Positive DCIS Research Articles

Oestrogen-induced genes in ductal carcinoma in situ: their comparison with invasive ductal carcinoma

It is well known that oestrogens play important roles in both the pathogenesis and development of invasive ductal carcinoma (IDC) of human breast. However, molecular features of oestrogen actions have remained largely unclear in pure ductal carcinoma in situ (pDCIS), regarded as a precursor lesion of many IDCs. This is partly due to the fact that gene expression profiles of oestrogen-responsive genes have not been examined in pDCIS. Therefore, we first examined the profiles of oestrogen-induced genes in oestrogen receptor (ER)-positive pDCIS and DCIS (DCIS component (DCIS-c)) and IDC (IDC component (IDC-c)) components of IDC cases (n=4 respectively) by microarray analysis. Oestrogen-induced genes identified in this study were tentatively classified into three different groups in the hierarchical clustering analysis, and 33% of the genes were predominantly expressed in pDCIS rather than DCIS-c or IDC-c cases. Among these genes, the status of MYB (C-MYB), RBBP7 (RBAP46) and BIRC5 (survivin) expressions in carcinoma cells was significantly higher in ER-positive pDCIS (n=53) than that in ER-positive DCIS-c (n=27) or IDC-c (n=27) by subsequent immunohistochemical analysis of the corresponding genes (P<0.0001, P=0.03 and P=0.0003 respectively). In particular, the status of C-MYB immunoreactivity was inversely (P=0.006) correlated with Ki67 in the pDCIS cases. These results suggest that expression profiles of oestrogen-induced genes in pDCIS may be different from those in IDC; and C-MYB, RBAP46 and survivin may play important roles particularly among oestrogen-induced genes in ER-positive pDCIS.

OT1-02-05: A Phase III Clinical Trial Comparing Trastuzumab Given Concurrently with Radiation Therapy to Radiation Therapy (RT) Alone for Women with HER2−Positive DCIS Resected by Lumpectomy: NSABP B-43.

Abstract Background Because a substantial portion of DCIS is ER negative and overexpresses HER2, therapy targeting this protein is a promising strategy for HER2−overexpressing DCIS. Preclinical studies have shown that trastuzumab (T) boosts the effectiveness of RT in xenograft models and in cell lines with no detrimental effect on irradiated HER2−normal cells. Studies correlating clinical response with molecular markers in T-treated patients show that apoptosis occurs within 1 wk of starting singleagent T, with little effect on proliferation. Shorter duration treatments with this agent require investigation. Adjuvant trials using T during breast irradiation have already provided ample safety evidence. Will T administered during WBI improve lumpectomy + WBI results in women with HER2−positive DCIS? This trial will allow us to better understand the biology of breast cancer and its prevention and will extend the benefits of breast-conserving surgery for women with DCIS. Trial Design: Post lumpectomy for DCIS without evidence of an invasive component, a central review of each patient's pure DCIS lesion is carried out for HER2 by IHC analysis. If the HER2 is 2+, FISH analysis is done, and patients whose tumors are HER2 3+ or FISH positive can be randomly assigned to receive 2 doses of T 3 wk apart during WBI or to receive WBI alone. Eligibility criteria: Women 18 years or older with an ECOG status of 0 or 1 who have undergone a margin-clear lumpectomy for DCIS and whose tumors are clinically or pathologically node negative are eligible. DCIS must be HER2 positive by central testing. ER and/or PR status must be known before random assignment. Specific aims: The primary aim is to determine if T given concurrently with WBI is more beneficial in preventing IBC recurrence, ipsilateral skin cancer recurrence, or ipsilateral DCIS compared with WBI alone for HER2−positive DCIS resected by lumpectomy. Secondary aims are to compare the possible benefit of T given during WBI to that of WBI alone in preventing regional or distant recurrence and contralateral invasive or DCIS breast cancer. B-43 will determine if invasive or DCIS DFS, recurrence-free interval, and OS can be improved with the addition of T to WBI. The effects of T on ovarian function in premenopausal women will also be assessed. Statistical methods and accrual: Our design calls for accrual of 2000 patients during a 7.9-year period. As of May 31, 2011, 578 patients have been entered. A definitive analysis of primary endpoints will be performed when 163 ipsilateral breast cancer events occur (7.5 and 8 years after protocol initiation). This number of events affords 80% power to detect a hazard reduction of 36%, from 1.73 ipsilateral breast cancer events per 100 patient-years to 1.11 events per 100 patient-years. The 36% observed reduction in the hazard of IIBCR-SCR-DCIS on the T arm is based on a projection of 40% hazard reduction if the compliance were perfect, with a 10% noncompliance rate. Supported by PHS grants NCI-U10-CA-69651, NCI-U10-CA-12027, and NCI P30-CA-14599 from the US NCI and Genentech, Inc. Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr OT1-02-05.

