Concurrent lobular neoplasia increases the risk of ipsilateral breast cancer recurrence in patients with ductal carcinoma in situ treated with breast-conserving therapy

Authors

Errata

This article is corrected by:

  1. Errata: Correction Volume 116, Issue 8, 2064, Article first published online: 11 February 2010

  • Presented in part at the 2008 Annual Meeting of the American Society of Clinical Oncology, Chicago, Illinois, May 30 to June 3, 2008.

Abstract

BACKGROUND:

Multiple clinicopathologic factors have been analyzed for their association with an increased risk of ipsilateral breast tumor recurrence (IBTR) after women receive breast-conserving treatment (BCT) for ductal carcinoma in situ (DCIS). The reported incidence of proliferative lesions, such as atypical ductal hyperplasia (ADH), columnar cell changes (CCC), and lobular neoplasia associated with breast cancer, has been as high as 23%; however, the relevance of these lesions on the natural history of DCIS and the risk of IBTR remains unknown.

METHODS:

Two hundred ninety-four patients with DCIS who received BCT between 1991 and 1995 were identified from the authors' institutional database. Slides were reviewed by a dedicated breast pathologist with particular attention to the presence of lobular neoplasia, ADH, and CCC. The actuarial 5-, 10-, and 15-year IBTR rates were calculated using the Kaplan-Meier method and were compared using the log-rank test.

RESULTS:

Concurrent lobular neoplasia was present in 41 of 294 patients (14%), ADH was present in 37 of 294 patients (13%), and CCC was present in 71 of 294 patients (24%). The median follow-up was 11 years. IBTR occurred in 40 of 227 patients without lobular neoplasia (18%) versus 15 of 41 patients with lobular neoplasia (37%; P=.005; hazard ratio [HR], 2.49). The 5-, 10-, and 15-year cumulative incidence rates of IBTR were twice as high in women who had DCIS and lobular neoplasia compared with women who had DCIS alone (P=.002). Concomitant ADH (HR, 1.53) and CCC (HR, 1.24) were not associated significantly with IBTR (P=.20 and P=.44, respectively).

CONCLUSIONS:

Concurrent lobular neoplasia is associated with a significantly higher risk of IBTR in women with DCIS who received BCT. Women with coexisting DCIS and lobular neoplasia who receive BCT should consider using additional risk-reducing strategies. Cancer 2009. © 2009 American Cancer Society.

Because of the increased use of screening mammography, the number of women diagnosed with ductal carcinoma in situ (DCIS) per year in the US has risen to >67,000.1-4 The majority of these women will opt for breast-conserving therapy (BCT) over mastectomy, accepting the long period of follow-up and a higher local recurrence rate.5 Although the search for molecular markers that predict the biologic behavior of DCIS, including the risk of local recurrence in women who receive BCT, has proven unsuccessful to date, clinical research over the last 2 decades has established several clinicopathologic features that carry an increased risk of local failure: young age,6, 7 high nuclear grade,8 the presence of necrosis,9 positive or close margins,10 and the omission of radiation therapy.8, 11, 12

Recently, invasive breast cancer research efforts have questioned the role of the uninvolved ductal epithelium in breast cancer pathogenesis.13 Breast cancer arises as genetic aberrations accumulate in precursor epithelial cells. Although considerable information is available regarding the molecular alterations that characterize breast cancer cells, it has been appreciated only recently that similar abnormalities exist in histologically normal appearing epithelium. These abnormalities include allelic imbalance or loss of heterozygosity,14, 15 loss of insulin-like growth factor 2 imprinting,16 aberrant response to estrogen,17 loss of retinoic acid receptor β expression,18 aberrant phosphorylation of p38,19 and up-regulation of the enhancer of zeste homolog 2 protein EZH2,20 to name a few. Some of these abnormalities have been detected in normal appearing tissue adjacent to the tumor, and others have been detected at a distance from it. Some abnormalities are concordant, and others are discordant with the gene expression abnormalities of the tumors themselves.

