead-time bias is also manifested in the comparison of extent of disease at diagnosis for IBC detected in the general population versus IBC detected after LCIS (Fig 1). Observed extent of disease would be expected to be lower in cohorts having greater screening. Breast cancers detected in patients who had prior LCIS in the current study were smaller and less likely to be associated with lymph node metastasis (ie, consistent with increased screening). The slightly higher rate of IBC after LCIS observed in the later time period of the study (eg, 8.4% at 10 years; Table 4) may, therefore, be closer to the true rate in the current era.
A general limitation for population-based data is the possibility of under-reporting or imperfect ascertainment, and reported rates are generally considered underestimates. Nonetheless, ascertainment levels for the SEER database are high. A case-finding study performed in six different SEER areas (63% of the total SEER caseload) assessed the completeness of reporting for 6,342 patients diagnosed and/or treated in 74 hospitals.42 In each hospital, 2 study months were selected at random, and patients in hospital department files were then matched against lists of patients entered in the SEER system. From the 6,342 reportable cancer diagnoses, 149 had not been entered into the SEER program, giving an estimated completeness of reporting for 1987 of 97.7%. Patients with breast malignancy accounted for nine of the 149 patients not entered into SEER.42 Incompleteness was highest for hospitals with the smallest caseload.42
Other limitations regarding the current study include the fact that both practice patterns and pathologic definitions for LCIS changed during the study time period. Because pathologic review is impossible, we cannot make any definite statement regarding the extent of LCIS or nuclear grade, both of which may have prognostic value. The reported data reflect the common usage of the criteria for the LCIS category by pathologists within the SEER reporting areas during the study time period. Similarly, current practice with chemoprevention for LCIS is not reflected in the SEER data presented here. Although second malignancy in patients with breast primaries may be influenced by radiotherapy,29 this factor is largely absent for LCIS. Despite these limitations, the SEER data indicate clearly that the risk of IBC after LCIS is equal in both breasts. This finding is exemplified by the fact that LCIS patients who had partial mastectomy had equal rates of ipsilateral versus contralateral IBC, whereas patients who had mastectomy had few secondary ipsilateral IBCs (Table 6). It should be noted that possible drawbacks of population-based studies (ie, lead-time bias and underascertainment) do not influence the laterality of the subsequent cancers detected.
In conclusion, LCIS may be considered a predisposing determinant of risk for subsequent invasive disease in either breast. Risk reduction with unilateral mastectomy is only modest, and complete excision after lumpectomy is unlikely to result in significant risk reduction. Further work is necessary to determine whether ALH or LCIS diagnosed on core-needle biopsy may require excision to fully evaluate the extent of disease.9 Prophylactic bilateral mastectomy may be an option for selected patients. The minimum cumulative risk of developing IBC after LCIS was 7.1% at 10 years, with risk equally distributed between both breasts.
Authors' Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
NOTES
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Foote FW Jr, Stewart FW: Lobular carcinoma in situ: A rare form of mammary cancer. Am J Pathol 17:491-496, 1941
Walt A, Simon M, Swanson GM: The continuing dilemma of lobular carcinoma in situ. Arch Surg 127:904-907, 1992
Goldschmidt RA, Victor TA: Lobular carcinoma in situ of the breast. Semin Surg Oncol 12:314-320, 1996
Anderson JA: Lobular carcinoma in situ of the breast: An approach to rational treatment. Cancer 39:2597-2602, 1977
Li CI, Anderson BO, Daling JR, et al: Changing incidence of lobular carcinoma in situ of the breast. Breast Cancer Res Treat 5:259-268, 2002
Yarbro JW, Page DL, Fielding LP, et al: American Joint Committee on Cancer Prognostic Factors Consensus Conference. Cancer 86:2436-2446, 1999
Page DL, Schuyler PA, Dupont WD, et al: Atypical lobular hyperplasia as a unilateral predictor of breast cancer risk: A retrospective cohort study. Lancet 361:125-129, 2003
Haagensen CD, Lane N, Lattes R, et al: Lobular neoplasia (so-called lobular carcinoma in situ) of the breast. Cancer 42:737-769, 1978
Arpino G, Allred DC, Mohsin SK, et al: Lobular neoplasia on core-needle biopsy: Clinical significance. Cancer 101:242-250, 2004
Page DL, Kidd TE Jr, Dupont WD, et al: Lobular neoplasia of the breast: Higher risk for subsequent invasive cancer predicted by more extensive disease. Human Pathol 22:1232-1239, 1991
Fisher ER, Land SR, Fisher B, et al: Pathologic findings from the National Surgical Adjuvant Breast and Bowel Project: Twelve-year observations concerning lobular carcinoma in situ. Cancer 100:238-244, 2004