P4-08-04: Preoperative Thrombin Pathway Activation Identifies Node Positive Patients in Early Breast Cancer.

Abstract Activation of the thrombin pathway (TP) is seen in 50% of cancers and plays a significant role in promoting metastasis. Cancer-induced TP activation is exacerbated by surgery, and this activation may be greater in the presence of metastases. Our aim was to determine if pre-operative plasma d-dimer, a marker of TP activation, correlated with the presence of LN metastases in early breast cancer. We also examined whether surgery-induced TP activation is affected by tumour-associated factors. Methods Plasma d-dimer was measured using automated ELISA pre-operatively and at days 1 and 42 following surgery in 103 prospectively recruited patients undergoing surgery for early breast cancer(87 invasive,13 DCIS). D-dimer values were log-transformed and correlated with LN metastases, tumour size and grade and hormone receptor status using parametric statistical tests. Results Median age was 68 years (range 46–82). Mean tumour size was 14.6 mm (range 7.5-40). Seventy invasive cancer patients had wide local excision(WLE) and sentinel node biopsy(SNB) and 17 had mastectomy(MX) +/− axillary clearance(ANC). Twenty-two patients had positive lymph nodes at primary surgery (MX 8,WLE 14). Of these 3(14%) were ER-ve and 19(86%) were ER+ve. Mean pre-operative d-dimer was higher (p&amp;lt;0.01) in LN positive (704ng/ml, 95% confidence interval, CI, 455–1088) versus LN node negative invasive cancers (443ng/ml, CI 304–627) or DCIS (366ng/ml, CI 284–471). D-dimer was also higher in patients with lymphovascular invasion (612ng/ml, CI 423–886) than in those without (454ng/ml, CI 385–534) and this approached significance (p=0.084). Post-operative d-dimer was higher (p&amp;lt;0.05) after MX compared to WLE at all timepoints. Following WLE + SNB for invasive cancer, post-operative d-dimer rose and was higher at day 1 and day 42 in LN positive vs LN negative and in ER negative versus ER positive cancers (see table). Repeated measures ANOVA analysis showed that the change in d-dimer over time was significantly different between subjects according to LN status (p=0.037) or ER status (p=0.015). Only one patient in the ER negative group had LN metastases. There was no association between pre or postoperative d-dimer and tumour size, tumour grade or PR status. Conclusion Pre-operative TP activation and the thrombotic response to surgery (as measured by d-dimer) is greater in the presence of LN metastases regardless of tumour size. ER negative cancers (7/8 of which were LN negative) also resulted in a greater d-dimer rise, suggesting a difference in the thrombotic response to surgery between different cancer phenotypes. Ongoing work will determine if peri-operative TP activation can act as an independent predictor of node status and cancer recurrence. Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P4-08-04.

OT1-01-01: Prospective Clinical Trial Evaluating Efficacy of Zoledronic Acid (ZA) Prophylaxis for Prevention of Aromatase Inhibitor Associated Musculoskeletal Symptoms: ZAP-AIMSS Trial.