It has been suggested that some of these genetic abnormalities harbor a distinct phenotype and develop over time into morphologically distinct lesions, such as columnar cell changes (CCC),21-23 atypical ductal hyperplasia (ADH),24, 25 or even lobular neoplasia.26-28 Therefore, the presence of concurrent proliferative lesions associated with in situ or infiltrating breast carcinoma may reflect underlying gene perturbations of cancer-related pathways in the uninvolved ductal epithelium, which could be markers of disease risk, occult disease, or, also, the tissue response to an existing tumor. The possibility that concurrent lobular neoplasia in early-stage invasive breast cancer has prognostic significance and may have an impact on the natural course of the disease has recently been suggested by the finding that the presence of lobular neoplasia was associated with an increased risk of ipsilateral breast tumor recurrence (IBTR) in women with invasive breast cancer who received BCT.29

It is in this setting that we undertook to investigate the biologic significance of concurrent lobular neoplasia, ADH, or CCC in noninvasive breast cancer. We hypothesized that the risk of IBTR in women with DCIS who received BCT is increased in the presence of lobular neoplasia, ADH, or CCC in the surrounding tissue. Our study objective was to determine the prognostic significance of a concomitant diagnosis of lobular neoplasia, ADH, and CCC in women with DCIS who received BCT on the risk of IBTR.

MATERIALS AND METHODS

After obtaining approval from the institutional review board, a review of institutional databases identified 304 women with DCIS who received BCT at Memorial Sloan-Kettering Cancer Center (MSKCC) from 1991 to 1995. Ten patients were excluded from the analysis because a rereview of pathology did not confirm the presence of DCIS without invasion, resulting in a population of 294 women.

All available histologic sections from each woman were reviewed by an MSKCC dedicated breast pathologist. In case of insufficient or unattainable material, original tissue blocks were reprocessed and new slides were created. All sections were reviewed for a comprehensive list of pathologic features, including margins (close margins were defined as tissue-free margins <1 mm), the presence of concomitant lobular neoplasia, ADH, CCC, architectural type, nuclear grade, and the presence of necrosis.

DCIS and lobular carcinoma in situ (LCIS) were diagnosed according to the World Health Organization Histological Typing of Breast Tumors 2nd Ed. guidelines.30 CCC included a spectrum of lesions in the terminal duct-lobular unit characterized by columnar epithelial cells with prominent apical cytoplasmic snouts, intraluminal secretions, and various degrees of nuclear atypia and architectural complexity. ADH lesions were defined as intraductal proliferations with atypia that exhibited some but not all of the features of DCIS. Lesions that were diagnosed as atypical lobular hyperplasia (ALH) were defined as lobular proliferations with some of the features of LCIS but with incomplete filling (<50%) of acini within the lobular unit. Lobular neoplasia was defined as either LCIS or ALH. Necrosis was defined as present if >10% of ducts contained degenerated cells characterized by amorphous, acellular, eosinophilic granular material from dying cells.

Radiation treatment was delivered as external beam radiation that consisted of conventional tangential breast irradiation. No patient received cytotoxic chemotherapy, and 18 patients (6.1%) received tamoxifen after the diagnosis of DCIS was made.

Patients were followed with routine physical examinations and yearly mammography. Clinical information was obtained from initial history questionnaires, including family history (number of first- and second-degree relatives affected by breast cancer), age, and clinical presentation. IBTR was defined as any subsequent DCIS or infiltrating carcinoma within the treated breast.

The number of patients in each category of clinicopathologic variable, treatment factor, and presence of proliferative lesion (lobular neoplasia, ADH, CCC) was calculated. For purposes of analyses, age, tumor size, and nuclear grade each were dichotomized. Specific attention was directed toward the following 3 covariates: presence or absence of lobular neoplasia, ADH, or CCC in any excision specimen. The Pearson chi-square test was used to compare the distribution of each clinicopathologic variable according to the 3 covariates of interest. Overall estimates of IBTR were calculated using the Kaplan-Meier method. The length of follow-up and the time to recurrence were defined relative to the date of definitive surgical excision of DCIS. The log-rank test and Cox regression analysis was used to examine whether recurrence-free curves were different among groups defined by each of the 3 covariates of interest and by other categorical demographic and clinicopathologic factors. A multivariate Cox proportional hazards model was fit, which included known risk factors of IBTR as well as lobular neoplasia, the only proliferative lesion variable that was significant in bivariate findings. For an exploratory analysis, a subset of patients with lobular neoplasia was assessed with the Kaplan-Meier method to determine the effect on IBTR of the presence of lobular neoplasia at the resection margin. All P values are 2-sided, and statistical significance was defined as P ≤ .05.