Levi F, Te VC, Randimbison L, et al: Trends of in situ carcinoma of the breast in Vaud, Switzerland. Eur J Cancer 33:903-906, 1997
Franceschi S, Levi F, La Vecchia C, et al: Second cancers following in situ carcinoma of the breast. Int J Cancer 77:392-395, 1998
Haagensen CD, Bodian C, Haagensen DE: Lobular Neoplasia (Lobular Carcinoma In Situ) Breast Carcinoma: Risk and Detection. Philadelphia, PA, WB Saunders, 1981, p 238
Fisher ER, Costantino J, Fisher B, et al: Pathologic findings from the National Surgical Adjuvant Breast Project (NSABP) Protocol B-17: Five year observations concerning lobular carcinoma in situ. Cancer 78:1403-1416, 1996
Ottesen GL, Graversen HP, Blichert-Toft M, et al: Carcinoma in situ of the female breast. 10 year follow-up results of a prospective nationwide study. Breast Cancer Res Treat 62:197-210, 2000
Fisher B, Constantino JP, Wickerman DL, et al: Tamoxifen for prevention of breast cancer: Report of the National Surgical Adjuvant Breast and Bowel Project P-1 Study. J Natl Cancer Inst 90:1371-1388, 1998
National Cancer Institute: Surveillance, Epidemiology, and End Results (SEER) Program Public-Use CD-ROM (1973-1998). Bethesda, MD, National Cancer Institute, 1999
Percy C, Fritz A, Jack A, et al: (eds): International Classification of Diseases for Oncology (ed 3). Geneva, Switzerland, World Health Organization, 2000
Kaplan EL, Meier P: Nonparametric estimation from incomplete observations. J Am Statist Assoc 53:457-481, 1958
Lee ET: Statistical Methods for Survival Analysis. Belmont, CA, Lifetime Learning Publications, 1980
Snedecor GW, Cochran WG: Statistical Methods (ed 7). Ames, IA, The Iowa State University Press, 1980
Salvadori B, Bartoli C, Zurrida S, et al: Risk of invasive cancer in women with lobular carcinoma in situ of the breast. Eur J Cancer Clin Oncol 27:35-37, 1991
Akashi-Tanaka S, Fukutomi T, Nanasawa T, et al: Treatment of non-invasive carcinoma: Fifteen year results at the National Cancer Center Hospital in Tokyo. Breast Cancer 7:341-344, 2000
Rosen PP, Braun DW, Lynggholm B, et al: Lobular carcinoma in situ of the breast: Preliminary results of treatment by ipsilateral mastectomy and contralateral breast biopsy. Cancer 47:813-819, 1981
Rosen PP, Lieberman PH, Braun DW Jr, et al: Lobular carcinoma in situ of the breast. Am J Surg Pathol 2:225-251, 1978
Devarahally SR, Severson RK, Chuba P, et al: Second malignant neoplasms after primary central nervous system malignancies of childhood and adolescence. Pediatr Hematol Oncol 20:617-625, 2003
Hamre MR, Severson RK, Chuba P, et al: Osteosarcoma as a second malignant neoplasm. Radiother Oncol 65:153-157, 2002
Yap J, Chuba PJ, Thomas R, et al: Sarcoma as a second malignancy after treatment for breast cancer. Int J Radiat Oncol Biol Phys 52:1231-1237, 2002
Ottesen GL, Graversen HP, Blichert-Toft M, et al: Lobular carcinoma in situ of the female breast: Short-term results of a prospective nationwide study. Am J Surg Pathol 17:14-21, 1993
Wheeler JE, Enterline HT, Roseman JM, et al: Lobular carcinoma in situ of the breast: Long-term follow-up. Cancer 34:554-563, 1974
Bodian CA, Perzin KH, Latter R: Lobular neoplasia: Long term risk of breast cancer in relation to other factors. Cancer 78:1024-1034, 1996
Marshall LM, Hunter DJ, Connolly JL, et al: Risk of breast cancer associated with atypical hyperplasia of lobular and ductal types. Cancer Epidemiol Biomarkers Prev 6:297-301, 1997
Colditz GA, Rosner B: Cumulative risk of breast cancer to age 70 years according ot risk factor status: Data from the Nurses' Health Study. Am J Epidemiol 152:950-964, 2000
Sasson AR, Fowble B, Hanlon AL, et al: Lobular carcinoma in situ increases the risk of local recurrence in selected patients with stages I and II breast carcinoma treated with conservative surgery and radiation. Cancer 91:1862-1869, 2001
Stone JG, Coleman G, Gusterson B, et al: Contribution of germline MLH1 and MSH2 mutations to lobular carcinoma in situ of the breast. Cancer Lett 167:171-174, 2001
Goldstein NS, Kestin LL, Vicini FA: Clinicopathologic implications of E-cadherin reactivity in patients with lobular carcinoma in situ of the breast. Cancer 92:738-747, 2001
Rieger-Chris KM, Pezza JA, Dugan JM, et al: Disparate E-cadherin mutations in LCIS and associated invasive breast carcinomas. Mol Pathol 54:91-97, 2001
Lu YJ, Osin P, Lakhani SR, et al: Comparative genomic hybridization analysis of lobular carcinoma in situ and atypical lobular hyperplasia and potential roles for gains and losses of genetic material breast neoplasia. Cancer Res 58:4721-4727, 1998
Etzell JE, Devries S, Chew K, et al: Loss of chromosome 16q in lobular carcinoma in situ. Hum Pathol 32:292-296, 2001
Abner AL, Connolly JL, Recht A, et al: The relation between the presence and extent of lobular carcinoma in situ and the risk of local recurrence for patients with invasive carcinoma of the breast treated with conservative surgery and radiation therapy. Cancer 88:1072-1077, 2000
Zippin C, Lum BS, Hankey BF: Completeness of hospital cancer case reporting from the SEER program of the National Cancer Institute. Cancer 76:2343-2350, 1995(Paul J. Chuba, Merlin R. )