Abstract Brief background: Aromatase inhibitor associated musculoskeletal symptoms (AIMSS) occur in approximately 50% of patients receiving AIs (Henry 08). However, interventions to prevent or treat AIMSS have not been established. In a retrospective study (Muslimani 09), patients receiving bisphosphonates along with AIs were less likely to report AIMSS compared to those not taking bisphosphonates (35% vs. 60%). However, the efficacy of bisphosphonates in reducing incidence of AIMSS has not been studied prospectively, so it cannot be recommended for routine clinical practice. Trial design: We are conducting a single arm, phase II clinical trial of 4 mg intravenous zoledronic acid (ZA) given at baseline and at 6 months, in combination with letrozole 2.5 mg daily for one year. Development of AIMSS will be assessed using the standardized Health Assessment Questionnaire (HAQ-DI) and pain Visual Analog Scale (VAS) at baseline,1, 3, 6, and 12 months. Secondary endpoints include mammographic breast density (when intact contralateral breast), bone mineral density, bone turnover metabolites, circulating inflammatory markers, and patient reported quality of life measures. Prevalence of AIMSS will be compared to historical controls from a recently completed multi-institutional study designated Exemestane and Letrozole Pharmacogenetics (ELPh trial, ClinicalTrials.gov #NCT00228956). The current study has the same eligibility criteria, method and intervals of outcome assessment, and AI medication, as the ELPh trial, ensuring that the two cohorts are comparable. Eligibility criteria: Postmenopausal women who have completed local therapy and chemotherapy for hormone receptor positive DCIS or stage I-III breast cancer and who are scheduled to receive adjuvant AI. Prior tamoxifen therapy is permitted. Specific aims: 1. Percentage of women experiencing AIMSS at 1, 3, 6, and 12 months after initiation of ZA and letrozole, as compared to historical controls. 2. Change in bone mineral density and breast density between baseline and 12 months for those receiving ZA and letrozole, as compared to historical controls. 3. Change in bone turnover markers and inflammatory markers between baseline and 1, 3, 6 and 12 months for those receiving ZA and letrozole, as compared to historical controls. Statistical methods: Allowing for a 20% dropout rate, a total sample size of 59 patients yields 80% power to detect reduction in AIMSS incidence from 50% to 30% with a two sided type I error rate of 5%. The rates of AIMSS and other endpoints at each time point and across all time points between controls and patients will be compared with a logistic regression model that adjusts for potential confounding variables and include random effects as appropriate to account for correlation between outcomes in the same patient. Present accrual and target accrual: The Johns Hopkins Institutional Review Board approved the study and it opened to accrual in January 2011. Since that time, 12 participants have signed consent and started therapy, and 2 have completed the 3 month evaluation. Funding: Trial supported by BCRF. ZA and letrozole kindly supplied by Novartis. Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr OT1-01-01.

Tamoxifen added to radiotherapy and surgery for the treatment of ductal carcinoma in situ of the breast: A meta-analysis of 2 randomized trials

BackgroundSurgical excision with adequate margins is the treatment of choice for ductal, in situ carcinoma of the breast (DCIS). The addition of radiotherapy (RT) halved local in situ and invasive recurrence. The purpose of our meta-analysis is to evaluate the reduction in recurrence (in situ or invasive) with the addition of tamoxifen (T), in particular in patients with DCIS treated with surgery+RT. Patients and methodsThe eligible studies (NSABP-B24 and UK ANZ DCIS trials) included prospective, randomized, controlled trials in which the addition of T had been compared with surgery+RT without T in women with DCIS of the breast. Relative risks (RRs) with 95% confidence intervals (CIs) were calculated for both in situ and invasive recurrence (local and controlateral). ResultsTamoxifen does not reduce breast cancer-specific or overall mortality when added to loco-regional therapy for DCIS of the breast (surgery plus or minus RT). Tamoxifen reduces overall breast cancer recurrence by 29% in all patients and by 33% in those treated with both surgery and RT. Only ipsilateral invasive (RR 0.61 [95% CI 0.41, 0.92]; p=0.02) and controlateral in situ relapses (RR 0.40 [95% CI 0.16, 0.96]; p=0.04) are significantly lowered when T is added to RT. Tamoxifen seems to exert a local synergistic effect with RT. Both young and older women (< and >50years) achieve some benefit from the addition of T (RR 0.6 and 0.74, respectively). ConclusionThe addition of T to surgery and RT for DCIS of the breast reduces the risk of local invasive and controlateral in situ relapses, but not the survival. The benefit is independent of age. In conclusion, surgery associated with RT and T is the treatment of choice for patients with (estrogen-receptor positive) DCIS of the breast.