RESULTS

Patient demographic, clinical, pathologic, and treatment characteristics for the 294 women with DCIS who were included in the study are listed in Table 1. The median age at diagnosis was 55 years (range, 26-89 years). The median follow-up from the time of definitive surgery for those who did not develop recurrent disease was 11 years (range, 0-16 years).

Table 1. Patient Characteristics (N=294)
CharacteristicNo. of Patients%
Demographics
 Patient age at diagnosis, y
  <454615.6
  ≥4524884.4
 Family history
  No21874.2
  Yes6321.4
  Unknown134.4
Clinical presentation
 Mammographic findings
  Abnormal26790.8
  Normal175.8
  No mammogram10.3
  Unknown93.1
 Palpable mass
  No24884.4
  Yes4013.6
  Unknown62.0
 Nipple discharge
  No27994.9
  Yes93.1
  Unknown62.0
Adjuvant therapy
 Radiation
  No20068
  Yes9131
  Unknown31.0
 Tamoxifen
  No27392.9
  Yes186.1
  Unknown31.0
Histopathologic characteristics
 Tumor size, cm
  ≤1.018061.2
  >1.010134.4
  Unknown134.4
 Nuclear grade
  Low3612.2
  Intermediate13445.6
  High9532.3
  Unknown299.9
 Necrosis
  Absent12442.2
  Present16355.4
  Unknown72.4
 Margins
  Negative: >1 mm16455.4
  Positive83.1
  Close: ≤1 mm9231.3
  Unknown3010.2
Concurrent proliferative lesion
 Lobular neoplasia
  Absent22777.2
  Present4113.9
  Unknown268.8
 Atypical ductal hyperplasia
  Absent23078.2
  Present3712.6
  Unknown279.2
 Columnar cell changes
  Absent19967.7
  Present7124.1
  Unknown248.2

Ipsilateral Breast Tumor Recurrences

There were 63 IBTRs among the 294 women, corresponding to 5-, 10-, and 15-year actuarial recurrence rates of 17%, 22%, and 29%, respectively (Fig. 1). Thirty-nine patients developed a recurrence as DCIS (62%), 22 developed a recurrence as invasive breast cancer (35%), and 2 patients developed a recurrence with unknown DCIS/invasive status (3%).

Figure 1.

Actuarial overall ipsilateral breast tumor recurrences among women with ductal carcinoma in situ who received breast-conserving treatment (N=294). The 5-, 10-, and 15-year cumulative ipsilateral breast tumor recurrence incidence rates (95% confidence intervals) were 0.17 (0.12-0.22), 0.22 (0.17-0.27), and 0.29 (0.23-0.36), respectively.

On univariate analysis, younger patient age, palpable mass, the omission of radiation therapy, and the presence of lobular neoplasia were associated with an increased risk of IBTR (Table 2, Figs. 2 and 3A). The presence of close or positive margins was of borderline statistical significance (P=.052). Nuclear grade (P=.52) and necrosis (P=.66) were not associated with IBTR; however, 34% of patients with intermediate/high nuclear grade and 40% of patients with necrosis received radiation versus 17% of patients with low-grade DCIS and 21% of patients with no necrosis.

Figure 2.

Association of age (A), palpable mass (B), and radiation therapy (C) with the risk of ipsilateral breast tumor recurrence in patients with ductal carcinoma in situ who received breast-conserving treatment.

Figure 3.

Association of lobular neoplasia (A), atypical ductal hyperplasia (ADH) (B), and columnar cell changes (CCC) (C) with the risk of ipsilateral breast tumor recurrence in women with ductal carcinoma in situ who received breast-conserving treatment.

Table 2. Univariate Analysis of Patient Demographic, Pathologic, and Treatment Characteristics and the Risk of Ipsilateral Breast Recurrence
CharacteristicHR95% CIP
  1. HR indicates hazard ratio; 95% CI, 95% confidence interval.

Demographics
 Patient age at diagnosis, y
  <451.00  
  ≥450.540.30-0.96.04
 Family history
  No1.00  
  Yes1.300.75-2.25.34
Clinical presentation
 Mammographic findings
  Normal1.00  
  Abnormal1.750.75-4.06.20
 Palpable mass
  No1.00  
  Yes1.861.02-3.36.04
 Nipple discharge
  No1.00  
  Yes2.550.80-8.15.11
Adjuvant therapy
 Radiation
  No1.00  
  Yes0.380.20-0.72.003
 Tamoxifen
  No1.00  
  Yes0.200.03-1.41.11
Histopathologic characteristics
 Tumor size, cm
  ≤1.01.00  
  >1.00.900.52-1.57.72
 Nuclear grade
  Low1.00  
  Intermediate/high0.790.39-1.62.52
 Necrosis
  Absent1.00  
  Present0.890.54-1.48.66
 Margins
  Negative: >1 mm1.00  
  Positive or close: ≤1 mm1.671.00-2.80.052
Concurrent proliferative lesion
 Lobular neoplasia
  Absent1.00  
  Present2.581.42-4.67.002
 Atypical ductal hyperplasia
  Absent1.00  
  Present1.530.79-2.96.20
 Columnar cell changes
  Absent1.00  
  Present1.240.72-2.14.44

Radiation treatment was associated with a lower incidence of IBTR (Table 2) (Fig. 2C). The 10-year local recurrence rate for women who underwent surgery and received radiation therapy was 11.9% compared with 27.9% for women who underwent surgery alone. The median follow-up for patients who did and did not receive radiation treatment was 11.8 years and 7.8 years, respectively.

Effect of Concurrent Lobular Neoplasia, Atypical Ductal Hyperplasia, and Columnar Cell Changes on the Risk of Ipsilateral Breast Tumor Recurrence

The presence of concurrent lobular neoplasia was associated with a significantly higher risk of IBTR (Table 2) (Fig. 3A). There was a nonsignificant trend toward an association between the presence of ADH and, to a lesser degree, the presence of CCC with a greater risk of IBTR (Table 2) (Fig. 3B and 3C).

Lobular neoplasia was present in the excision specimens from 41 patients (14%) and was not identified in 227 specimens (77%). The median age of those with and without lobular neoplasia was 60 years (range, 34-79 years) and 55 years (range 26-89 years), respectively (P=.21). There was no statistically significant difference in the clinicopathologic or treatment features, including radiation therapy, between the groups with and without concurrent lobular neoplasia (Table 3). Patients who had lesions with coexisting ADH less frequently received radiation, more frequently received adjuvant tamoxifen, and less commonly displayed necrosis on histology. Patients who had lesions with CCC also more frequently received tamoxifen and less commonly had necrosis on histology.

Table 3. Comparison of Clinicopathologic and Treatment Features According to the Presence of Lobular Neoplasia, Atypical Ductal Hyperplasia, or Columnar Cell Changes
 Lobular Neoplasia: No. of Patients (%)*PAtypical Ductal Hyperplasia: No. of Patients (%)PColumnar Cell Changes: No. of Patients (%)P
PresentAbsentPresentAbsentPresentAbsent
  • *

    Unknowns were excluded in the calculation of percentages.

Demographics
 Patient age at diagnosis, y
  <453 (7.3)39 (17).116 (16)34 (15).8213 (18)27 (14).34
  ≥4538 (93)188 (83) 31 (84)196 (85) 58 (82)172 (86) 
 Family history
  No28 (68)175 (81) 30 (83)168 (77) 52 (75)147 (78) 
  Yes13 (32)41 (19).076 (17)51 (23).3817 (25)42 (22).68
  Unknown011 111 210 
Clinical presentation
 Mammographic findings
  Abnormal37 (90)207 (94) 35 (100)210 (94) 67 (94)179 (94) 
  Normal4 (9.8)12 (5.5).530 (0)14 (6.3).134 (5.6)11 (5.8).83
  No mammogram0 (0)1 (0.5) 0 (0)0 (0) 0 (0)1 (0.5) 
  Unknown07 26 08 
 Palpable mass
  No35 (85)190 (86) 26 (77)197 (87) 59 (84)167 (86) 
  Yes6 (15)32 (14).978 (24)30 (13).1111 (16)27 (14).71
  Unknown05 33 15 
 Nipple discharge
  No39 (95)216 (97) 34 (97)219 (97) 68 (97)188 (97) 
  Yes2 (4.9)7 (3.1).571 (2.9)7 (3.1).942 (2.9)6 (3.1).92
  Unknown04 24 15 
Adjuvant therapy
 Radiation
  No32 (80)148 (66) 32 (87)148 (65) 46 (65)137 (70) 
  Yes8 (20)77 (34).085 (14)79 (35).0125 (35)59 (30).43
  Unknown12 03 03 
 Tamoxifen
  No37 (90)212 (95) 31 (84)218 (96) 63 (89)189 (96) 
  Yes4 (9.8)12 (5.4).286 (16)9 (4).0038 (11)7 (3.6).02
  Unknown03 03 03 
Histopathologic characteristics
 Tumor size, cm
  ≤1.029 (73)139 (62).1822 (60)144 (65).5349 (69)119 (62) 
  >1.011 (28)87 (39) 15 (41)78 (35) 22 (31)72 (38).32
  Unknown11 08 08 
 Nuclear grade
  Low5 (13)30 (13).974 (11)31 (15).5812 (18)23 (12) 
  Intermediate/high33 (87)194 (87) 32 (89)182 (85) 55 (82)162 (88).27
  Unknown33 177 414 
 Necrosis
  Absent21 (51)97 (43) 26 (70)89 (39) 44 (62)72 (37) 
  Present20 (49)129 (57).3311 (30)137 (61).00127 (38)123 (63).0003
  Unknown01 04 04 
 Margins
  Negative: >1mm20 (53)129 (62) 27 (73)131 (60) 50 (71)109 (58) 
  Close: ≤1 mm16 (42)74 (35) 9 (24)81 (37) 19 (27)71 (38).12
  Positive2 (5.3)6 (2.9).501 (2.7)8 (3.6).301 (1.4)8 (4.3) 
  Unknown318 010 111 