Abstract P2-06-17: Biological Study of the Effects of GW572016 (Lapatinib) in Ras/Raf/MAK and PI3K/Akt Signaling Pathways in HER2 Positive Ductal Breast Carcinoma in Situ (DCIS)

Abstract Background: DCIS is a heterogeneous disease, with increasing incidence. Up to 37% of DCIS overexpress HER2. Lapatinib is a tyrosine kinase inhibitor that directly inhibits HER2 receptor signaling by blocking the receptor's kinase activity. This work tested the hypothesis that Lapatinib inhibits Ras/Raf/MAK and PI3K/AKT signaling in HER2 positive DCIS. Patients and Methods: Patients with HER2 positive (FISH rate &amp;gt;2.2) DCIS diagnosed by core needle biopsy with low risk of having a hidden invasive component were eligible. Neoadjuvant Lapatinib 1500mg daily was administered for 4 weeks prior to surgical resection. EGFR was analyzed by FISH in the biopsy specimen. Ki67, PTEN, AKT, pAKT and pERK were evaluated by IHQ. PI3KCA mutations were screened at exons 9 and 20 by PCR. Apoptosis was analyzed by TUNEL assay. All biological markers were measured before and after Lapatinib treatment. Results: Eleven pts with HER2 positive DCIS have been included to date. Median FISH amplification was 7 (range 2, 70-14). Evaluation pre-Lapatinib showed that: none of the pts had EGFR amplification; six pts (54%) had a Ki67 &amp;gt;25%; pAKT was overexpressed in 3 pts (27%) and pERK in 2 pts (18%); PTEN was deleted in 3 pts (27%). PI3KCA mutations were detected in only 1 pt (9%). After Lapatinib treatment, 9 pts were evaluable. We observed a trend for inactivation of the Ras/Raf/MAK signaling pathway in 7 pts (70%) patients. Interestingly, 3 of these pts had paradoxical activation of PI3K/Akt suggesting a compensatory feed back loop. One out of these 3 pts has PTEN deleted. In addition, 3 pts (33%) showed an increase in the apoptotic index. Conclusions: In our study, four weeks of Lapatinib against HER2 positive DCIS resulted in inactivation of Ras/Raf/MAK pathway; however an activation of the PI3K/Akt was seen as the same time in 3 pts. Nevertheless this trend needs to be confirmed with the inclusion of more pts. Citation Information: Cancer Res 2010;70(24 Suppl):Abstract nr P2-06-17.

Abstract P1-03-12: The First Reproducible Model of Primary Human Ductal Carcinoma In Situ (DCIS) Using the Mouse Intraductal (MIND) Method

Abstract Past studies of the molecular mechanisms regulating human DCIS progression to invasion have been limited by a lack of suitable human models. The current models have not utilized primary human DCIS placed in their natural microenvironment, i.e., inside the mammary ducts, and do not mimic the heterogeneity of human DCIS. Previously, we reported the development of a mouse xenograft model using established breast DCIS cell lines, SUM-225 and DCIS.COM by the mouse intraductal (MIND) method. MIND involves injection of human DCIS cells into the primary ducts of immunodeficient mouse mammary glands via the nipple. In this study we demonstrate the reproducible growth of human primary DCIS cells derived from patient biopsy samples using the MIND model. Core biopsies from consenting patients are processed to isolate the epithelial cell component and injected. To date, 69 patients have consented to provide core biopsy tissue and we have received tissue from 38. Of these, 14 samples have been confirmed as positive for DCIS, and 2 as atypical hyperplasia, based on the pathology report of adjacent biopsies. The biopsy samples were heterogeneous with respect to ER, PR, Ki67, Her-2 and histology. A total of 24 mammary glands from immunodeficient mice have been injected with these cells and analyzed for growth by immunofluorescent (IF) with antibodies directed at human-specific cytokeratin-5 and/or 19 and smooth muscle actin. Table 1. Summary of grewlh of human primary biopsies in (tie MIND xenograft The average number of cells recovered per mg of tissue are shown in Table 1. Biopsies diagnosed as normal yielded only 91±34 cells/mg compared with most DCIS cases that showed an average cell number of 971 per mg tissue. Our results indicate that human DCIS and hyperplasia cells from biopsy samples are capable of growing within the NOD-SCID IL2Rgammanull (NSG) mouse mammary ducts. NSG mice mammary ducts are more suitable hosts for the growth of human primary DCIS cells, compared to SCID-Beige. Furthermore, the higher number of cells recovered following digestion predicted the positive DCIS growth rate. DCIS cells formed single and multi-layered epithelium inside the ducts and were heterogeneous with respect to the expression of human specific cytokeratins. The MIND xenografts recapitulated the patient's original DCIS as evidenced by IF staining for the biomarkers ER and Her-2 in the primary human biopsies and MIND xenografts. These results provide the first reproducible model of primary human DCIS for studying the temporal processes of early breast cancer progression. Future experiments will characterize the cellular basis for the subtypes of DCIS and delineate the distinct molecular and cellular mechanisms of early breast cancer invasion, and cancer stem cell growth and self-renewal. Citation Information: Cancer Res 2010;70(24 Suppl):Abstract nr P1-03-12.