IBTR occurred in 15 of 41 patients (37%) with concurrent lobular neoplasia versus 40 of 227 patients (18%) without lobular neoplasia (P=.001) (Fig. 3A). The 5-, 10-, and 15-year cumulative incidence of IBTR was twice as high in women who had both DCIS and lobular neoplasia compared with women who had DCIS alone (P=.002) (Table 4). Women with concomitant ADH had more IBTR (hazard ratio, 1.5), but the difference was not statistically significant (P=.2) (Fig. 3B).

Table 4. Cumulative Incidence and Hazard Ratios for Ipsilateral Breast Tumor Recurrence
Variable5-YearCumulative Incidence*15-YearHRP
10-Year
Estimate95% CIEstimate95% CIEstimate95% CI
  • 95% CI indicates 95% confidence interval; HR, hazard ratio.

  • *

    Based on Kaplan-Meier curve.

  • Based on Cox proportional hazard regression.

Lobular neoplasia
 Absent0.140.09-0.190.190.17-0.250.240.17-0.31  
 Present0.260.11-0.410.360.20-0.520.550.33-0.772.577.002
Atypical ductal hyperplasia
 Absent0.160.11-0.210.210.15-0.270.280.20-0.36  
 Present0.220.08-0.360.250.10-0.400.410.21-0.611.533.20
Columnar cell changes
 Absent0.170.11-0.230.210.15-0.270.270.19-0.35  
 Present0.170.08-0.260.240.13-0.350.370.22-0.521.239.44

Multivariate analysis indicated that concomitant lobular neoplasia was associated with an increased risk of IBTR in women with DCIS who received BCT after adjusting for other risk factors (P=.005) (Table 5). The risk associated with concurrent lobular neoplasia was >2-fold that of women without concurrent lobular neoplasia and was similar to the risk associated with not receiving radiation.

Table 5. Multivariate Analysis for Risk of Ipsilateral Breast Tumor Recurrence*
VariableAdjusted HR95% CIP
  • HR indicates hazard ratio; 95% CI, 95% confidence interval.

  • *

    Only variables that were identified as significant on univariate analysis were included in the multivariate model. There were 260 observations used and 55 events (recurrences) in the multivariate model.

Age at diagnosis, y
 <451.00  
 ≥450.500.26-0.97.04
Palpable mass
 No1.00  
 Yes2.051.10-3.81.02
Radiation
 No1.00  
 Yes0.330.17-0.67.002
Concurrent lobular neoplasia
 No1.00  
 Yes2.491.33-4.67.005

Impact of Lobular Carcinoma in Situ Versus Atypical Lobular Hyperplasia on the Risk of Ipsilateral Breast Tumor Recurrence

There were 20 women who had concurrent LCIS with 10 recurrences, and 18 women had concurrent ALH with 4 recurrences in the study population. The 10-year cumulative incidence rate of IBTR among women who had DCIS and LCIS was 47% compared with 21% among women who had DCIS and ALH (P=.08).

Impact of Lobular Neoplasia Margin Status on the Risk of Ipsilateral Breast Tumor Recurrence

Of the patients who had concurrent lobular neoplasia, 63% had lobular neoplasia at the final margin. There were 12 IBTRs in these 26 women (46%) compared with 3 IBTRs in 12 women (25%) with no lobular neoplasia at the final pathologic margin (P=.18) (Fig. 4).

Figure 4.

Ipsilateral breast tumor recurrence among women who had ductal carcinoma in situ and lobular neoplasia at or close to the resection margin compared with women who had margins that were clear of lobular neoplasia.

DISCUSSION

With the increased use of screening mammography, there has been a dramatic increase in the number of women diagnosed with DCIS. Although DCIS mortality is low, local recurrence rates are higher than those associated with invasive carcinoma, and approximately 50% of all recurrences are invasive. With the increasing use of BCT, it is estimated that $65 million currently is spent annually in the US for surveillance of previously excised DCIS.31 Although several molecular markers, such as cell surface-associated mucin 1 (epithelial membrane antigen),32 p53,33 vascular endothelial growth factor,33 p16 and Ki-76,34 and the v-src sarcoma viral oncogene homolog c-SRC,35 have demonstrated promise in risk stratification for IBTR among women with DCIS who receive BCT, their low positive and negative predictive values have hampered the translation of their use into clinical practice.22, 36 The objective of the current study was to determine whether lobular neoplasia or other proliferative lesions observed on routine histopathology in conjunction with DCIS are associated with higher IBTR rates in women who receive BCT.

The presence of concurrent proliferative lesions in invasive and noninvasive breast cancer is not an infrequent event, occurring in up to 23% of patients.29, 37-40 However, its impact on the natural course of the disease remains largely unknown. The reported prevalence of concurrent proliferative lesions associated with DCIS is 15%,39 and concomitant LCIS reportedly occurs in 14% of patients with DCIS.40 This is similar to the current series, in which we identified lobular neoplasia in 41 of 294 patients (14%) with DCIS who received BCT. These rates also are similar to those described for patients with early-stage invasive breast cancer and LCIS (5%-12%).29, 37, 38 It is not known, however, whether DCIS associated with concurrent lobular neoplasia, ADH, or CCC represents subgroups with different biologic behavior.

In the current series, the presence of lobular neoplasia was associated with 5-, 10-, and 15-year cumulative incidence rates of IBTR of 0.26, 0.36, and 0.55, respectively. In contrast, patients without lobular neoplasia had cumulative incidence rates of only 0.14, 0.19, and 0.24, respectively, and this association was independent of other clinical, pathologic, and treatment factors by multivariate analysis. These included age, radiation therapy, and clinical presentation (palpable mass). The presence of necrosis and nuclear grade was not associated with IBTR in our univariate and multivariate analyses. Solin et al reported that, in a large cohort of patients with DCIS who received BCT and definitive radiation, after long-term follow-up (>10 years), initial differences in the IBTR rate between patients who had high nuclear grade and necrosis and patients who had tumors with favorable histopathology disappeared because of the ongoing incidence of late recurrences in the favorable histopathology group.41

Our finding that lobular neoplasia was associated with an increased IBTR risk in women with DCIS who received BCT is consistent with recent publications regarding the impact of lobular neoplasia on IBTR in women with invasive breast cancer. In 1 of the largest series to date, Sasson et al recently demonstrated that patients with LCIS and early-stage invasive breast cancer had a significantly increased risk of IBTR29; IBTR occurred in 57 of 1209 of patients (5%) who did not have LCIS compared with 10 of 65 of patients (15%) who had LCIS (P=.001). Griem et al reported a 5-year actuarial breast recurrence rate of 17% in patients with LCIS who underwent conservative surgery and received radiation compared with 5% in patients without LCIS.42 Moran and Haffty reported a 10-year actuarial breast recurrence rate of 23% in 51 patients with invasive carcinoma and LCIS who underwent BCT versus 17% in 1045 women without LCIS (P=.05).37 Jolly et al examined 607 women with invasive breast cancer who received BCT, including 56 women (9%) who had LCIS in association with invasive cancer43; in that study, the 10-year IBTR rate was 14% for women who had LCIS versus 7% for women without LCIS (P=.04). In contrast, in another series, Abner et al reported an 8-year crude rate of IBTR of 13% for 137 patients with invasive carcinoma and LCIS versus 12% for 1062 patients without LCIS.38 Although those authors concluded that the presence or extent of LCIS was not associated with an increased risk of IBTR, they acknowledged several shortcomings of their study. Because of the long study accrual period from 1968 to 1986, no centralized pathology review was possible, and interpretation depended on several independent pathologists over time, which may have led to significant under-reporting and case misclassification. In addition, 67% of patients in that series had unavailable data on resection margins, and 39% of patients who had evaluable margins had margins that were positive for DCIS or invasive carcinoma.