Abstract P1-15-02: Impact of Retinoblastoma (RB) Tumor Suppressor Pathway on Ductal Carcinoma In Situ (DCIS) Recurrence

Abstract The retinoblastoma (RB) tumor suppressor is functionally inactivated at high frequency in human malignancies including breast cancer. By immunohistochemistry loss of RB function is associated with p16ink4a overexpression and high Ki67 proliferation index.We have previously shown that loss of RB results in dramatic upregulation of genes associated with tumorigenesis and is a key determinant of therapeutic response in invasive breast carcinoma. Recently publish data indicate that RB pathway dysregulation plays an important role in DCIS progression to invasive carcinoma. The aim of this study was to evaluate the expression p16ink4a and Ki67 (indicative of RB loss) in a large cohort of DCIS patients treated at one institution with wide-excision and close follow-up. Material and Methods 149 DCIS patients were included in the study. Expression of p16ink4a and Ki67 was assessed by employing a standard immunoperoxidase method with ani-p16ink4a (MTM Laboratories Westborough, MA) and anti-Ki67 (AbCam Cambridge, MA) antibodies. The p16ink4a staining in DCIS epithelium was scored semi-quantitatively as negative (0; no staining), weak (1;staining in less than 25% of cells), moderate (2; defined as staining in 25-75% of cells) and strong (3; &amp;gt;75% of cells staining). For statistical analysis, the original p16ink4a scores (0, 1, 2, 3) were dichotomized as low (0,1) vs. high (2-3) and original ki67 scores were dichotomized as low (&amp;lt;= 10%) vs. high (&amp;gt; 10%). Stromal p16 staining was scored as absent or present. Association between p16ink4a and Ki67 and nuclear grade, necrosis, hormone receptor and Her2 expression as well as recurrence was assessed using Fisher's exact test. The Cox proportional hazard model was used for analyses of time to recurrence. Results There was a statistically significant association between high p16ink4a and Ki67 expression in DCIS epithelium and triple negative as well as Her2 positive DCIS subtypes. Significant association was detected between high expression of p16 and Ki67 in DCIS epithelium and DCIS recurrence (p=0.005). Stromal expression of p16 was the strongest predictor of DCIS recurrence (p &amp;lt; 0.0001). In multivariate Cox model DCIS cases with high expression of p16 and Ki67 in epithelium and stromal p16 expression had the shortest time to recurrence. Conclusion RB pathway dysregulation in DCIS epithelium as well as stromal expression of p16ink4a indicative of senescence are associated with DCIS recurrence. Citation Information: Cancer Res 2010;70(24 Suppl):Abstract nr P1-15-02.

Method of detection of DCIS to predict HER2 overexpression.

1557 Background: Overdiagnosis of DCIS has become a major concern. Overexpression of the human epidermal growth factor receptor-2 (HER2) has been introduced as a prognostic marker that correlates with a higher biologic aggressiveness of breast cancer. We investigated the rate of HER2 overexpression of DCIS diagnosed by breast MRI compared to the DCIS diagnosed by mammography. Aim was to find out whether the additional DCIS diagnosed by MRI represents prognostically relevant disease or whether it will lead to additional overdiagnosis. Methods: During a 5-year period 7,319 women underwent breast MRI in addition to mammography for screening or for diagnostic assessment. A total of 167 women received the final surgical pathology diagnosis of pure DCIS. HER2 receptor status was established by standardized methods and dichotomized as positive (DAKO score 3+ or c-erbB2-amplifikation at FISH) or negative (DAKO score < 3+ or FISH negative). It was available in 127/137 women. We investigated the detection rate of breast MRI and of mammography for DCIS. We compared the HER2 receptor status of all MRI-detected versus all mammography-detected DCIS. Furthermore we compared the HER2 receptor status off only mammography-detected versus only MRI-detected DCIS. Results: Of the 127 DCIS, 95 (75%) were categorized as HER2 negative, and 32 (25%) as HER2 positive. Of the 32 HER2 positive DCIS, 13 (41%) were diagnosed through mammography, and 31 (97%) through MRI. Of the 10 DCIS which were only mammography-detected, one (10%) was HER2 positive. Of the 52 DCIS which were only MRI detected, 21 (40%) were HER2 positive. In the entire cohort, two-thirds of HER2-positive DCIS (66%, 21/32) were only detected by MRI (p < 0.005). Conclusions: DCIS not visible at MRI is virtually always HER2 negative. Two-thirds of HER2 positive DCIS are not visible on mammography, but only diagnosed by MRI. The method of detection of a DCIS correlates with its biologic aggressiveness: Detectability by mammography is correlated with a lower aggressiveness compared with detectability by MRI. MRI-detected additional DCIS are associated with a higher biologic aggressiveness. We conclude that MRI-detected DCIS do not add to mammography-induced overdiagnosis, but represent prognostically relevant disease. No significant financial relationships to disclose.

Biological Markers Predictive of Invasive Recurrence in DCIS

DCIS is a heterogeneous group of non-invasive cancers of the breast characterized by various degrees of differentiation and unpredictable propensity for transformation into invasive carcinoma. We examined the expression and prognostic value of 9 biological markers with a potential role in tumor progression in 133 patients with pure DCIS treated with breast conserving surgery alone, between 1982–2000. Histology was reviewed and immunohistochemical staining was performed. Pearson correlation coefficient was used to determine the associations between markers and histopathological features. Univariate and multivariate analysis examined associations between time to recurrence and clinico-pathologic features and biological markers. Median age at diagnosis was 55 years (25–85). With a median follow up of 8.91 years, 41/133 patients recurred (21 as invasive recurrence). In this cohort 13.5% had low, 43% intermediate and 42% high nuclear grade. Comedo necrosis was found in 65% of cases. Expression of ER (62.4%), PR (55.6%), HER2/neu (31.6%), MIB1 (39.8%), p53 (22.6%), p21 (39.8%), Cyclin D1 (95.5%) calgranulin (20.5%), psoriasin (12%), was found in DCIS. HER2/neu was overexpressed in 45% that recurred as DCIS and 42.9% that recurred as invasive cancer, and only in 26.1% in cases that never recurred. On univariate analysis, HER2/neu overexpression was the only marker associated with an increased risk for any recurrence (p = 0.044). The hazard ratio for recurrence for HER2/neu positive DCIS was 1.927 (confidence interval 1.016–3.653) compared to HER2 negative DCIS. On multivariate analysis, HER2/neu overexpression remained the only independent variable significantly associated with any recurrence (p = 0.014) and with invasive recurrence (p = 0.044). This data suggest that HER2/neu testing may become an important parameter in the management of DCIS and the treatment of cases with positive HER2/neu status could be modified accordingly, similar to the current approach for HER2/neu positive invasive disease.