Adepoju et al, in 1 of the few series on DCIS and concomitant high-risk proliferative lesions, recently reported on the relation between the presence of lobular neoplasia, ADH, or both and the risk of IBTR in patients with DCIS.40 In that study, there were only 14 patients with concurrent ALH and 11 patients with concurrent LCIS (25 of 307 patients; 7.7%). The 10-year actuarial local recurrence rate reported for DCIS and concomitant ADH, lobular neoplasia, or both was 8% versus 15% without concurrent proliferative lesions (P=.38). However, those authors acknowledged several limitations of their findings. Because of the prolonged study period (1968-1998), their analysis was based almost entirely on pathology reports only. No centralized pathology rereview was possible for the majority of cases, and significant under-reporting and misclassification may have occurred over time.

In a nationwide, prospective Danish study of in situ carcinoma of the breast, 275 women who underwent excision alone were registered from 1982 to 1989.39 In that series, the risk of invasive recurrence after DCIS excision did not differ statistically between those with (26 patients) and without (142 patients) concurrent LCIS. Specimen margin status, which is a major determinant of outcome, was not included in that analysis.

The influence of lobular neoplasia margin status on IBTR in patients with DCIS and concurrent lobular neoplasia who receive BCT remains to be defined. In the series by Sasson et al,29 no IBTRs in 9 women who had LCIS at the final surgical margin occurred compared with a 35% 10-year cumulative IBTR incidence rate in 56 women who had surgical margins that were negative for LCIS. The influence of LCIS at the surgical margins on IBTR rates was not evaluated in the series by Moran and Haffty37 or Abner et al.38 Fisher et al suggested that LCIS-negative surgical margins in patients with LCIS- only registered on the National Adjuvant Surgical Breast and Bowel Project Protocol B-17 trial may have been responsible for the lower than expected rate of subsequent recurrence.11 Despite the limited clinical data addressing the influence of LCIS margin status on the risk of IBTR, the recently discovered, shared molecular abnormalities between DCIS and LCIS and the existence of potential common precursors for both DCIS and LCIS44 raise the possibility that margin status may be relevant. Moinfar et al45 and Boecker et al46 reported that genetic changes can precede the development of breast precursor lesions, and Lakhani et al44 recently reviewed similar findings for LCIS. Furthermore, a recent immunophenotypical analysis of >1000 breast lesions, including DCIS, ADH, CCC, typical epithelial hyperplasia, and lobular neoplasia, suggested a common route of pathogenesis.47 Molecular and genetic aberrations in the uninvolved ductal epithelium may develop over time into phenotypically distinct precancerous lesions, including CCC, ADH, or lobular neoplasia. Therefore, our finding that lobular neoplasia is associated with a higher risk of recurrent disease may hint at a more cancerous potential of the entire ductal epithelial ‘field’ in these patients. Its prognostic significance and biologic information, as demonstrated here, warrant further study.

In conclusion, the current study demonstrated the association of lobular neoplasia with a >2-fold greater risk of IBTR in patients with DCIS who received BCT, even after adjusting for important clinicopathologic features. Although the findings were not statistically significant, the data suggest that women who have concomitant DCIS and LCIS my have an increased risk of IBTR compared with women who have DCIS and ALH, and that the risk of local recurrence possibly may be decreased by the complete excision of lobular neoplasia. Concurrent lobular neoplasia may be a phenotypic manifestation of more extensive genetic abnormalities, resulting in a higher risk for local recurrence. Further study of the molecular basis of these findings is warranted, and future clinical studies of women with DCIS who receive BCT should include assessment of the surrounding epithelium. Patients with DCIS and concurrent lobular neoplasia may particularly benefit from use of risk-reducing strategies, including radiation and antiestrogen treatment.

Conflict of Interest Disclosures

The authors made no disclosures.

Ancillary