DCIS and aromatase inhibitors

In patients with hormone receptor positive DCIS tamoxifen reduces recurrence rates by almost 50%. Few data are available with aromatase inhibitors from randomised studies. In the ATAC study there were three DCIS lesions in the anastrozole arm and four in the tamoxifen arm in the women with ER positive invasive cancer. In the MA17 study which randomised patients to up to 5 years of letrozole or placebo there was only one DCIS event in the contralateral breast in patients taking letrozole and five on placebo. There were also four patients in this study who had DCIS in the conserved breast on placebo and none in the letrozole treated group. The few clinical data that are available therefore suggest the aromatase inhibitors are likely to be effective in DCIS. A histological review of a study of 206 postmenopausal women with invasive oestrogen receptor positive breast cancer who were randomised as part of a 14 day preoperative study to receive 2.5 mg of letrozole or 1 mg of anastrozole identified 27 patients with 28 pairs of tumours in whom there was sufficient ER positive DCIS in invasive cancer in the initial core biopsy and in the subsequent surgery specimen, to evaluate for PgR activity and proliferation. Within the DCIS both aromatase inhibitors significantly reduced PgR expression and both drugs also produced a significant fall in proliferation. There was a moderate degree of agreement between the fall in PgR in both the invasive cancer and DCIS (Kappa = 0.5; p = 0.0013) and between the fall in proliferation and between the invasive and in situ components (correlation coefficient = 0.68; p < 0.001). This study has shown significant effects of aromatase inhibitors on DCIS indicating that these agents are therapeutically active in this condition.

BS18 DCIS AND TREATMENT

Ductal carcinoma in situ now represents 30% of all screen detected breast cancers. Breast conserving surgery for ductal carcinoma in situ is now established although the extent of margin clearance necessary to prevent recurrence after breast conserving surgery remains controversial. None of the previous trials which compared radiotherapy with breast conserving surgery alone have achieved clear margin status in all patients and analysis of their data suggests that the recurrence rate when clear margins around the ductal carcinoma in situ are achieved is low (approximately 10% at six years). Data from randomised trials indicates that a margin of greater than 1 mm clearance is sufficient to minimise recurrence in the breast. Several centres have published that clear margins are essential in the management of ductal carcinoma in situ whether radiotherapy is used or not.Tamoxifen has been studied after wide local excision and radiotherapy in the NSABP24 trial and in the UK DCIS trial. In the former trial a 40% reduction in recurrence in the breast using tamoxifen was seen but the majority of this effect was in women under 50 years of age and only a marginal effect was seen in older patients. In the UK DCIS trial a 20% non‐significant reduction in recurrence was seen in women who were not given radiotherapy. Studies show that younger women (less than 50 years of age) have a significantly higher risk of recurrence after ductal carcinoma in situ treatment and tamoxifen is therefore recommended in women under 50 years of age who are undergoing wide local excision for ductal carcinoma in situ. Oestrogen receptor status was retrospectively assessed in the NSABP B24 trial and ER positive DCIS had a 60% reduction in recurrence whilst ER negative DCIS had no benefit. The BASO DCIS II trial assesses the benefit of adding radiotherapy in ER positive DCIS treated with 5 years Tamoxifen on local recurrence. Aromatase inhibitors are more effective in the presence of CerbB2 oncogene, which is frequently expressed in ductal carcinoma in situ and studies comparing aromatase inhibition with Tamoxifen in ductal carcinoma in situ are already underway (IBIS II). These studies should take account of oestrogen receptor status in the ductal carcinoma in situ to allow us to more adequately define the role of oestrogen receptor status in predicting response to therapy in patients with ductal carcinoma in situ.

Can intraductal breast carcinoma be excised completely by local excision? Clinical and pathologic predictors.

Microscopic evaluation of excised intraductal breast carcinoma (DCIS) specimens using a serial subgross technique reveals that in many patients the lesion is larger than expected, often making complete excision impossible with less than a true quadrantectomy. Data is presented on 181 patients with DCIS in whom the initial biopsy was performed using a more cosmetic wide local excision rather than a true quadrantectomy. Clear margins were defined as no tumor within 1 mm of any inked or dyed margin. All of these patients subsequently underwent mastectomy or reexcision of the initial biopsy site. This allowed pathologic evaluation for residual disease. At mastectomy or reexcision, 76% of patients with initially involved margins had residual DCIS, as did 43% of patients with initially clear margins (P < 0.0001). Larger tumor size was a statistically significant predictor of initial margin involvement and residual DCIS (P < 0.05). Patients with comedo-DCIS had a greater tendency toward positive initial histologic margins and residual DCIS, but this trend was not statistically significant (P < 0.1). DCIS presents major problems to both surgeons and pathologists. It is difficult to excise completely using a wide local excision. Histologically negative margins do not guarantee that residual DCIS has not been left behind. Inadequate excision of the primary lesions may be the most important cause of local failure after conservative treatment for intraductal breast carcinoma.